1 Family Court Review April 2005 Debate Master

EMPIRICAL AND ETHICAL PROBLEMS WITH CUSTODY..., 43 Fam. Ct. Rev.
193

43 Fam. Ct. Rev. 193
Family Court Review
April, 2005
Symposium Issue on Child Custody Evaluations
EMPIRICAL AND ETHICAL PROBLEMS WITH CUSTODY RECOMMENDATIONS: A CALL FOR CLINICAL
HUMILITY AND JUDICIAL VIGILANCE
Timothy M. Tippins
a1
J effrey P. Wittmann
a2

Copyright 2005 by Association of Family and Conciliation Courts; Timothy M. Tippins, J effrey P. Wittmann
This article proposes a four-level model of clinical inferences to analyze the psychological evaluation process in custody
matters. At each level the authors summarize the status of the relevant psychological literature and conclude that, as
clinicians respond to the ultimate issues (e.g., who should be the custodial parent) the empirical foundation for such
conclusions is tenuous or non-existent. A jurisprudence argument is also made that such opinions should be routinely
excluded from the fact-finding process. Given the significant potential for specific custody recommendations to limit
personal liberties and the trajectory of a childs life, the paucity of relevant research available in this area, and profound
evidentiary issues, such recommendations should be viewed as ethically inappropriate. A model for what clinicians can
ethically say to courts is proposed.

INTRODUCTION
Forensic psychological assessments in contested custody matters are often pivotal documents that can have a dramatic effect
on the trajectory of the litigation and, ultimately, on the form a particular childs life will take after judicial disposition. The
courts afford culturally sanctioned weight to behavioral science, and these documents are often eagerly awaited because of
their potential value in providing leverage for one side over the other and for their capacity to move cases toward stipulation.
Indeed, recent empirical work has confirmed the high perceived value of clinicians involvement in matrimonial matters by
both attorneys and judges (Bow & Quinnell, 2004) despite earlier findings that suggested a less favorable view by legal
professionals (Felner, Rowlison, Farber, Primavera, & Bishop, 1987; Melton, Weithorn, & Slobogin, 1985).

Most relevant to the current review is the finding by Bow & Quinnell (2004) that one of the components most valued by legal
consumers is a specific recommendation for custody and that the overwhelming majority of judges and attorneys believe that
psychologists should directly address the ultimate issue before the court. This latter belief is incongruent with the ongoing
controversy about whether evaluators should make such specific recommendations, a controversy that caused the American
Psychological Association itself to hedge on the issue in its Guidelines For Child Custody Evaluations In Divorce
Proceedings (APA, 1994) and that has caused some scholars to draw the conclusion that this practice should cease altogether
due to the serious limits of the psychology disciplines knowledge base on custody topics (ODonohue & Bradley, 1999;
Melton, Petrila, Poithress, & Slobogin, 1997; Tippins, 2003; Wittmann, 2003) and due to philosophical and legal arguments
against ultimate issue testimony in general (e.g., Heilbrun, 2001).

This review will briefly summarize the general boundaries of our psychological knowledge regarding variables relevant to
making custody recommendations. We will then *194 outline the jurisprudential and evidentiary problems regarding
psychological opinions about custody matters, followed by an outline of what we would propose should be the parameters of
ethical practice in this area. The summary will be organized initially around a four-level model for understanding clinical
inferences in custody matters.

CLINICAL INFERENCE HIERARCHY
2014 Thomson Reuters. No claim to original U.S. Government Works. 1

Cal. Coalition for Families and Children v. San Diego County Bar Ass'n
13 cv 1944 CAB BLM
CCFC Master Exhs. P3594
EMPIRICAL AND ETHICAL PROBLEMS WITH CUSTODY..., 43 Fam. Ct. Rev. 193

In an attempt to simplify, for discussion purposes, the very complex clinical process involved in custody assessments, we
would propose the following stratification of the data and inferences garnered by clinicians in these matters:

Level I (What the clinician observes). This category includes anything that the clinician observes with his/her senses, without
the addition of higher-level abstractions about the observations. Included in this category would be comments in a report such
as She hung her head low and was often tearful, Aaron clung to his mothers leg throughout our session, There is an
elevation on scale two of the MMPI-2, and Father would not compliment mother in any way regarding her parenthood.

Level II (What the clinician concludes about the psychology of a parent, child, or family). This category includes inferences
and higher-level abstractions about what was observed in Level I without reference to custody/best-interests constructs.
Examples would include the following: J acquelines attachment to her father appears quite insecure, Ms. J ones has strong
tendencies toward substance abuse, Fathers style of parenting is very authoritarian, or Mothers blue mood, the childs
report of her chronic sadness, and her MMPI-2 elevations suggest the conclusion that she is moderately depressed.

Level III (What the psychologist concludes about the implications of Level II conclusions for custody-specific variables). At
this level, the clinician typically uses an even higher level of abstraction about the case, making reference to custody-specific
constructs such as global parenting capacity, potential psychological risks of primary custody with father, the fit
between a childs needs and a parents parenting capacity, and so forth. Level II inferences about individual or family
psychology (Mother is depressed, There is an overly close father--son alliance against mother, etc.) are used in Level III
to make inferences specifically relevant to custody and access issues (e.g., Multiple days away from mother, this childs
primary psychological parent, will be emotionally stressful, There is a poor fit between J asons need for parental empathy
regarding his ADHD and his fathers blunt and authoritarian approach, etc.). At this level, therefore, the psychologist is still
not making overt comments about what should happen in a family (as in the shoulds of Level IV), but the inferences drawn
begin to have more clear connections to the ultimate issues before the court and therefore, by definition, also begin to have
greater potential impact on judicial decisions regarding important personal liberties (e.g., the right to have access to ones
children, the right to do so unencumbered by supervision, the right to make decisions regarding a childs medical care, etc.).

Level IV (The psychologists conclusions about the custody-related shoulds in the matter). Here, the clinician takes the
Level III inferences and presents prescriptive conclusions about what access schedule and other plan parameters (e.g.,
supervised contact, etc.) should be put in place by the court. Although Level II and Level III include *195 a certain level of
value judgment on the clinicians part (e.g., It is better for a child to not display depressive or aggressive symptoms than to
display these symptoms), it is at Level IV that the value judgments become much more overt, prescriptive, sweeping, and
potentially life-changing for the litigants. There is a shift at this level from lower-level assessments of the whats in the
matter, to taking overt positions about the shoulds (with the implication exuded by the expert witness status that
behavioral science has been brought to bear on the process of achieving a specific, prescriptive recommendation about a
childs postdivorce life).

THE LIMITS OF CUSTODY-RELEVANT PSYCHOLOGICAL KNOWLEDGE
Level I
At this very basic level of observation and information/reporting, mental health professionals bring specialized knowledge to
the table in the form of (a) a scientific method for gathering data that should be more orderly/cautious than what would be
used by a lay person and (b) clinical decision rules about what observations may have psychological significance. They arrive
at the clinical room with the same sense organs as nonexperts yet can interpret what is seen and heard through the lenses of
an empirical and clinical knowledge base so that early psychological hypotheses can be developed. For example, whereas a
lay person may hear a dad say, J ason and I think his mom has a problem, and make little of the comment, a clinician will
note the behavior for later consideration due to the possibility of father--son alignment.

Interestingly, it is at this level that the psychologist may be able to offer some of the most useful information to the finder of
fact who has to sort through the dueling distortions of family life that often characterize custody petitions. As highly trained
interviewers and listeners, psychologists can set the stage of the clinical setting to maximize self-disclosure about issues that
may be disputed, yet important to a best-interests decision. For example, the simple and cautious recording of

13 cv 1944 CAB BLM

best-interests-related admissions on the part of parents, or reports about family life on the part of children, can provide a
valuable glimpse into how the family functions outside of the polarizing effects of litigation (e.g., hearing a father admit that
he actually did strike mother in front of the children, hearing a childs tearful report of abuse at the hands of a drunken
mother, listening to and recording a teenagers fervent desire to remain close to friends in a particular school system, etc.). It
is the Level I observations that can assist in creating a more efficient fact-gathering process for a court and that form the basis
of an investigative function for custody evaluators that can be quite valuable.

The limitations of our knowledge at Level I are best discussed in the context of the rich literature on clinical judgment error.
Although we are not aware of any substantive research on the clinical accuracy and objectivity of custody evaluators, there is
a long line of judgmental heuristics research to suggest that even at the level of basic data gathering, clinicians can be
extremely prone to distortions of what they observe due to various cognitive biases, attribution effects, labeling effects,
illusory correlations, flawed estimation rules, and so on (e.g., Garb, 1998; Turk & Salovey, 1988). We now also have data
raising questions about the accuracy of what is actually recorded by clinicians during forensic evaluations (Lamb, Orbach,
Sternberg, Hershkowitz, & Horowitz, 2000) that highlights how weighty ultimate conclusions can sometimes rest on clinical
sand.

Although such judgmental distortions are now well documented, and while both clinician and lay person share these
distorting tendencies, it is the clinician who makes observations *196 equipped with a scientific method for processing what
is observed, a method that is designed to at least reduce the effects of these cognitive errors. There is evidence that when
clinicians appropriately apply diagnostic strategies forcing them to consider alternative hypotheses (the strategy taught in
scientist-practitioner training programs) they become more accurate in their case assessments (e.g., Tutin, 1993). We believe,
therefore, that the problem in the area of custody assessment is not whether mental health professionals have evaluative
knowledge or methods that supercede those of lay persons. Rather, the problem is whether this knowledge is overextended
into questionable higher-level inferences and whether the scientific method is consistently brought to bear on what is
observed in order to correct for ubiquitous judgment errors and to avoid having an inappropriate effect on personal liberties.

Level II
At this level the clinician begins to interpret and combine the observations made at Level I (observations, test results, record
review, collateral contacts) to make inferences about the presence, absence, or severity of general psychological constructs,
inferences that are made based on strands of consistent and contradictory data within the case itself. The body of knowledge
that can be brought to bear here essentially represents the universe of empirically supported psychological constructs about
individual and family functioning. This includes, but is not limited to, inferences about psychopathology, basic and discrete
parenting skills and skills-deficits, intellectual/cognitive functioning, developmental status, developmental variables
(needs) potentially relevant to access plans (e.g., age-differences in time perception), child-temperament variables,
substance abuse tendencies, attachment constructs (e.g., primary psychological parent), interpersonal style, criminality,
domestic violence tendencies, adequacy of parent--child boundaries, available social support, impulse control, and
family-level constructs (emotional boundaries, enmeshment, etc.).

The capacity for psychological tests to validly and reliably assess a host of individual and family functioning variables is well
established and, in certain areas, rivals the accuracy of medical tests (Meyer et al., 2001). Even more germane to issues
related to custody and access is the empirical studies elucidating relevant aspects of the parent--child relationship as they
relate to positive and negative child outcomes. For example, there is a literature on the association between certain
traits/disorders in parents and adjustment problems in children (Waxler, Duggal, & Gruber, 2002), on the link between
specific parenting styles and positive child outcomes (Teti & Candelaria, 2002), and on the association between substance
abuse patterns and negative child outcomes (Mayes & Truman, 2002). As this line of research has advanced in sophistication
and complexity away from simple notions of parents affecting children in a unidirectional manner, literature has developed
around the bidirectional nature of the parent--child relationship and the ways in which children impact adults psychologically
(Patterson & Fisher, 2002). There are now well-established lines of empirical evidence that allow circumscribed conclusions
about secure/insecure attachment patterns (e.g., Rothbaum, Weisz, Pott, Miyake, & Morelli, 2000) and the association
between certain attachment patterns and psychological adjustment problems (Rubin & Burgess, 2002).

We now have a relatively robust literature about the short and long-term symptomatic effects of parental divorce on children
(e.g., Hetherington & Kelly, 2002) and about a host of factors (e.g., diminished parenting quality and socioeconomic status)

13 cv 1944 CAB BLM

that may mediate some of these effects (Hetherington & Stanley-Hagen, 2002). The negative effects on *197 children of
parental hostility and violence are also now well documented and researchers have begun to analyze in more detail the factors
that diminish or amplify the negative effects (e.g., Cummings & Davies, 1994).

This literature allows clinicians to draw circumscribed conclusions about the likely meaning of a childs symptoms (The
interview data and collateral contacts suggest the oppositional behavior on this childs part that is alleged to be the result of
harsh behavior on fathers part is also consistent with the symptom profile that is commonly associated with normative
divorce-related adjustment challenges) and about family variables that impact on children (The narrative descriptions of
frequent angry exchanges, combined with comments by the children and with the data from police reports suggest that the
relationship between these parents is very prone to overtly hostile exchanges). In addition, clinicians can serve the court at
this level by applying a rigorous hypothesis-testing approach to strands of family data so that cautious psychological
conclusions can be drawn to assist the finder of fact in sorting through contradictory allegations (e.g., Although mother
admits to shyness and father asserts that she is disabled by social anxiety, the psychological testing, combined with both
mothers presentation and comments by her children and her therapist do not indicate that her social anxiety is seriously
limiting her interpersonal functioning.). Finally, the judicious use of psychological language to articulate the stressful
circumstances that litigating families often find themselves in can assist the finder of fact in achieving a more humanized
understanding of the emotional nuance behind petitions and counter-petitions.

The limitations of psychological knowledge at Level II are largely related to the limited usefulness of the data at this level to
draw justifiable conclusions about custody-related Level III constructs, and the procedural dilemma of whether such data
should be mentioned at all if it cannot be reliably connected to such constructs. That is, a clinician has a wealth of
psychological literature and case-specific observations that can be opined about at Level II (dad is quick tempered, mom is
authoritarian, dad is lacking in empathy skills, mom is irrationally suspicious, this child has a reactive temperament, etc.).
Such information could easily be used, within a clinical relationship that is therapeutic or non-court related, to attempt to
assist a client with their personal or family problems even if many of the inferences are only hypotheses that rest on
significant clinical uncertainty. However, a forensic custody evaluation is conducted in a milieu where the words that are
chosen (a) will be interpreted and used by nonclinicians and (b) may be used to truncate important human liberties for parents
and, by association, for children. It therefore appears that the most ethically sensitive approach to Level II inferences is to
avoid their misuse by mentioning only those that on the basis of convergent, case-specific information and the specialized
knowledge base of the profession will be able to be reliably tied to Level III, custody-specific variables (i.e., to global
parenting capacity, global statements about a childs psychological needs, parent--child fit, risk factors associated with
different access plans, etc.). As we will assert below, this means that we must be ethically circumspect at Level II: Despite
the wealth of clinically relevant information that can be gleaned at Level II, much of it should often be withheld because of
the unique socio-cultural milieu in which the forensic expert is functioning. For example, although testing, interviews, and
records support the conclusion that a parent displays serious obsessive-compulsive traits, obsessive compulsive should be
highlighted in the forensic clinicians conclusions only if the clinician can infer, based on confirmed case data and the
specialized knowledge base of the profession, that these traits can be reliably associated with custody-specific inferences
(diminished parenting capacity, poor fit with a childs unique emotional needs, etc.).

*198 We are in agreement with ODonohue and Bradley (1999) that there are profound definitional, assessment, reliability,
and validity problems associated with the interview protocols and psychological tests used in custody matters that mean that,
even at Level II, many erroneous inferences are likely presented to courts on a regular basis. There is no reliably established
association between traditional psychological test findings (MMPI-2, Rorschach, etc.) and different parenting
styles/competencies, although we do know that test profiles predict certain states of parental turmoil and maladjustment, and
studies find associations between these states and child adjustment problems (Waxler, Duggal, & Gruber, 2002). Other than a
small amount of literature on associations between certain MMPI-2 profiles and an increased risk of child abuse (summarized
by Pope, Butcher, & Seelen, 2000), we are unaware of any robust line of research that allows the prediction of parenting
behavior through the use of traditional personality instruments. There is no custody-related, structured child interview
protocol that is viewed as having acceptable psychometric properties. There are also severe threats to valid inferences from
parent--child observation sessions due to the reactivity of the parties to the unique custody evaluation setting, unreliable
coding systems, constricted behavioral sampling, and due to the lack of research-based guidance regarding which
parent--child relationship variables can be viewed as reliably predicting different levels of child adjustment under different
access plans (ODonohue & Bradley, 1999). However, researchers and theorists are beginning to offer guidance to
practitioners about clinical principles that increase the validity of inferences derived from parent--child sessions (Hynan,

13 cv 1944 CAB BLM

2003).

One of the most significant gaps in the knowledge base that forensic clinicians bring to custody matters is related to base
rates and normal distributions of various child, parent, and child-parent relationship variables (ODonohue & Bradley, 1999).
For example, a common assumption among mental health professionals is that a parents capacity to empathize with, and
emotionally understand a child is a critically important skill. However, we do not have sufficient data on how the capacity
for empathy is distributed in the population to help a court understand whether the specific level of empathy-deficit in a
particular parent--child dyad should be viewed as so extreme as to warrant a preference for one access plan over another. The
empathy a child experiences from her father may appear greater than what she experiences from her mother, yet full
normative data might show that the mother--child empathy pattern still falls within normal limits. If it does, should this be
weighted against the mother? Certainly, this is more of a socio-moral decision than a clinical one. We are even lacking
information about the base rates of various kinds of parenting skills in the general population, raising the same kind of
challenging problem: We might note for a court that a particular father appears extremely passive with regard to limit setting,
yet we have little literature to guide us in deciding whether his level of passivity should be viewed as statistically extreme
enough to be relevant to best interests. The clinician is often left with personal values and clinical experience to make this
decision and we believe that these sources of clinical guidance are, in most cases, wholly inappropriate as a basis for
suggesting access plans that might limit certain personal liberties (i.e., by constricting parental access, constricting a childs
access to a parent, etc.).

Our trial consultations have made it evident that, quite unfortunately, many forensic clinicians are willing to opine about
Level II inferences (a) despite being ultimately unable to make the connection to Level III or (b) more problematically, with a
willingness to state a connection to Level III that represents clinical speculation rather than empirically sound conclusions.
This problem is nowhere more evident than with psychological testing: clinicians are often willing to include extended, test
interpretative narratives despite being *199 unable to make an empirically supportable connection between the many
adjectives used to describe a parent or child and custody-relevant variables (leaving posturing attorneys or judges to decide if
they should view a particular description as worrisome). Many clinicians are also unaware of the literature on normative
elevations on personality tests among custody litigants (e.g., Bagby, Nicholson, Buis, Radovanovic, & Fidler, 1999; Butcher,
1997). They therefore over-interpret scale elevations to the detriment of the litigant despite the fact that certain scale profiles
should actually be adjusted downward because of these base rate findings (McCann et al., 2001). Finally, despite the
admirable efforts by certain forensic experts to develop instruments specifically designed for custody and access issues
(Bricklin, 1989, 1990a, 1990b, 1992; Ackerman & Schoendorf, 1992), most forensic reviewers agree that there are profound
definitional, validity, and reliability questions that need to be resolved before these instruments should be viewed as
psychometrically sound for measuring the Level III, ultimate issue-related constructs that they were designed to measure
(Brodzinsky, 1993; Otto, Edens, & Barcus, 2000). However, it does appear to us that these instruments may hold great
potential value for generating hypotheses about psychological variables at Level II (e.g., for gathering information about such
issues as parenting skills, how a parent is perceived by a child, a parents intellectual competence, etc.).

At this inferential level psychological constructs are often presented as scientifically discovered fact without full disclosure
regarding the lack of professional consensus about such constructs. Clinicians are willing to easily include comments about
worrisome enmeshment, a childs lack of individuation from a parent, primary attachment figure and so on while not
stating that there continues to be vigorous debates about the meaning and implications of such constructs. J udges and
attorneys are usually not aware that many of the psychological constructs that they read about, and that they use to form
strategy/opinions, are not always sturdy, enduring truths and are often not transgenerational or transcultural. For example,
forty years ago a parents homosexuality was defined as a psychiatric illness whereas now it is viewed as an orientation.
The importance of a child achieving independence and individuation from a parent essentially represents a North American
value-judgment that is not made as frequently in other cultures (Rothbaum, Weisz, Pott, Miyake, & Morelli, 2000). A family
viewed as enmeshed to a worrisome degree by a clinician might be viewed as evidencing a loyalty and emotional closeness
that is honored in another culture. Unfortunately, few clinicians make courts aware that the constructs they use to argue for
different custody plans are plastic over time, very value-laden, and can change in importance depending on ones geographic
longitude and latitude.

Level III
As clinicians begin to draw inferences regarding custody-specific constructs they use language that can have substantial

13 cv 1944 CAB BLM

impact on the trajectory that a custody matter takes because such inferences, at least at a theoretical level, are more closely
related to the ultimate issue before the finder of fact. A statement in a clinical report should be defined as Level III if it
conveys conclusions about aspects of a childs psychological best interests, potentially nudging the consumer of the report in
certain directions regarding access planning. Conclusions about the fit between a childs needs and a parents skills, the
global, relative parenting capacities of the two parents, custody-specific test results about preferred parent (Bricklin, 1984)
or custody quotients (Ackerman & Schoendorf; 1992), and the psychological risks/benefits of different custody plans all
have more *200 perceived relevance than Level I and Level II to the custody and access decisions before the court (insofar as
these constructs begin to limit the field of psychologically preferable access plans for the consumer of the report). However,
such statements are still in the realm of whats because they ostensibly are still conveying conclusions about the
psychological state-of-affairs of the evaluated family without specifically stating what the court should do.

The American Psychological Association, in its guidelines for custody evaluations (APA, 1994), begins to suggest a
nomenclature for Level III in its conclusion that the fit between a childs psychological functioning and developmental
needs and a parents functional ability to meet these needs should be the focus of custody assessments. It is at this level
where the clinician begins to make a leap from Level II inferences regarding such matters as security of attachment, parenting
skills, developmental needs, and so on to opinion statements about the goodness of fit between what was concluded about the
child at Level II (psychological functioning and developmental needs) and what was concluded about the parents at Level II
(functional skills of parents). It is, therefore, at this level of inference that the clinician begins to have a more potentially
substantial impact on the ultimate decision (and, therefore, we would argue that the ethical mandate for the conclusions to be
empirically sound and grounded in specialized knowledge of the profession becomes substantially more paramount).

We would argue that the only ethical statements that can be made by clinicians at Level III include a summary for the court
of notable psychological risks and/or advantages associated with various access plans, but that such statements can only be
made with clearly articulated qualifications, cautionary statements to the court, and references to the limitations of the
evaluation methods used. It will be argued below that the statements that clinicians can make at this level that are within
acceptable ethical and scientific parameters are extremely limited. However, we also believe that helpfulness to the finder
of fact (Melton et al., 1997) as a guiding principle suggests that child-focused, yet constricted, statements about potential
risks/advantages, as long as they are grounded in case-specific facts and reliable empirical literature, represent a forensic
work product that is ethical, useful to the court, and potentially valuable to both the child in question and society at large.
Examples of the kinds of constricted, risk-focused conclusions include the following: (1) Primary placement with the father
in this matter includes the risk of long-term exposure to Mr. Smiths chronic tendencies toward derogatory and violent
behavior with women. (2) Primary placement with mother includes the risk of substantial, weekly exposure to her now,
well established and self-admitted tendencies toward alcohol abuse while under stress, tendencies that interrupt her ability to
meet this toddlers needs for careful supervision. (3) J onathan has been diagnosed with severe ADD and a mild psychotic
disorder that all clinicians agree interferes with his academic and social functioning--a notable risk associated with primary
placement with his father, therefore, involves the fact that Mr. Smith is openly resistant to giving his son the medications that
appear to improve his emotional adaptation. (4) A risk factor associated with primary placement with mother includes her
chronic inability to set firm and predictable limits on her eight-year-old sons behavior, as indicated by her behavior in
session, the record review, and consultation with the childs therapist--an inability that is especially relevant given the fact
that J ason has begun to engage in delinquent behavior in the community. Placement with father appears to hold the advantage
of firm and predictable limit setting for J onathan, a parental pattern that has been established in the research literature to have
a positive association with better child outcomes than more passive approaches.

*201 Such a risk-focused approach does not take an affirmative stance about whether the specific risks should be
determinative and allows substantial room for negotiation, as well as for the finder of fact to make the ultimate, socio-moral
value judgments about whether such risks should be attended to in the final access plan and how they should be weighted. It
also rests on an implicit value judgment that we should make explicit: the value to children and families of minimal
invasiveness by mental health professionals except where case-specific and research-based data can reliably suggest
important, child-relevant risks/advantages.

Unfortunately, with the exception of such constricted statements of risk/advantage, we conclude that there is little that can be
reliably said by mental health professionals at Level III. The concept of the adequacy of fit between a childs needs and a
parents capacity has not been operationalized by the psychology discipline; that is, individual clinicians are left to their own
idiosyncratic definitions of what adequate fit means. There is a quite paltry, yet growing, empirical literature on reciprocal

13 cv 1944 CAB BLM

effects between parents and children that might be relevant to custody (Patterson & Fisher, 2002) but no one to our
knowledge has tested these constructs across different access plans. There is no established metric for assessing the goodness
of parent-- child fit and there has been no attention in the literature to the ways in which parent--child mismatches might be
associated with eventual, positive child outcomes (a mothers superior capacity for emotional attunement may align nicely
with the mental health disciplines emphasis on empathy and compassion, whereas the father-CEOs high academic
expectations may produce other valuable achievement-focused outcomes for a child that are less valued by the same
discipline).

Although the substantially increased complexity of our understanding of child and family functioning has caused the
somewhat simplistic notion of primary psychological parent (Goldstein, Freud, & Solnit, 1973) to be less emphasized in
recent years, we continue to find it present as an operative construct in a fair number of custody assessments. It is important
to emphasize that, at least as a psychological construct that should be determinative for access planning, this construct is
hotly debated and very controversial; recent reviewers have suggested that it may be time to set it aside in favor of more
complex and multidimensional views of childrens attachment needs (Kelly & Lamb, 2000). As such, it has little professional
consensus as a determinative variable and its use as a centerpiece by a clinician for forming a rationale for a particular
custody plan should be viewed as highly suspect (see Level II comments above).

Other custody-specific constructs often used to argue for particular access plans include the parent of choice construct
presented by Bricklin (1984) and the parental custody index offered by Ackerman and Schoendorf (1992). Both systems
offer a welcome attempt to operationalize custody-relevant psychology variables and to standardize assessment approaches.
Unfortunately, as noted above, there are numerous substantive questions about the validity and reliability of these instruments
and, even more apropos to the current discussion, a lack of consensus among mental health professionals regarding how such
instruments should be validated (Brodzinsky, 1993). Until such a consensus is reached we believe that the use of such
constructs in forensic reports, especially at Levels III and IV, implies a level of scientific consensus and certainty that is
inappropriate. In addition, we believe that a more procedurally appropriate approach to developing custody-related
instruments is to focus research on better assessment techniques for Levels I and II, and that trying to measure such
socio-moral constructs as best interests or best custodian (the Level III constructs and the shoulds at Level IV) should
be deemphasized because they represent, in a sense, an attempt to measure the incalculable (partially *202 because such
value-laden constructs can reasonably be given different definitional contours by legal and mental health professionals who
have different preferences/values, yielding little consensual agreement about what is important to a child and to the definition
of the construct).

Another challenge with Level III inferences is the fact that the vast majority of psychological research that informs Level II
inferences was not designed specifically to inform those at Level III. For example, while the developmental literature might
allow a clinician to draw reasonable conclusions about attachment patterns in a particular family and about potential risks to
healthy attachment associated with different custody schedules, there is virtually no empirical literature that allows the
clinician to then, on the basis of this literature, reliably suggest specific access plans. This is because we do not know how the
effect of such patterns changes in the context of the enormously complex emotional life of the separating family. Long-held
truisms about attachment may morph or become invalidated under particular schedules, with particular children and parents
who display a unique array of psychological attributes.

It is our opinion that the most egregious manner in which Level III custody inferences are used in forensic reports involves
the lack of open admission to the court of the degree to which such inferences rest on substantial uncertainty and on the
subjective values of the individual clinician or the mental health profession at large. Informed clinicians can take reasoned
positions against a wide variety of custody-relevant assumptions (e.g., ODonohue & Bradley, 1999; Melton et al., 1997)
including: (1) continuity in social relationships should be paramount (What about the capacity of a child to be strengthened
and enriched by change?); (2) continuity in the relationship with the primary psychological parent, often assumed to be the
mother in this culture, should be paramount (because there is a rich literature regarding the benefits of father involvement that
may argue against this position); (3) children should never be placed with a parent who has engaged in domestic violence
(but occasionally domestic violence is circumscribed to separation-related events and the nonviolent parent may have other
meaningful forms of dysfunction such as alcoholism that also place children at risk); and (4) siblings should be kept together
(yet occasionally children have very distinct and contradictory personal needs that may suggest the value of dividing the
siblings). These represent only a small sampling of the highly subjective value choices that often lay behind the positions
clinicians take at Level III and that are often opined about in forensic reports without adequate disclosure of professional

13 cv 1944 CAB BLM

uncertainty.

The problem here is not that clinicians have nothing to say about these matters (we know about the effects of domestic
violence, alcoholism, etc. on children). However, we have virtually no empirical literature to guide us in the selection of
access plans among competing, complex clinical realities (i.e., in the weighing of different detriments and advantages).
Despite this state of affairs, it is at Level III that clinicians often present their logic and personal values under the guise of
behavioral-scientific truth without disclosing that we have no reliable clinical method to do this weighing. Simple logic and
subjective values do not represent specialized knowledge.

Finally, it is not uncommon for clinicians to use a childs current symptomatic response to a family plan as one argument for
an alternate access plan. However, while it is true that substantial data is accumulating on the normative effects of divorce on
children (e.g., Hetherington & Kelly, 2002), we have no empirically proven method for teasing out whether a childs
symptoms are best viewed as a time-limited response to a court dispute (that will subside postdisposition regardless of the
access plan) or a direct response to a specific custodial schedule.

*203 Level IV
As noted above, it is the specific custody recommendation that is most valued by the consumers of forensic evaluations (Bow
& Quinnell, 2004). At Level IV, the clinician begins to communicate a substantial narrowing of the preferred custodial plans
for a particular child and, quite often, specifically chooses a single plan as the most beneficial or least detrimental (right down
to the hours of the day or night a child should be with each parent and whether or not there should be sole or joint legal
custody). The professional posture of the forensic clinician who chooses to communicate at this level shifts dramatically from
having described for the court the whats of a family (Levels I, II, and III) to a specific, prescriptive recommendation
regarding what a court should do.

We are in agreement with forensic scholars who conclude that clinicians should routinely avoid addressing the ultimate issue
before the court in the form of a specific recommended access plan (e.g., Karras & Berry, 1985; Melton et al., 1997).
Drawing conclusions about the ultimate issue before the finder of fact blurs critically important boundaries between the
person invested with the power to make socio-moral and social-control decisions (the judge) and the expert witness who is
hired to assist the court (e.g., the psychologist). This argument holds that, even if one were to develop a family blood test that
could perfectly predict the custody plan that would lead to the best psychological functioning on the part of a child, clinicians
should still refrain from making that specific recommendation (although they might be able to say many things at Levels I, II,
and III to tip their hand). A clinician might be able to say that the family custody blood test predicts with 99% certainty
that primary custody with mother would lead to less behavioral disturbance on the part of the girls in this family. However,
specifically saying therefore, the court should place the children with their mother would still arguably be inappropriate.
We are not prepared to assert, and it is beyond the purview of this article to consider, that the policy judgment of Federal
Rule of Evidence 704 and its common law parallels generally permitting ultimate issue testimony should be repudiated.
However, it is imperative that mental health witnesses carefully circumscribe their testimony so as not to blur the critical
distinction between the function of an expert witness to expound from the specialized knowledge base (i.e., the whats) and
that of the judge to make the ultimate legal-socio-moral determination (i.e., the shoulds).

However, even if one asserts that experts should be able to address the ultimate issue as a general proposition, our review of
the literature suggests that we are quite far from the development of a family custody blood test and that the likelihood of
developing predictive power that would warrant specific custody recommendations is extremely remote. Yet an extremely
common part of forensic assessments, and often the part of the report read first by lawyers and judges, is the section in which
clinicians dare to tell the court what days and times the children should be with what parent.

The magnetic forces of trying to make private practices survive and of having legal advocates and judges tug at clinicians for
very specific direction make the regular choice to provide specific custody recommendations understandable, yet no less
egregious. Any clinician who recalls training in multivariate statistics and prediction will, if honest, attest to the
astronomically complex cognitive task of making relatively circumscribed behavioral predictions, let alone combining and
weighing the many inferences drawn at Level II and III into a specifically recommended access plan. However, there is not
one piece of research that we know of that supports the notion that clinicians can reliably and validly engage in this
multidimensional inferential process in a way that yields recommendations that are *204 correlated with positive child

13 cv 1944 CAB BLM

adjustment. There is no psychological technology existing for weighing and combining the complex factors assessed at
Levels I, II, and III into access plan recommendations that are empirically established to yield better child adaptation than
alternative plans. Theoretically, at Level IV, the clinician has somehow combined and weighed intra-psychic, behavioral,
dyadic, family-level, extended family-level, and cultural issues into a direction for the court that is presented as representative
of behavioral science, and this occurs despite the complete lack of evidence that we are able to do so reliably and validly.

There is a growing literature on the correlations between various custody plans (sole vs. joint) and positive/negative
adjustment in children (e.g., Bauserman, 2002). Assuming that this literature grows and becomes more robust, it may serve as
a useful source of knowledge for judges and policy makers. However, given the scores of variables that are idiosyncratic to
specific families, it may well be impossible to generate empirical prediction data that could be used to back up specific
custody recommendations for specific families (Melton et al., 1997). We will paraphrase and extend the phrase offered by
ODonohue and Bradley (1999) in the following fashion: when clinicians make a custody recommendation they are
essentially portraying themselves as knowing the best plan for a child after answering the question What initial state of
mother, father, children, extended family, and other important relationships, and what contextual variables, when considered
in the context of the scores of potential access plans for a given family, will result in what later negative or positive state in
the children? To say that the mental health profession is inadequately prepared to make such multivariate decisions is an
understatement, yet day in and day out forensic reports imply that we are quite able to do so.

Another way of stating the dilemma at this level is the following: there is no evidence in the empirical literature that current
interview protocols, traditional psychological tests, or custody-specific tests are in any way able to reliably predict child
adjustment to different access plans, yet 94% of evaluating psychologists still make such recommendations (Bow & Quinnell,
2004).

At this level, as in Level II and III, there is often a long list of behind-the-scenes values on the part of the clinician that is
guiding their choice of data, how they interpret it, and what they recommend. Unfortunately, many clinicians make specific
custody recommendations without explicating their often subjective beliefs about what is good and bad for children. It
should be emphasized that the purpose of behavioral science is to empirically guard against the excessive encroachment of
subjective values, or at least to guard against unwarranted assumptions. Consequently, clinicians arrive at court cloaked in the
presumed objectivity of behavioral science when many subjective values are directing their conclusions (e.g., boys do best
with their fathers, children do best with their primary attachment figure, joint plans are bad for young children, etc.). Few
clinicians are up front with the legal consumers of their reports about the degree to which the assumptive underpinnings of
their recommendations are highly subjective and controversial. In addition, the often emotionally infused desire to advocate
for a particular childs interests (rather than for the data) can itself introduce profound and distorting bias into the clinical
process (Wittmann, 1985).

Given the virtual nonexistence of empirical data supporting the predictive validity of psychological interviews and testing
with regard to specific access plans, ethical issues become quite problematic. The ethical principals of psychologists and
Code of Conduct (APA, 2003) state the following in Ethical Standard 2.04:
*205 Psychologists work is based upon established scientific and professional knowledge of the
discipline.

However, we have no empirical knowledge/support for our ability to make specific custodial predictions for specific families.
Ethical Standard 9.02(a) states the following:
Psychologists administer, adapt, score, interpret, or use assessment techniques, interviews, tests, or
instruments in a manner and for purposes that are appropriate in light of the research on or evidence of
the usefulness of the proper application of the techniques.

Unfortunately, because there are no clinical techniques currently available which have been shown to reliably predict positive
or negative child adaptation under different access plans, it would seem that any specific custody recommendations based on
current instruments and techniques would be ethically inappropriate.

Ethical Standard 9.02(b) states the following:

13 cv 1944 CAB BLM

Psychologists use assessment instruments whose validity and reliability have been established for use
with members of the population tested. When such validity or reliability has not been established
psychologists describe the strengths and limitations of test results and interpretation.

While there are many assessment techniques with established reliability and validity for assessing Level I and Level II
variables, the lack of similarly valid and reliable instruments for Levels III and IV (i.e., for predicting adjustment under
various plans) is an ethical indictment of any specific custody recommendation.

Ethical Standard 9.06 states the following:
When interpreting assessment results ... psychologists take into account the purpose of the assessment ...
They indicate any significant limitations of their interpretations.

Again, given the paucity of research noted for Level IV, psychologists are clearly required to inform custody litigants of the
lack of validity and reliability of the interview and test techniques they use if they intend to provide specific
recommendations.

Finally, Principal A states the following:
Psychologists strive to benefit those with whom they work and take care to do no harm.

We assert that specific custody recommendations, given their lack of adequate empirical foundation, can hold significant
potential for harming a child because of the dramatic effect that such recommendations can have on the individual, familial,
academic, and social trajectory of that childs life (given the absence of a knowledge base to reassure the finder of fact that
such a trajectory has been shown to lead to reliable, positive child outcomes).

Ethical considerations evolve as a discipline becomes more self-aware. To date, most clinicians have offered specific
recommendations in a well-intentioned and sincere effort to provide guidance that will help children vulnerable to the
stressors of family reorganization. In addition, they have been taught by the legal profession that their work product is highly
valued, reinforcing this professional behavior. However, given the various ethical dictates noted above, and our more evolved
self-understanding as a discipline regarding the limits *206 of our knowledge, it is our opinion that the time has come for
specific custody recommendations to be viewed as unethical for at least the foreseeable future.

CUSTODY RECOMMENDATIONS: AN EVIDENTIARY ANALYSIS
Given the enormous impact that custody evaluations can have on the lives of litigants and their children, and given the
remarkable limitations on the underlying professional knowledge base and lack of consensus as to uniform methodology, it is
important to examine these opinions within the framework of the law of evidence that governs admissibility. Notwithstanding
the increased attention to the gatekeeper responsibility of the trial court sparked by the U.S. Supreme Court decision in
Daubert v. Merrell Dow Pharmaceuticals (1993), custody opinions by mental health witnesses have largely escaped the
serious judicial scrutiny of reliability and relevance that is so central to the fundamental question of admissibility. As one
scholar has trenchantly noted, It is striking that one of the most important categories of cases, the future of our children, has
ignored the call for trial judges to address threshold scrutiny of the reliability and relevance of expert testimony (Shuman,
2002, p. 139).

Opinion evidence, whether from lay witnesses or experts, has long been a vexing and multi-dimensional issue. Writing of
opinion evidence generally, McCormick, at 12, states: In the outer circle of collateral fact near the rim of relevancy,
evidence will be received with relative freedom; but as we come closer to the hub of the issue, the courts are more careful to
call for details instead of inferences.

J uxtaposition of McCormicks statement with the four-level analysis presented above is graphically depicted in Figure 1.


13 cv 1944 CAB BLM

TABULAR OR GRAPHIC MATERIAL SET FORTH AT THIS POINT IS NOT DISPLAYABLE
Only Levels II, III, and IV are levels at which inferences are being drawn. At Level I, the clinician is reporting factual
observations made during the clinical interview. While his or her investigative expertise and interviewing skill may well have
uncovered the observed fact and the observation may strike right at the core of the best interest issue, his or her testimonial
status in reporting the observation is nonetheless that of a fact witness. For example, if during the interview, the father
became angry and struck the child physically, his conduct would impact upon the best interest issue quite directly. Yet, from
an evidentiary perspective, the clinicians secretary would be quite as competent as the evaluator to testify to the occurrence,
assuming he or she observed it, because its relevance does not depend upon any inference being drawn by the witness.

Once we move to the inferential levels, however, one can readily observe that the closer to the core issue, i.e., the centermost
circle, the opinion strikes, the greater its potential impact on the outcome of the case. Accordingly, under McCormicks
formulation the greater the judicial vigilance and evidentiary rigor must be. Similarly, it may be said that the more
questionable the reliability and validity of the premises underlying the opinion, that is, the principles, methodology, and
techniques of the profession whose expertise is being proffered, the more muscular the court must be in exerting its
prerogatives as evidentiary gatekeeper.

In terms of the multilayered paradigm posited above, as the witness proceeds from Level I to Level IV, from concrete factual
observations to the higher levels of abstraction, the opinion strikes closer and closer to the heart of the case and at the same
time becomes weaker and weaker in terms of its scientific validity and evidentiary reliability.

In this regard, McCormick makes an interesting statement:
Undoubtedly there is a kind of statement by the witness which amounts to no more than an expression of
his general belief as to how the case should be decided or as to the amount of *207 damages which would
be just. All courts exclude such extreme, conclusory expressions. There is no necessity for this kind of
evidence; to receive it tends to suggest that the judge and jury may shift responsibility for the decision to
the witnesses. In any event, it is worthless to the trier of fact [italics added] (McCormick, 12).

Interestingly, the specific custody recommendations under discussion fall precisely into this category of extreme, conclusory
expressions of opinion as to how the case should be decided. It is particularly troublesome because, as noted above, such
recommendations are not anchored to sufficient empirical research. Exacerbating the issue further, the very overreliance upon
the opinion of the witness, noted by McCormick and cautioned against even in some high judicial quarters,
1
is exactly what
often comes to pass with respect to custody evaluations. As one commentator observes,
The use of court-appointed experts whose opinions determine the outcome of many, if not most, contested custody cases
effectively delegates judicial power without formal legislative approval (Shuman, 2002, p. 161).

When that happens, the court may be basing its decision on personal value judgments of witnesses who happen to have
professional credentials. However, those credentials do not entitle the witnesss personal, as opposed to professional or
scientific, judgments to be admitted, let alone carry weight. When an expert witness does not testify on the basis of the
collective view of his scientific discipline and where no understandable scientific basis is stated, personal opinion, not
science, is testifying .... (Daubert, 1995; Turpin, 1992). Cogently stated, something doesnt become scientific knowledge
just because its uttered by a scientist .... (Daubert, 1995).

*208 To the extent such personal opinions are admitted under the guise of science and are allowed to impact the outcome of
the case, the process is tainted and judicial power is usurped. Therefore, when confronting the proffer of such opinions, the
court, in its role as gatekeeper, has every reason to be Draconian in its application of the rules of evidence and traditional
common law safeguards.

Evidence doctrine is well suited to the task of safeguarding the judicial process from pseudoscientific opinions. Indeed, the
raison-dtre of its many rules is the exclusion of unworthy proof. If the law of evidence were concerned only with allowing
evidence in, it could be stated in a single sentence: Evidence is admissible when it is relevant
2
(People v. Scarola, 1988). If
that were the only thing to be said on the subject, such legal scholars as Wigmore, McCormick, and others could have saved
immeasurable hours and intellectual energy, not to mention trees. Significantly, the vast body of evidence law is aimed at

13 cv 1944 CAB BLM

keeping unworthy evidence from tainting the mind of the fact finder.
3

ASSESSING RELIABILITY: THE DAUBERT STANDARD
In 1993, the U.S. Supreme Court, in Daubert, established a new standard for assessing the admissibility of expert scientific
testimony. The Court subsequently held that the new standard is to be applied to the assessment of all expert testimony,
whether grounded in science or some other species of specialized knowledge (Kumho, 1999).

In establishing a new standard, the Court explicitly rejected the so-called Frye rule, under which, the trial court, instead of
directly determining the validity of the science underlying the experts opinion, looked to whether the principle or method
had gained general acceptance within the relevant scientific community.
J ust when a scientific principle or discovery crosses the line between the experimental and demonstrable stages is difficult to
define. Somewhere in this twilight zone the evidential force of the principle must be recognized, and while courts will go a
long way in admitting expert testimony deduced from a well-recognized scientific principle or discovery, the thing from
which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in
which it belongs (Frye, 1923, p. 47).

Under Frye, the court engaged in counting scientists votes, rather than on verifying the soundness of a scientific
conclusion (People v. Wesley, 1994; Jones v. United States, 1988; State v. Montalbo, 1992).

It has been noted that Daubert could not have come at a better time because of the proliferation of harsh criticism of expert
testimony that is predicated upon junk science, with some of the harshest criticism being directed toward the testimony of
mental health witnesses (Lyon & Koehler, 1996; Foster & Huber, 1997).

Daubert is often cited as authority for the declaration that the trial judge is the gatekeeper, who is charged with meaningful
scrutiny of proffered scientific evidence (Reagan, 1999). In fact, judges have always served in that role. Not only with respect
to expert testimony, but, indeed, with respect to all evidence, the common law trial judge sits as referee, calling balls and
strikes in accordance with the law of evidence, deciding what proffered evidence is admissible and what is not. The distinct
contribution made by Daubert was that it reclaimed this vital role that the Frye test essentially delegated to extra-judicial
venues.

Yet, the judicial mission under either Frye or Daubert is the same. It is to ensure that only opinions grounded in valid
scientific principle and methods make it across the *209 drawbridge, and that those that only appear to be science, that is,
pseudoscientific assertions, do not (People v. Wesley, 1994). The difference is that under Frye the court effectively delegates
this determination to the relevant professional community, accepting as a given that the science is good science where it
has gained general acceptance within that community. Under Daubert, the court directly assesses whether proffered
testimony is so grounded, upon consideration of scientifically relevant criteria, such as whether the theory or technique can
be or has been tested, whether it has been subjected to the rigors of peer review and publication, how precise it is as measured
by the known error rate, as well as whether it is generally accepted in the relevant scientific community (Daubert, 1993).

Although, under Daubert, the applicable criteria of reliability may vary from issue to issue depending upon what subject
matter is at hand (Kumho, 1999), the standard it enunciated is applicable to all expert testimony and was sufficiently clear
that it was later incorporated into an amended Rule 702 of the Federal Rules of Evidence:
If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the
evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience,
training, or education, may testify thereto in the form of an opinion or otherwise, if (1) the testimony is
based upon sufficient facts or data, (2) the testimony is the product of reliable principles and methods,
and (3) the witness has applied the principles and methods reliably to the facts of the case.

It is because specific custody recommendations are not based on sufficient empirical data and therefore, cannot currently be
the product of reliable principles and methods, that they ought not to be admissible.

13 cv 1944 CAB BLM

GENERAL ACCEPTANCE--DEFINING THE RELEVANT SCIENTIFIC COMMUNITY
Frye and Daubert are both federal court decisions. While some states adopted the Daubert standard, others continue to
adhere to the Frye test of general acceptance (People v. Wesley, 1994). Even under Daubert, however, one of the criteria for
determining reliability is to look at whether the principle or method in question has gained general acceptance in the relevant
scientific community.

Both Frye and Daubert establish a test of admissibility, the purpose of which is to ensure that the evidence presented is
reliable (People v. Wernick, 1996). The distinction between the two is simply that Daubert undertakes to do so directly.
However, when determining admissibility under Frye, the court does not directly assess the scientific validity, that is, the
evidentiary reliability of the proffered testimony. Rather, as noted earlier, it looks to whether the scientific principle or
procedure at hand has gained general acceptance in the particular field in which it belongs (Frye, 1923, p. 47).

Under Frye, the court exclusively looks to general acceptance of the procedures and methodology as reliable within the
scientific community, whereas under Daubert general acceptance is simply one of the factors considered (People v.
Wernick, 1996). Under either standard, proper definition of the relevant scientific community is important. For a Frye court,
because general acceptance is the only factor considered, proper definition is nothing less than critical.

Defining the relevant scientific community with respect to the reliability of specific custody recommendations is not as
simple a proposition as it may first seem. Because custody evaluations usually lie within the realm of psychology, certainly
the evidence of reliability *210 ought to come from its domain-- but from which quarter? Here, one must take note of the
schism that exists within the field between those who anchor their conclusions upon empiricism and those who proceed by
intuition.
Even more deep-seated than the division between scientists and practitioners in clinical psychology has been the split
between empiricists and romantics (a terminology suggested by psychiatrist Paul McHugh of J ohns Hopkins University).
According to scholars of intellectual history, two broad trends can be discerned in European and American thought over the
past few centuries. One trend (which McHugh calls empiricist but historians might call enlightenment) holds that reason,
objectivity, and empirically verified evidence provide the surest road to knowledge. The other trend (which both McHugh
and the historians call romantic) teaches that intuition, empathy, and subjective insights can provide deeper and more
authentic understanding than can mere reason
4
(Wood, Nezworski, Lilienfeld, & Garb, 2003, p. 93).

Notwithstanding the fact that mental health witnesses sometimes attempt to escape the clutches of cross-examination by
claiming psychology is not a science, but an art, or a hybrid of art and science (Gould, 1998, p. 53), a claim that carries many
other evidentiary implications, none of which operate in favor of admissibility, the mainstream view or consensus within the
profession is to the contrary. As one historian of the mental health profession states:
As a result, empiricism has become the dominant perspective of contemporary psychology, gaining
almost universal acceptance. There seems to be widespread agreement that scientific advances are
optimally produced and conveyed under the procedures of empirical verification; other forms of inquiry
do not appear to offer the compelling attraction of empiricism (Brennan, 2003, p. 332).

Accordingly, it is both appropriate and necessary to hold the mental health witness accountable for the application of
empirically supportable principles and methods. The courts have both the right and the obligation to insist that the experts
whose opinions can change lives in dramatic fashion support each and every one of their inferences with specific empirical
evidence. In other words, the legal system must demand that the premises and reasoning of the expert be scientifically valid.

The deep schism between the scientists of the profession and the romantics ought to be of critical concern to the legal
profession when assessing admissibility. When a court assesses the general acceptance issue, under either Frye or Daubert, it
must look to those within the profession who adhere to the scientific method and who insist upon empirical verification of
their premises and conclusions.


13 cv 1944 CAB BLM

Unfortunately, scientifically minded clinicians are becoming a rapidly dwindling minority within their profession
(Lilienfeld, Lynn, & Lohr, 2003, p. xi). Indeed, the term schism may be too mild to describe the phenomenon. Today,
however, calling it a gap is like saying there is an Israeli-Arab gap in the Middle East. It is a war, involving deeply held
beliefs, political passions, views of human nature and the nature of knowledge, and--as all wars ultimately involve--money,
territory, and livelihoods (Lilienfeld et al., 2003, p. xiv).

The implications of this dwindling attachment to scientific method with respect to therapeutic settings aside, in the forensic
context such unanchored opinions can determine whether or how often a specific child gets to see a particular parent. Here,
with substantial individual rights hanging in the balance, it clearly becomes the responsibility of the courts *211 to keep
unscientific opinions regarding access plans out of the process entirely. Whatever epistemological merit there might be to the
romantic position, the courts, in assessing evidentiary reliability, must look to the empiricists for their assessment of the
existing research data that supports or disconfirms tendered theories.

It is therefore critical that the courts define the relevant community with care because that definition can determine the
outcome of the general acceptance issue. If, for example, the Frye court had defined the relevant community as polygraphists,
rather than looking to physiological and psychological authorities (Frye, 1923, p. 54), the polygraph likely would have
been found admissible. Likewise, opinions based on astrology would be admissible were the relevant community defined as
astrologers and inferences drawn from the Ouija board would come in if we look only to those who believe it has predictive
validity. The Supreme Court acknowledged as much in Kumho, finding the general acceptance factor of little assistance
where the discipline itself lacks reliability, as, for example, do theories grounded in any so-called generally accepted
principles of astrology or necromancy (Kumho, 1999, p. 1175).

Likewise, the reliability of custody evaluations and specific recommendations cannot be accurately assessed by reliance upon
those in the clinical community who have largely departed the realm of science. The court, for an accurate assessment of the
state of specialized knowledge, must turn to psychologists who emphasize science and research, and empiricists, who
recognize that intuition and subjective insight are inappropriate bases for forensic conclusions and who bind themselves to
the scientific method. To do otherwise is akin to looking to astrologers or necromancers to assess the scientific reliability and
validity of star-based predictions and sances.

A court looking to general acceptance either as a standard or as one of several criteria of reliability is tracking for the
presence or absence of controversy, a point made by Chief J udge Kaye in her very instructive concurrence in People v.
Wesley (1994):
The point of noting controversy about the reliability of the forensic technique is not for our Court to determine whether the
method was or was not reliable in 1988, but whether there was consensus in the scientific [italics added] community as to its
reliability. The Frye test emphasizes counting scientists votes, rather than on verifying the soundness of a scientific
conclusion. (citations omitted) Where controversy rages, a court may conclude that no consensus has been reached (p. 439).

Two points are worth making in this regard. First, as the writings of the several scholars and commentators referenced herein,
as well as others not cited, make clear, controversy rages with respect to the validity of specific custody recommendations.
Somewhat curiously, it is a rather one-sided controversy in that much has been published in the peer-reviewed literature
delineating the absence of empirical support for specific recommendations, while scholarly argument supporting the
empirical foundations for such recommendations is scant to nonexistent even though the practice of rendering
recommendations has become commonplace.

Second, in determining the admissibility of such recommendations, it is clearly the opinion of the scientific community as to
general acceptance that is required. That is the essence of the Frye test. It is also why the opinions of nonscientific
practitioners will not be helpful and why the courts should turn only to carefully chosen scientist-practitioners and, more
often, to research psychologists. It is this latter group of professionals who have been trained from the start to understand that
replicated behavioral research in conjunction with cautious hypothesis testing is the only way to guard against subjective
bias, idiosyncratic *212 and unverified theory, and the excessive encroachment of personal values into custody-related
inferences that can change the trajectory of a childs life (Levels III and IV).

Looking to scientist-practitioners and researchers, rather than those who rely on intuition and theory that is without a solid

13 cv 1944 CAB BLM

research base, serves another purpose as well. New Yorks highest court has noted that courts ought to be cognizant of the
potential economic interest of those to whom it looks for evidence of general acceptance: A Frye court should be
particularly cautious when--as here--the supporting research is conducted by someone with a professional or commercial
interest in the technique (citation omitted) (People v. Wesley, 1994, p. 464) (Kaye, C. J ., concurring). Thus, the question of
reliability ought not to be submitted to the custody evaluators who have an economic interest in the admissibility of their
opinions.

BURDEN OF PROOF
Finally, when it comes to establishing that there is a valid scientific basis for the opinion of the expert, the burden of proof
rightly falls upon the proponent of the evidence. A basic tenet of science is that the burden of proof always falls squarely on
the claimant, not the critic. Consequently, it is up to the proponents of these techniques to demonstrate that they work, not up
to the critics of these techniques to demonstrate the converse (Lilienfeld et al., 2003, p. 3; Tyson v. Keane, 1998; People v.
Wesley, 1994).

QUALIFICATION: A QUESTION-BY-QUESTION ISSUE
The first step in taking the direct testimony of an expert is to have the witness accepted by the court as qualified as an
expert. All too often, once the witness has been so accepted, what ensues is an intellectual imbroglio, with the witness
being allowed to opine on matters that do not fall within the range of his or her expertise.

When looking to the issue of witness qualification, a distinction must be made between the question of whether the witness is
generally qualified to speak for his or her profession and the discrete, but often overlapping question of whether that
profession has anything relevant to say about the issue before the court that is grounded on scientifically valid principles as
established by empirical research. As one commentator has noted:
By virtue of their qualifications alone, experts do not provide any assurance that their opinions rest on
reliable methods and procedures. Instead, relying on experts without testing the reliability of their
methods and procedures cloaks experts value judgments under the veil of science and risks that their
personal and professional characteristics bias the evaluation and the importance of information learned
(Shuman, 2002, p. 160).

Significantly, under Rule 702, the witness must be qualified as having scientific, technical, or specialized knowledge with
respect to those issues that the court determined will be helpful to the trier of fact. Once so qualified, the witness is allowed to
testify in the form of opinion or otherwise, but only to the extent that (1) the testimony is based upon sufficient facts or data,
(2) the testimony is the product of reliable principles and methods, [italics added] and (3) the witness has applied the
principles and methods reliably to the facts of the case.

This implicates Wigmores delineation that the expert witnesss testimonial capacity in every case is a relative one, i.e.,
relative to the topic about which the person is asked to *213 make his statement. The object is to be sure that the question to
the witness will be answered by a person who is fitted to answer it (Wigmore, 1979, 555, p. 750).

Further, Wigmore states that: Since experiential capacity is always relative to the matter at hand, the witness may, from
question to question, [italics added] enter or leave the class of persons fitted to answer, and the distinction depends on the
kind of subject primarily, not on the kind of person (Wigmore, 1979, 555, p. 750).

Assuming, therefore, that the witness is generally qualified as an expert in the particular scientific discipline, the related
question is whether that discipline has something to say in response to each particular question posed, a something that can
be proven to be scientifically valid by the empirical research that constitutes the specialized knowledge base of that
discipline. It was precisely the lack of such valid underlying methodology that was the subject of comment by the Supreme
Court in Kumho where the Court sustained the lower courts rejection of expert testimony. The District Court had found the
witness qualified as a mechanical engineer, yet rejected the proffered testimony because despite those qualifications ... the
methodology employed by the expert in analyzing the data ... and the scientific basis, if any, for such an analysis was not

13 cv 1944 CAB BLM

shown to be reliable (Kumho, 1999, pp. 1176-1177; Shuman, 2002, p. 139).

In the absence of empirical research that establishes that there are reliable principles and methods at the disposal of the
experts profession, then the testimony falls outside the domain of professional opinion and into the realm of personal value
judgment and theoretical assumptions. This distinction between expert-based conclusions and personal opinion is implicit in
the requirement that the experts opinion be stated in terms of reasonable professional certainty and is of critical
importance in assessing the admissibility of the opinions offered in custody evaluations:
There is an important difference between an expert opinion and a personal opinion. When an expert has
formulated an opinion, it is reasonably presumed that the expert has drawn upon information
accumulated and published over the years. The defining attributes of an expert opinion relate not to the
credentials held by the individual whose fingers type the words or from whose mouth the words flow;
rather, the requisite characteristics relate to the procedures that were employed in formulating the
opinion and the body of knowledge that forms the foundation upon which those procedures were
developed [italics added]. If the accumulated knowledge of the experts field was not utilized, the opinion
expressed is not an expert opinion. It is a personal opinion, albeit one being expressed by an expert
(Martindale, 2001).

Thus, a witness might well be qualified quite eminently as a clinical or forensic psychologist yet the opinion nonetheless be
inadmissible because it rests in some part on a principle or technique that is not scientifically reliable and valid, such as
concluding that a child has been sexually abused on the basis of play with anatomically detailed dolls (Lilienfeld et al., 2003,
pp. 57-61), or upon unproven theoretical assumptions or personal values (e.g., preschool children need a primary home,
siblings should not be separated, etc.). The question is not whether this person is qualified as an expert because of his or her
credentials as a psychologist or psychiatrist but, rather, whether any mental health witness, however well credentialed, is
competent to answer the particular question, for example, what arrangement is in the best interest of the subject child?

Otherwise stated, does this witness have specialized knowledge that allows each required inference to be drawn so that it
may be stated to a reasonable degree of professional certainty? That question, in due course, turns on the underlying
question as to whether the witnesss profession has produced a reliable and valid knowledge base upon *214 which to
predicate each inference. This issue, ultimately, turns on whether the empirical research supports scientifically reliable and
valid constructs to support the inferences. The process of witness qualification establishes only that this particular witness is
qualified to speak for his or her profession, to state in a courtroom what is known empirically to that profession. What is
known to the mental health profession is that which is established empirically, not what this particular witness may have
idiosyncratically intuited or concluded on the basis of personal value judgments.

OPINION ON ULTIMATE ISSUES
Even though evidence doctrine no longer precludes opinions on ultimate issues as a general matter, the courts ought not to
permit them in custody cases as a matter of evidentiary law and the mental health profession should not provide them as a
matter of professional ethics.

There are two related reasons why such opinions ought not to be permitted:
They exceed the boundaries of the empirical knowledge base of the mental health profession.

They implicitly misrepresent the limits of that knowledge base.

As one noted scholar expresses the problem:
Frequently, evidence that lacks reasonable scientific backing is presented to the courts as though it had
scientific merit. Although corrupt individuals occasionally may act as hired guns, testifying in favor of
those who retain them for a price, a much more common and pernicious problem is the uncorrupt but
inaccurate expert who presents pet theories and personal prejudices as scientific fact (Baerger,

13 cv 1944 CAB BLM

Galatzer-Levy, Gould, & Nye, 2002).

The Melton treatise also makes the point quite cogently:
Despite the fact that such opinions are commonly sought and, unfortunately, are commonly given, mental health
professionals ordinarily should refrain from giving opinions as to ultimate legal issues .... When experts give such opinions,
they usurp the role of the fact finder by suggesting that the opinions are based on specialized knowledge specific to the
profession.

Note in this regard that although Rule 704(a) allows experts to give opinions on ultimate issues, Rule 702 prohibits admission
of any opinion not based on specialized knowledge, a prohibition which presumably can include ultimate issue opinions.
Indeed, Rule 704(b) (an amendment to the original Rule 704 which was inspired by J ohn Hinkleys acquittal on insanity
grounds) makes this point concretely with respect to mental state testimony in criminal cases. The position we take is that the
same evidentiary prohibition should apply to all types of cases. In any event, even if a court permits such an opinion to be
admitted as a matter of law, it should not be offered as a matter of professional ethics because of the implicit
misrepresentation of the limits of expertise involved when a clinician acting as an expert witness gives a legal opinion in the
guise of mental health knowledge (Melton et al., 1997, p. 17).

Lest there be any doubt that Melton et al. include mental health testimony in connection with custody opinions, they state
quite compellingly:
Indeed, there is probably no forensic question on which overreaching by mental health professionals has
been so common and so egregious. Besides lacking scientific validity, such opinions have often been
based on clinical data that are, on their face, irrelevant to the legal questions in dispute (p. 484).

*215 Indeed, whatever position one might take on the ultimate issue rule with respect to other species of expert testimony,
such opinions by mental health witnesses on the ultimate question of a childs best interest ought not to be allowed. It is less
an ultimate issue matter than it is a question of basis.

The best interests standard is a legal and socio-moral construct, not a psychological construct. There is no empirically
supportable method or principle by which an evaluator can come to a conclusion with respect to best interests entirely by
resort to the knowledge base of the mental health profession. In this regard, Melton et al. (1997) have noted that there has
been remarkably little research meeting minimal standards of methodological rigor about the effects of various custody
arrangements on children and families of different characteristics (p. 484). Krauss and Sales (1999) observed that ... few
studies have focused on the directly relevant legal questions or used the methodological rigor that is necessary for a
psychologist to offer exclusively science-based opinions on the best interest of a particular child (p. 89).

ODonohue and Bradley (1999) called for a moratorium on all participation by mental health professionals in the child
custody arena because ... the mental health professional currently cannot ethically conduct these evaluations. In so doing,
they refer to the lack of empirical research that clearly identifies relevant variables involved in child custody and their
possible causal or correlative relationships to key outcomes.

The custody court is confronted with the challenge of determining with which parent the childs interest will best be served.
Sometimes this involves selection of the better parent. Too often, the court must choose the one who is less bad. There is
no empirically verified psychological construct of good parent or bad parent, let alone a construct for the comparative
analysis that the court is called upon to perform. While we have established certain correlations between parenting style and
child outcome, good and bad are socio-moral constructs that, by definition, are not within the purview of psychology as a
behavioral science. There is also a noted lack of scientific methodology, empirical grounding, and psycho-legal relevance
in the custody evaluation field (Baerger et al., 2002, p. 36). Yet, these opinions are highly influential:
The recommendations contained in child custody evaluations (CCEs) exert considerable influence on
the course of ongoing custody litigation. Many courts accord significant weight to the opinions of child
custody evaluators, often accepting the evaluators recommendations without challenge. An evaluators
recommendations can also precipitate case settlement or material concessions once both parties become

13 cv 1944 CAB BLM

aware of the evaluators findings (Baerger et al., 2002; Horvath, Logan, & Walker, 2002).

In effect, the legal, socio-moral construct of best interest has no specific, operational definition in terms of psychological
assessment and is so ambiguous and value laden that reaching a consensus that would permit accurate measurement of this
characteristic in a forensic assessment is not possible (Shuman, 2002, p. 144).

As one prominent writer in the field states:
We often use scientifically crafted data-gathering tools, such as standardized tests. We organize our data and form inferences
based on theoretically developed models of human behavior that guide our conclusions. We are, in most aspects of our work,
scientists struggling to fit our observations into models of behavior that can inform the court about what is in the childs best
psychological interests.

We should not delude ourselves, however. Our techniques and models are only in their infancy. Our tests are often not
validated on forensic populations, let alone the specific *216 populations of those undergoing custody evaluations. Our
interview protocols often are homegrown, with little, if any, reliability data available concerning test-retest or internal
consistency. Our recommendations are based not on empirically driven conclusions but on best-guess strategies.

For example, no empirical work has been done in which a matched set of children with similar test and interview data is
placed randomly into different custodial arrangements to examine overall effectiveness of one placement over another. Such
an experiment would be blatantly unethical and inhumane. We cannot manipulate families to serve our science. Yet we often
provide testimony to the court about custodial arrangements with the arrogance of true science implied. We are but
sophisticated guides for the trier of fact through a confusing array of psychological technology, a technology never intended
for use in custodial assessment. Through our learned and judicious use of psychological theories, methods, and data
gathering, we determine our best guesses possible. Our tools are often not valid for custodial assessment. Our models are
often rationally, not empirically, derived. And our opinions are more educated guesses than truth. We need to be careful in
how we present our data and opinions to the court so as not to mislead (Gould, 1998).

Psychology has a long way to go before it has empirically tested competing theories about the long-term adaptation of
children under various access plans and, as noted above, the multidimensional complexity of such a task, along with our
inability to randomly assign children to different custody conditions, make the likelihood of an adequate test extremely
remote. Until custody-specific theory and replicated empirical research on this topic is a fact, rather than a hope, the custody
evaluator has little empirically based guidance about what relevant variables to analyze, let alone their priority or relative
weight. A uniform methodology for conducting custody evaluations likewise must await the empirical research that indicates
what the relevant variables are that must be studied and development of a uniform method of assessing them. Absent that
research, the evaluator cannot possibly draw the necessary inferences without seeking guidance from clinical experience,
which is often nothing more than accumulated bias, personal moral/value judgments, or unverified theoretical assumptions
(Wittmann, 1985).

The issues presented here go far beyond questions of best practice because one cannot begin to consider practice standards
in the absence of an empirical knowledge base that informs the discussion. They are not problems that can be subsumed
within the rubric of it goes to weight either, or the time-honored though doctrinally unsupportable, Ill take it for what its
worth. Rather, these issues go to the heart of the law of evidence, that is, relevance and reliability, and thus present a
question not of weight, but of admissibility.

DISCUSSION AND CONCLUSIONS
Notwithstanding the strong case that exists that it is unethical for mental health witnesses to tender specific custody
recommendations and that their admission at trial violates fundamental evidentiary doctrine, it would be a mistake to blame
the current situation solely on the mental health profession. That profession is simply responding to the demand of the legal
consumers, the lawyers and judges who actively seek such invalid opinions. Privately, many evaluators will acknowledge
that they deliver these recommendations with great internal reluctance, but yield to consumer pressure for fear of being out of
business if they refuse.

13 cv 1944 CAB BLM

Psychologists frequently assume roles in which theyre expected to offer opinions, even though uncertainty is enormous and
the scientific knowledge base is extremely limited. For example, *217 in custody evaluations that involve reasonably normal
individuals, its often impossible to know which parent will provide a child with a better environment. Research in this area is
limited, and there are few or no well-validated procedures for arriving at recommendations.

... Nevertheless, the lawyers and judges in such cases often press psychologists for a firm opinion.

Under such demands, psychologists sometimes resort to unvalidated procedures. The reasoning seems to be that any test,
even one of dubious validity, will perform better than merely flipping a coin (Wood et al., 2003, p. 297).

The legal system has both the right and the duty to exclude opinions that are not supported by good science. It is
understandable that lawyers and judges actively seek recommendations that are not so supported. This reflects both the
natural human desire for easy answers to difficult problems and the legal professions lack of understanding of scientific
method in general and, more specifically, its proper application in the behavior science milieu.

Lawyers and judges perceive a value to specific recommendations in large part because they often facilitate settlement
without trial. It is important to understand, however, that the leverage the recommendation exerts on litigants has little, if
anything, to do with its underlying validity. As a general proposition, litigants settle cases when they perceive that they likely
will do worse if they go to trial. Thus, the typical scenario in custody litigation is that the evaluators report comes in with a
specific recommendation. The attorney for the disfavored party tells the client the judge is highly likely to go with the
recommendation. Loosely translated, you are going to lose, so why not settle now and save yourself the risk and cost of
trial. It can fairly be said that the client does not hear the recommendation and experience an epiphany, suddenly realizing
that the other parent is indeed the better choice. But for the perception that the judge will credit the recommendation and act
upon it, the report would exert little leverage on the settlement process. Hypothetically, the same leverage could be obtained
if everyone believed that the judge had tossed a coin and the result would be determinative or highly persuasive. The
distinction is that no one would believe the coin toss is a relevant predictor of parenting capability but many within the legal
system, being uninformed of behavioral science precepts, do believe that the mental health witnesss recommendation is.

Yet, specific recommendations can also have unintended effects. When a judge receives a specific custody recommendation
from an evaluator and then decides that it ought not be followed, the recommendation may well pose an obstacle that the
judge needs to surmount in writing the decision. Unfortunately, some appellate courts have given quasi-presumptive status to
the opinions of court-appointed evaluators, mistakenly confusing court-appointed status with freedom from bias and other
assessment errors. (Rentschler v. Rentschler, 1995; In re Custody of Rebecca B., 1994, but see, Chait v. Chait, 1995; Edgerly
v. Moore, 1996). Such decisions leave cautious trial judges feeling the burden of justifying any departure from the mental
health opinion by writing their way out of it.

Given the pervasiveness and potentially enormous impact of behavioral science testimony, it is imperative that lawyers and
judges become more educated and informed consumers of mental health opinions. Advanced education in scientific method
and empirical behavioral science research will position lawyers and judges to bring more qualitative analysis to bear upon
forensic testimony and produce an essential understanding that the legal system ought not make demands upon behavioral
science that it is presently unequipped to meet nor admit recommendations that are not predicated upon an empirically
established specialized knowledge base.

*218 RECOMMENDATIONS
As implied above, it is our position that clinicians with adequate forensic training can provide matrimonial and family court
judges with very useful and helpful information at Levels I and II. For example, admissions by litigants regarding disputed
questions before the court, compilations of descriptions regarding parenting behavior over time, and data based inferences
regarding parenting skills/deficits. In addition, the articulation of a childs expressed concerns, emotional pain,
developmental needs, and preferences all represent ethical contributions to the fact finding process as long as each conclusion
that is drawn is based in a cautious hypothesis-testing model, convergent data, and the specialized knowledge of the
profession. The simple listing for the court of psychological risk factors associated with various access plans (Level III) can
also be done in an ethical manner as long as it is carefully grounded in case specific data and the specialized knowledge base

13 cv 1944 CAB BLM

of the mental health profession. In these areas, we believe that forensic reports can have very important value to the
fact-finding effort and that they help to humanize an often excessively polarized and distorting adversarial process. However,
our literature review suggests the following changes in professional practice:
(1) Given our conclusion that psychologists can only offer ethical and helpful information at Levels I and II, and a small bit
of useful information at Level III, the courts should continue to engage qualified and circumspect forensic evaluators to assist
the fact-finder in bringing forth custody-relevant information. However, we do suggest a moratorium on the practice of
psychologists providing recommendations regarding specific custody arrangements and schedules.

(2) Psychologists should routinely include warnings in forensic reports about the very significant limitations in our capacity
to predict the best or worst custody plans for a particular family as a step toward educating the judiciary (particularly if they
choose to continue the practice of making specific custodial recommendations).

(3) The American Psychological Association should work toward articulating a more clear position against specific custody
recommendations (given that it is almost impossible to argue that adequate data exists to make such prescriptions).

(4) J udges should begin to help the psychology discipline rein itself in by including in court orders a specific statement
precluding the expert witness from addressing the ultimate custody-plan issue.

(5) Clinicians should avoid addressing custody-related issues that are largely non-psychological (economic benefits of
different child placements, qualitative differences between school districts, etc.).

(6) J udges and lawyers should routinely be provided with more comprehensive training regarding seminal psychological and
developmental issues relevant to access planning. More broadly, law school curricula ought to include basic and advanced
courses to educate the legal profession with respect to the many areas where law and the behavioral sciences intersect.

(7) Clinicians should be allowed to function as expert evaluators in custody matters only if they have had adequate forensic
training (to help ensure respect on the part of the clinician for the critical importance of an empirical grounding for inferences
when such inferences can result in the limitations on human liberties).

(8) Future research should focus on developing valid and reliable instruments for measuring Level I and Level II variables. In
addition, attempts should be made to study the multivariate combinations of individual, family, and access plan variables that
tend to be correlated with negative and positive child outcomes (admittedly, a daunting task).

*219 REFERENCES
Ackerman, M., & Schoendorf, K. (1992). ASPECT: Ackerman-Schoendorf Scale for Parent Evaluation of Custody--Manual.
Los Angeles: Western Psychological Services.

American Psychological Association. (2003). Ethical principles of psychologists and code of conduct. American
Psychologist, 57(12), 1060-1073.

American Psychological Association. (1994). Guidelines for child custody evaluations for in divorce proceedings. American
Psychologist, 49, 677-682.

Baerger, D. R., Galatzer-Levy, R., Gould, J . W., & Nye, S. G. (2002). A methodology for reviewing the reliability and
relevance of child. Journal of the American Academy of Matrimonial Lawyers, 18, 35-74.

Bagby, M. R., Nicholson, R. A., Buis, T., Radovanovic, H., & Fidler, B. (1999). Custody evaluations detecting malingered
and defensive responding on the MMPI-2 in family custody and access evaluations. Journal of Personality Assessment,
62(2), 191-203.

Bauserman, R. (2002). Child adjustment in joint-custody vs. sole custody arrangements: A meta-analytic review. Journal of

13 cv 1944 CAB BLM

Family Psychology, 16(1), 91-102.

Bow, J . N., & Quinnell, F. A. (2001). Psychologists current practices and procedures in child custody evaluations five years
after American Psychological Association guidelines. Professional Psychology: Research and Practice, 32(3), 261-268.

Bow, J . N., & Quinnell, F. A. (2004). Critique of child custody evaluations by the legal profession. Family Court Review, 42,
115-126.

*220 Brennan, J . F. (2003). History and systems of psychology (6th ed.). Upper Saddle River, NJ . Prentice-Hall.

Bricklin, B. (1989). Perception of Relationships Test manuel. Furlong, PA: Village Publishing.

Bricklin, B. (1984, 1990a). Bricklin Perceptual Scales manual. Furlong, PA: Village Publishing.

Bricklin, B. (1990b). Parent Awareness Skills Survey manual. Furlong, PA: Village Publishing.

Bricklin, B. (1992). Data-based tests in child custody evaluations. American Journal of Family Therapy, 20, 254-265.

Brodzinsky, D. M. (1993). On the use and misuse of psychological testing in child custody evaluations. Professional
Psychology: Research and Practice, 24(2), 213-219.

Butcher, J . N. (1997). Frequency of MMPI-2 scores in forensic evaluations. MMPI-2 News and Profiles, 8(1), 2-4.

Cummings, M. E., & Davies, P. (1994). Children and marital conflict: The impact of family dispute and resolution. New
York: Guilford Press.

Felner, R. D., Rowlison, R. T., Farber, S. S., Primavera, J ., & Bishop, T. A. (1987). Child custody resolution: A study of
social science involvement and impact. Professional Psychology: Research and Practice, 18, 468-474.

Foster, K. R., & Huber, P. W. (1997). Judging science: Scientific knowledge and the federal courts. Cambridge, MA: MIT
Press.

Galatzer-Levy, R. M., & Kraus, L. (1999). The scientific basis of child custody decisions. New York: Wiley.

Garb, H. N. (1998). Studying the clinician. Washington, DC: American Psychological Association.

Goldstein, J ., Freud, A., & Solnit, A. (1973). Beyond the best interests of the child. New York: Free Press.

Gould, J onathan W. (1998). Conducting scientifically crafted child custody evaluations. Thousand Oaks, CA: Sage.

Heilbrun, K. (2001). Forensic mental health assessment. New York: Kluwer Academic.

Hetherington, M., & Kelly, J . (2002). For better or for worse: Divorce reconsidered. New York: W.W. Norton & Company.

Hetherington, M., & Stanley-Hagen, M. (2002). Parenting in divorced and remarried families. In M. H. Borstein (Ed.),
Handbook of parenting (Vol. 3). NJ : Erlbaum.

Horvath, L. S., Logan, T. K., & Walker, R. (2002). Child custody cases: A content analysis of evaluations in practice.
Professional Psychology: Research and Practice, 33, 557-562.

Hynan, D. L. (2003). Parent-child observations in custody evaluations. Family Court Review, 41(2), 214-223.

Karra, D., & Berry, K. (1985). Custody evaluations: A critical review. Professional Psychology: Research and Practice,
16(1), 76-85.

13 cv 1944 CAB BLM

Karras, D., & Berry, K. K. (1985). Custody evaluations: A critical analysis. Professional Psychology: Research and Practice,
16(1), 76-85.

Kelly, J ., & Lamb, M. (2000). Using child development research to make appropriate custody and access decisions for young
children. Family & Conciliation Courts Review, 38(3), 297-311.

Krauss, D. A., & Sales, B. D. (1999). The problem of helpfulness in applying Daubert to expert testimony: Child custody
determinations in family law as an exemplar. Psychology of Public Policy and Law, 5, 78-99.

Lamb, M. E., Orbach, Y., Sternberg, K. J ., Hershkowitz, I. & Horowitz, D. (2000). Accuracy of investigators verbatim notes
of their forensic interviews with alleged child abuse victims. Law and Human Behavior, 24, 699-708.

Lilienfeld, S. O., Lynn, S. J ., & Lohr, J . M. (2003). Science and pseudoscience in clinical psychology. New York: Guilford
Press.

Lyon, T. D., & Koehler, J . J . (1996). The relevance ratio: Evaluating the probative value of expert testimony in child sexual
abuse cases. Cornell Law Review, 82, 43-78.

Martindale, D. A. (2001). Cross-examining mental health experts in child custody litigation. The Journal of Psychiatry &
Law, 29, 483-511.

Mayes, L. C., & Truman, S. D. (2002). Substance abuse and parenting. In M. H. Borstein (Ed.), Handbook of parenting (Vol.
4) (pp. 329-360). Mahwah, NJ : Erlbaum.

McCann, J . T., Flens, J . R., Campagna, V., Collman, P., Lazzaro, T., & Connor, E. (2001). The MCMI-III in child custody
evaluations: A normative study. Journal of Forensic Psychology Practice, 1, 27-44.

Melton, G. B., Weithorn, L. A., & Slobogin, C. (1985). Community mental health centers and the courts: An evaluation of
community-based forensic services. Lincoln, NE: University of Nebraska Press.

Melton, G. B., Petrila, J ., Poythress, N. G., & Slobogin, C. (1997). Psychological evaluations for the courts: A handbook for
mental health professionals and lawyers (2nd ed.). New York: Guilford Press.

Meyer, G. J ., Finn, S. E., Eyde, L. D., Kay, G. G., Moreland, K. L., Dies, R. R., et al. (2001). Psychological testing and
psychological assessment: A review of evidence and issues. American Psychologist, 56(2), 128-165.

ODonohue, W., & Bradley, A. R. (1999). Conceptual and empirical issues in child custody evaluations. Clinical
Psychology: Science and Practice, 6, 310-322.

*221 Otto, R. K., Edens, J . F., & Barcus, E. H. (2000). The use of psychological testing in child custody evaluations. Family
& Conciliation Courts Review, 38(3), 312-340.

Patterson, G. R., & Fisher, P. A. (2002). Recent developments in our understanding of parenting: Bidirectional effects, causal
models, and the search for parsimony. In M. H. Borstein (Ed.), Handbook of parenting (Vol. 5; pp. 55-88). Mahwah, NJ :
Erlbaum.

Pope, K. S., Butcher, J . N., & Seelen, J . (2000). MMPI, MMPI-2, & MMPI-A in court: A practical guide for expert witnesses
and attorneys (2nd ed.). Washington, DC: American Psychological Association.

Reagan, R. T. (1999). Book review--judging science: Scientific knowledge and the federal courts, by Kenneth R. Foster &
Peter W. Huber. Oklahoma Law Review, 52, 291-301.

Rothbaum, F., Weisz, J ., Pott, M., Miyake, K., & Morelli, G. (2000). Attachment and culture: Security in the United States

13 cv 1944 CAB BLM

and J apan. American Psychologist, 55(10), 1093-1104.

Rubin, K. H., & Burgess, K. B. (2002). Parents of aggressive and withdrawn children. In M. H. Borstein (Ed.), Handbook of
parenting (Vol. 1). Mahwah, NJ : Erlbaum.

Shuman, D. W. (2002). The role of mental health experts in custody decisions: Science, psychological tests, and clinical
judgment. Family Law Quarterly, 36, 135.

Strong, J . (Ed.). (1999). McCormick on evidence (5th ed.) (Vols. 1-2). St. Paul, MN: West Group.

Teti, D. M., & Candelaria, M. A. (2002). Parenting competence. In M. H. Borstein (Ed.), Handbook of parenting (Vol. 4; pp.
149-180). Mahwah, NJ : Erlbaum.

Tippins, T. M. (2003). Custody evaluations--Expertise by default? New York Law Journal, 230, 3.

Turk, D. C., & Salovey, P. (Eds.). (1988). Reasoning, inference, & judgment in clinical psychology. New York: Free Press.

Tutin, J . (1993). The persistence of initial beliefs in clinical judgment. Journal of Social and Clinical Psychology, 12(3),
319-335.

Waxler, Z. W., Duggal, S., & Gruber, R. (2002). Parenting and psychopathology. In M. H. Borstein (Ed.), Handbook of
parenting (Vol. 4; pp. 295-328). Mahwah, NJ : Erlbaum.

Wigmore, J . (1979). Wigmore on evidence. Boston: Little Brown.

Wittmann, J . J . P. (1985). Child advocacy and the scientific model in family court: A theory for pre-trial self-assessment.
The Journal of Psychiatry & Law, 13, 61-82.

Wittmann, J . J . P. (2003). Should forensic psychologists make custody recommendations? New York Family Law Monthly,
4(11), 1-3.

Wood, J . M., Nezworski, M. T., Lilienfeld, S. O., & Garb, H. N. (2003). Whats wrong with the rorschach? Science confronts
the controversial inkblot test. San Francisco: J ossy-Bass.

Table of Cases
Bennett v. J effreys, 356 N.E.2d 277 (N.Y. 1976).

Chait v. Chait, 638 N.Y.S.2d 426 (N.Y. App. Div. 1995).

Edgerly v. Moore, 647 N.Y.S.2d 773 (N.Y. App. Div. 1996).

Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579 (1993).

Daubert v. Merrell Dow Pharm., Inc., 43 F.3d 1311 (9th Cir. 1995).

Frye v. United States, 293 F. 1013 (D.C. Cir. 1923).

In re Custody of Rebecca B., 611 N.Y.S.2d 831 (N.Y. App. Div. 1994).

In re Sayeh R., 693 N.E.2d 724 (N.Y. 1997).

Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137 (1999).

13 cv 1944 CAB BLM

J ones v. United States, 548 A.2d 35 (D.C. 1988).

People v. Scarola, 525 N.E.2d 728 (N.Y. 1988).

People v. Wernick, 674 N.E.2d 322 (N.Y. 1996).

People v. Wesley, 633 N.E.2d 451 (N.Y. 1994).

Rentschler v. Rentschler, 611 N.Y.S.2d 523 (N.Y. App. Div. 1995).

State v. Montalbo, 828 P.2d 1274 (Haw. 1992).

Turpin v. Merrell Dow Pharm., Inc., 959 F.2d 1349 (6th Cir. 1992).

Tyson v. Keane, 991 F.Supp. 314 (S.D.N.Y. 1998).

Footnotes

a1

Timothy M. Tippins, Esq. is an Adjunct Professor of Law at Albany Law School and a private practitioner who has engaged in
matrimonial and family law practice for over thirty years. He now devotes his practice time to serving as special counsel and
consultant to other family law practitioners on a nationwide basis, serving as trial and appellate counsel, with special emphasis on
the presentation and cross-examination of expert mental health testimony. Tippins has served in all major professional leadership
positions in the New York family law community, including President of the American Academy of Matrimonial Lawyers--New
York Chapter, Chair of the New York State Bar Association Family Law Section, and Chair of the New York State Bar Association
Task Force on Family Law. Tippins is a regular feature columnist for the New York Law Journal, where his numerous articles
regarding custody evaluations have been published. He has lectured before many major professional organizations, such as the
Association of Family and Conciliation Courts, American Academy of Matrimonial Lawyers, ABA, NYSBA, and the New York
State Psychological Association. Together with Dr. Wittmann, he has taught the Tippins-Wittmann Empirical Forensic Model at
the NYS Judicial Institute and at other forums around the country.

a2

Jeffrey P. Wittmann is a licensed psychologist, trial consultant, and divorce mediator whose practice concentrates on consultation
in contested custody and access matters. He serves as a trial consultant for major law firms in the Manhattan area and elsewhere.
Dr. Wittmann is a nationally recognized expert on issues related to the intersection of law and psychology and on professional
practices in the child custody area, is regularly on the faculty for training seminars offered to attorneys and mental health
professionals, and has been a frequently utilized scholar-trainer for judges in Family, Supreme, and Appellate courts. He lectures
on forensic and divorce psychology at both the undergraduate and graduate level. Dr. Wittmann has provided invited testimony to
the New York State Senate regarding matrimonial statutes; provides opinions on divorce-related matters to print, television, and
radio media; and has served as a program consultant to ABC News. He is on the editorial board of the New York Family Law
Monthly and is the author of Custody Chaos, Personal Peace (Perigee, 2001) and of numerous scholarly articles regarding
forensic psychology.

1

See, for example, Mtr. of Bennett v. Jeffreys, 40 NY2d 543 (1976): In custody matters parties and courts may be very dependent
on the auxiliary services of psychiatrists, psychologists, and trained social workers. This is good. But it may be an evil when the
dependence is too obsequious or routine or the experts too casual. Particularly important is this caution where one or both parties
may not have the means to retain their own experts and where publicly compensated experts or experts compensated by only one
side have uncurbed leave to express opinions which may be subjective or are not narrowly controlled by the underlying facts.; and
Mtr. of Sayeh R. v. Monroe County DSS, 91 N.Y.2d 306, 693 N.E.2d 724: It is necessary also to make some reference to the role
of the psychological material presented in the instant case. Experts, who predict future consequences based on their professional
theories and examinations of subject children, should not be elevated to the singular importance of, in effect, overriding the array
of pertinently balanced jurisdictional protections afforded to decrees affecting one of societys most sacrosanct
relationships--parent and child. Courts must beware lest the unique juridical authority to decide these cases be sacrificed to the
sheer crosswinds of paid or even so-called independent experts.


13 cv 1944 CAB BLM

2

Even in the basic statement, the concept of relevance ought to keep out most of what passes in the trade today because proffered
testimony cannot be relevant if it is not reliable. The two concepts are inextricably intertwined.

3

The issue of keeping unreliable evidence fromprejudicing the factfinder is sometimes assumed to be less significant in custody
cases because they are usually tried to a judge, rather than a jury. This notion is not well-founded:
Because judges, not jurors, are responsible for the final decision in these cases, the judges may believe that they will not be
influenced by expert testimony that might fail the Daubert admissibility standard and, therefore, do not feel it is necessary to apply
the Daubert standard to expert testimony prior to its admission into evidence in this context. Obviously, assumptions that judges
can mentally differentiate good scientific expert testimony from bad scientific expert testimony, and differentially evaluate
scientific and clinical testimony, are unanswered empirical questions. Likewise, the assumption that judges are less influenced by
bad scientific testimony than jurors is also an unanswered empirical question. Yet, similar to the prior assumptions, it appears from
the existing psychological literature on decision making that these assumptions are questionable at best (Krauss & Sales, 1999, p.
84).

4

The same text also notes that tensions between professional psychology and research psychology reached crisis levels by the late
1980s when practitioners took control of the APA, prompting research psychologists to break away in 1988 and formthe American
Psychological Society whose defined mission is to promote, protect, and advance the interests of scientifically oriented
psychology in research, application, teaching, and the improvement of human welfare.

43 FAMCR 193
End of Document

2014 Thomson Reuters. No claimto original U.S. Government Works.


13 cv 1944 CAB BLM
A THIRD CALL: RESTORING THE NOBLE EMPIRICAL..., 43 Fam. Ct. Rev. 270

Family Court Review
April, 2005
Tippins and Wittmanns Rejoinder
A THIRD CALL: RESTORING THE NOBLE EMPIRICAL PRINCIPLES OF TWO PROFESSIONS
Timothy M. Tippins, J effrey P. Wittmann
a1

Copyright 2005 by Association of Family and Conciliation Courts; Timothy M. Tippins, J effrey P. Wittmann
In writing the lead article, we hoped to produce a deeper interdisciplinary discourse that will lead to a higher level of service
for the families and children we serve. The thoughtful responses that follow the lead article encourage us to believe that such
discourse and the consequential elevation of practice standards is, indeed, a realistic hope.

Kelly and J ohnston (this issue) state that they are in essential agreement with our analysis, but express concern that we may
have understated the breadth and depth of the empirical research base in the area of divorce psychology. They further express
the concern that our call for a moratorium on recommendations may create even greater risk for families.

With respect to their first point, we thank them for augmenting what we knew was only a partial review of the available
empirical research base. Our shorter description of the literature is intended to be illustrative. We repeatedly observe in the
lead article that the empirical literature available to custody evaluators, though insufficient as a foundation for ultimate
expressions of best interest, is quite substantial and can provide a firm anchor for many less abstract, yet still helpful
family-related conclusions. Kelly and J ohnston have performed an important service by delineating the scope of this
empirical database. This delineation provides additional support for our point that, even at Level II, the mental health
discipline is able to provide a great deal of objective and verified information that can assist courts in making decisions
relative to the needs of children.

Their position is that our call for a moratorium on recommendations may create even more risks for divorcing families. This
seems to rest upon an implicit assumption that the current state of affairs, wherein evaluators choose specific custody plans in
the absence of evidence that they can reliably do so, thus altering childrens life courses and truncating fundamental parenting
rights, poses little or no risk to those families. That assumption is not necessarily valid. In fact, whether children and families
are better off with or without explicit recommendations is an empirically unanswered question and, therefore, we would posit
that such an assumption is at least questionable.

For example, for decades the mental health profession used the simple metric of determining the primary psychological
parent according to the Goldstein, Freud, and Solnit model (1979); yet we essentially do not know, on average, how those
children fared in response to recommendations that placed them in largely sole-parent custodial arrangements. Best interest
recommendations were widely admitted during those decades on theoretical premises that many scholars now would view as
questionable. It is entirely possible, because these child custody decisions were based on overly simplistic notions of
attachment psychology, that children would have been at a lesser risk had such testimony *271 been excluded. Certainly, its
exclusion might have increased the possibility that alternative placements that could have better served the childrens futures
would have been explored and tried with greater frequency. In short, in the absence of longitudinal tracking and feedback
with respect to the developmental outcomes of children whose custody has been determined on the basis of forensic
recommendations, we do not know whether those recommendations have worked or not worked, or how much unwitting
harm they may have inflicted.

We heartily agree with Kelly and J ohnstons assertion that there has been a drift in the custody evaluation field toward a
willingness to split hairs with regard to the psychological advantages and disadvantages of mothers and fathers for their
children. This thereby raises the bar for what is considered adequate parenting in custody arrangements well beyond the bar

13 cv 1944 CAB BLM

that is established for matters in dependency courts. This unfortunate dynamic has made mental health professionals
inappropriate arbiters of social, moral, and cultural values, denoting a marked departure from the noble empirical roots of
their discipline.

Kelly and J ohnston suggest that we may have been excessively conservative in the limitations we suggested on what forensic
evaluators can do with Level III information. Most of the research examples that they reference for sound, empirically based
Level III statements would, according to our model, fall in Level II, for example, the associations between certain parental
disorders and child dysfunction, adaptive parenting behavior that is predictive of positive child outcome, attachment-related
research findings, and so forth. Level III statements are limited to those custody-specific comments by the forensic evaluator
that are more conclusory and directive in tone, such as the reporting of parenting quotients, the preferred parent, summary
statements regarding the risks associated with a particular parent, and so on. It is at this point that the evaluator begins to take
the Level II inferences about psychological disorders, parenting capacity, developmental needs, and so forth and summarize
them in a manner that can be used as a basis for choosing one parenting arrangement over another. It is here too that the risk
of inappropriate infringement on human liberties soars. Although it may be a fine-line distinction, what we are asserting is
that at Level II, forensic evaluators can reach a wide variety of basic conclusions about a child and his or her parents.
However, once these are arranged in a Level III format to present conclusions about the risks and benefits of different
custody arrangements, evaluators should seriously constrict their utterances to those conclusions that are firmly anchored in
the empirical knowledge base of the psychology discipline.

Finally, we entirely agree with Kelly and J ohnstons assertion that the presentation of a series of alternative hypotheses and
risk/benefit analyses with respect to future child functioning under different arrangements, when firmly supported by
empirical research findings, can be an ethical and valuable service to provide for the court. Our model simply suggests that
this would be defined as a Level III work product, not Level IV, as Kelly and J ohnston suggest, because it does not say what
the judge should specifically do regarding the custody plan to be chosen, and, therefore, would be entirely appropriate.

Gould and Martindale (this issue) respond to our position with an articulate call for forensic evaluators to become more
grounded in the empirical literature as they form their work products for the court. We agree with them that harm can come to
families when untrustworthy data are not identified as such and that opinions about specific predictive outcomes that are
expressed with a high level of confidence are not grounded in science. We also agree with their call for higher practice
standards in this area, an assertion that was also supported by Kelly and J ohnston in their critique. In addition, we found
Gould and Martindales position helpful and congruent with our own, that the problem we describe is consumer-driven and
that not much will change until the courts begin to do what they *272 should have been doing all along-- carefully
scrutinizing the basis of the opinions that are offered by forensic experts to ensure that a strict standard of evidentiary
reliability, designed to ensure due process and to protect human liberties, is observed.

When Gould and Martindale suggest that it is deceptive to offer custody opinions based on professional experience, because
it implies that the forensic field is able to accurately make such a prediction, we could not agree more and, in fact, we view
this as the heart of the matter we have attempted to delineate in our piece. Do educated guesses have a place in a court of law
where a child and parent can lose access to one another? Should courts make the custody decision based, even in part, upon a
witnesss intuitively inspired subjective belief, even when it is paraded with a fancy moniker like clinical judgment? We
think not. Indeed, evidentiary doctrine is organized around the principle that exactly such quasi-scientific and nonscientific
opinions ought not enter the realm of the courtroom.

Gould and Martindale, while recognizing the dubious empirical base for ultimate best interest opinions, seem to suggest that
it nonetheless may be acceptable for the mental health professional to testify to ultimate factual issues, such as
psychological best interests. The problem with this is that there is no agreed-upon operational definition of psychological
best interests, let alone consensus as to how such an amorphous construct could or should be measured. Thus, the question
remains: How can such recommendations align with, what we view as Gould and Martindales welcomed and respectable
emphasis on, the critical need for sound scientific bases for the conclusions expressed in forensics reports and testimony?
While we agree that the constriction that they suggest, couching the opinion in qualified terms of psychological best
interests, marks progress and makes the empirical overreach less egregious than the more sweeping and generalized best
interest assertions presently typical in the field, an overreach does remain.

It bears repeating that there is no operational construct of psychological best interests, a void that can easily be illustrated

13 cv 1944 CAB BLM

by considering the commonly experienced scenario of two imperfect parents, each conveying to a child a combination of
dysfunctional traits and positive attributes. One parent may be an empathetic, compassionate, emotionally attuned being who
has never held a productive job, who endlessly watches TV reruns, who obsessively surfs the Internet, and who sets the
academic achievement bar quite low for the child. At the same time, the other parent may be a high-achieving, extremely
productive, high-income CEO who is preoccupied with work, emotionally distant, and who sets the academic performance
standard quite high for the same child. Even assuming, arguendo, that the psychology discipline could perfectly predict that
if custody is given to the first parent, the child will develop into an underachieving, sporadically employed, but happy and
stress-free adult; while if committed to the custody of the latter parent, the child will become very professionally successful,
high-achieving, materially prosperous, anxious, ulcer-medication-popping adult, the choice between those two outcomes
remains a socio-moral decision. It is not a question that can be answered by the scientific knowledge base of the mental
health profession. Thus, even constricting forensic utterances to psychological best interests, does not resolve the
personal/moral/cultural/social subjectivity issue always implicit in a custody determination.

While Gould and Martindale underscore the importance of disclosing the absence of research data on any particular issue that
is addressed in a forensic report, we would emphasize that there are ethical mandates beyond just disclosing limitations, for
example, the mandate to use instruments, such as tests and interview protocols, only for their empirically established
purposes. And, even assuming that disclosure satisfied the evaluators ethical strictures, once the witness concedes a lack of
empirical basis for his or her conclusions, under Daubert v. Merrel Dow Pharmaceuticals, Inc. (1995), the court should rule
the testimony inadmissible.

*273 We completely agree with Gould and Martindales assertion that it is distressing to find that all too many evaluators
continue to base forensic opinions on clinical judgment and hunches and adamantly insist that this is acceptable. Stubborn
adherence to this circle the wagons position, defending at all costs the way the mental health profession has been doing
evaluations, rather than critically scrutinizing that process in empirical terms consistent with the very roots of the psychology
discipline, could very well signal that future day, a day perhaps not that far off, when custody evaluators are entirely excluded
from the legal system on an evidentiary basis. It is not as though such calls for categorical exclusion have not been heard
before. Our lead article referenced ODonohue and Bradleys (1999) call for a moratorium on mental health participation in
the custody process until greater scientific coherence is achieved. More recently, New Yorks Matrimonial Commission,
which scrutinized the forensic role, reportedly heard testimony from respected quarters of the legal profession recommending
that custody evaluations not be utilized at all, unless there is a history of mental illness in the family in question (Wise, 2004).

Gould and Martindale also refer to the fact that if the mental health profession were to withdraw from custody evaluation
work, the important research that we suggest is necessary would not be undertaken. We would point out that nowhere in our
article do we suggest such a retreat and, in fact, we would view its occurrence as a profound disservice to divorcing families.
The lead article endeavors to draw distinctions between admissible and inadmissible assertions in the belief that strict
observance of the evidentiary standards will preserve due process and allow into evidence only that expert testimony that is
soundly based on scientifically valid data, principles, and methods. Such stringent safeguards, which we should have been
observing all along, ought to temper calls for outright rejection of all mental health input. These are calls that the courts
might answer if they perceive that their only choice is between a wholesale rejection or a continuation of the current chaos.

Grisso (this issue) expresses concern, also voiced by Gould and Martindale, that exclusion of ultimate best interest opinions
and specific custodial arrangement recommendations may impair the livelihood of many mental health professionals. We
would respond by noting first that the possibility of such impairment, though an understandable practical concern, is not an
intellectually valid argument against elevating the ethical and professional level of practice in any field of endeavor. It is even
less a sound reason for continuing a practice that is potentially harmful, as we believe is the proffer of empirically
unsupported custody recommendations laden with subjective personal judgments that are potentially infused with various
biases and ripe with judgment error. When, for example, the Food and Drug Administration announces a ban on products
long on the marketplace, such as recent proscriptions of ephedrine or andro supplements, companies invested in the
manufacture and sale of those products fold, and jobs are lost. Yet, such industry-wide economic dislocation does not--and
cannot--deter the removal of risky products from the market to enhance the safety and well-being of consumers.

Second, we submit that the reforms we propose, which are really more a matter of restoration of bedrock principles of both
the legal and the mental health professions, need not carry dire economic consequences. The most acute economic risk would
be borne by individual evaluators who courageously adopt our position and henceforth decline to specify ultimate best

13 cv 1944 CAB BLM

interest solutions, irrespective of what others may do. The risk--and it is a real one-- is that less ethical or less scientifically
faithful evaluators would continue to provide such specious conclusions and, because lawyers and judges may like specific
recommendations, more evaluations would be assigned to these practitioners at the expense of those individuals who adhere
to the higher standard we propose. This is why we agree with Gould *274 and Martindale that the legal system, which bears
substantial responsibility for the current chaos in demanding and allowing unsupportable recommendations, should indeed, as
they so nicely put it, go first.

This is also why we recommend in the lead article that greater education in the empirical content of the behavioral science
field, as well as the principles of scientific method, is essential. Such education is vital not only so that lawyers, judges, and
mental health professionals are brought to a deeper understanding of the importance of empirical support for expert opinions,
and why conclusions lacking such support have no place in the courtroom, but also to underscore that the mental health field,
and the psychology discipline in particular, has much empirically established information to impart that can inform the court
in discharging its obligation to determine the best interest question. In other words, such educational efforts will lay the
groundwork for a new and broader understanding of the appropriate role of forensic evaluators.

As we have taught our judicial education seminars, two qualities of thought emerged from our audiences. First, many judges
expressed their gut feeling that something was amiss, a sense that the best interest opinions they were hearing daily in their
courtrooms simply did not have legs. Second, a surprising number of judges found our more conservative empirical model,
restricted to an empirically supported recitation of risks and benefits associated with different custody plans, to be quite
attractive because it still provides them with valuable guidance without risking a usurpation of their role as the final arbiter of
the custodial decision.

Grissos comment that judges prefer to hear specific recommendations, which require Level IV inferences that are
impermissible under our empirical model, points out another important practical dilemma presented by the reforms we are
proposing. Yet here, again, this is not an intellectually sound argument against elevating practice standards and responsibly
exercising the judicial role of gatekeeper.

J udges may prefer to hear many kinds of testimony, but judges are not laws unto themselves. An individual judge may,
because of his or her personal metaphysical belief system, prefer to hear from astrologers or necromancers. Yet, as
gatekeepers, they are duty-bound to exclude any such proffer by applying the rules of evidence as they have been fashioned
either by common law or codified by legislatures or rule-making authorities. It is worth the reminder to observe that the rules
of evidence, as well as the precepts of common law trial procedure more generally, though they may at times seem arcane,
were designed to ensure due process and to protect the rights and liberties of those who come before the courts. Certainly,
such safeguards are no less important when the rights of parents and children are at stake. In fact, it is those very safeguards
that draw the distinction between empirically grounded, reliable testimony from a competent forensic evaluator and the
intuitive inspirations of the astrologer, allowing the former into evidence while simultaneously excluding the latter.

We agree with Grisso that the issues we have delineated are not limited to the realm of custody work or even to mental health
testimony in general. The basic issue applies to all species of expert testimony. Where the particular type of testimony flows
from the realm of science, as does that which comes from the psychology discipline, the empirical standard needs to be
brought to bear, even if that means that opinions that have long been commonly received, for example, expressions of
causation based on clinical medical judgment, are henceforth excluded (Faigman, Kaye, Saks, & Sanders, 2002). Once one
accepts the premise that the rules of evidence stand as the guardian of the liberties of our litigants, one must allow the
evidentiary chips to fall where they may.

*275 Grissos suggestion that the kinds of multivariate research we recommend may, in the end, prove to be impossible and
that the research may never produce the requisite empirical support for an ultimate best interest opinion may be exactly
correct. In a sense, Grissos assertion in this regard underscores the concerns we have expressed: Day in and day out forensic
evaluators are engaging in an extremely complex, multivariate assessment, and weighting/prioritizing task for which they
have virtually no research to guide them. As Kelly and J ohnston observe, there is a substantial volume of literature on various
individual, family-level, and divorce-related constructs. However, when it comes to the combination, weighting, and
prioritization of the multiple and multilevel variable constellations presented in most custody matters, there simply is no
research to suggest that such a complex, predictive undertaking can be achieved with scientific validity and evidentiary
reliability. Notably, while the nine responding articles in this issue do not agree with all of what we have postulated in the

13 cv 1944 CAB BLM

lead article, no one suggests that such empirical support for the validity and reliability of ultimate best interest opinions does,
in fact, exist.

We do believe that the discipline of psychology has already advanced well beyond where it was even thirty years ago with
regard to its research and statistical procedures. In addition, we do not want to give up hope that energetic academicians and
scientist-practitioners will seek and find funding to begin gathering helpful and predictive data. However, even if one accepts
the possibility, alluded to by Grisso, that achieving a multivariate research base is impossible, then so be it. We may never
reach the Holy Grail and common law judges will have to decide custody cases in the absence of Level IV input, just as they
did for years and years before mental health experts entered the field and just as fact-finders presently have to resolve
credibility issues without the aid of the polygraph. Despite decades of controversy, industry advocacy, and actual usage in the
investigatory realm, the polygraph has never made it to an acceptable level of evidentiary reliability to become admissible.
The parallels of the polygraph to our subject are striking. The polygraph may be good enough to use outside the courtroom,
but not sufficiently valid to be admitted into evidence, just as many clinical approaches may be helpful in treatment, but
remain far removed from the evidentiary reliability required by law to be heard in court. We do not lower the bar for the
polygraph because that is the best that we can produce and there is no reason to do so when the rights of families are on the
line. One simply does not abandon standards of evidentiary reliability, designed to safeguard liberties, to suit the limited state
of knowledge of a discipline. Rather, the burden rests on the discipline to show that its principles, methods, and resulting
conclusions have adequate evidentiary reliability to be worthy of courtroom expression.

Grissos point about theory as a foundational aspect of scientific progression and his suggestion that custody assessment
requires a sound, empirically validated theory to explain the logic on which ones opinion is based, is a helpful reminder
about the true nature of science. However, if we return to the example of the slothful/compassionate parent versus the
hyper-driven CEO parent, one can readily see that by a slight turn of the forensic prism with any particular family, an entirely
new theory about what is best for a child can be selected. This is due to the subjective value-laden, socio-moral sand that
often forms the foundation for what sound like highly predictive and confident custody recommendations. One would need
multiple theory givers for any particular case in order to simply cover the bases. We have suggested in our lectures that
doing so can actually be a service to the court, as long as the evaluator holds back on choosing a specific preferred plan. One
forensic may present a cogent, emotional-intelligence-based theory in support of the slothful parent, while another may be
just as cogent in proffering an achievement-based theory as *276 the basis for recommending the CEO parent. What basis in
the research literature would the forensic have for preferring one over the other? This is ultimately a simple values-based
preference about desirable outcomes about which reasonable evaluators could disagree.

Kisthardt and Fines (this issue) assert that it is possible that judges are in no better position to determine the best interest of a
child than are psychologists, and that their unease in doing so has likely led to an excessive reliance on psychological
assessments. It is also possible that mental health professionals are in no better position than the judge to make such a
value-laden judgment call. Neither assertion has been empirically proven.

The real problem is that the legal system all too often indulges the assumption, with no underlying evidence to support it, that
mental health professionals are in a superior position to provide best interest answers. Perhaps they are, but we do not know
that to be true. A vicious cycle ensues that only compounds the problem. Every time a mental health professional gives a
recommendation, he or she implicitly communicates to the court, We can do this. And every time a court allows the expert
to make an ultimate best interest statement, the court tacitly says back to the mental health profession, Yes, you can do this.
Once judges and lawyers, acting on behalf of parents and children, begin to ask the right question, to wit: Can you do this?
the honest answer will be No. Should the occasional rogue say, Yes, the immediate response must be, Prove it, upon
which silence should reign in the courtroom.

It may also be true, as Kisthardt and Fines assert, that if custody evaluators increasingly pull their punches by presenting
their conclusions more conservatively, the legal system may conclude that they have little value. Yes, some may indeed so
conclude, but as we took pains to observe in the lead article, and which Kelly and J ohnston so nicely underscore in
augmenting our recitation of the many issues with respect to which empirical research does exist, that conclusion would be
wrong.

This means that we need to be energetic in our attempts to educate lawyers and judges, as well as many mental health
professionals, to appreciate the valuable contribution that can be made by putting forth empirically based conclusions without

13 cv 1944 CAB BLM

overreaching. The mental health profession must not fall prey to the temptation to act as if it can do something accurately, to
wit, make specific custody-related predictions, when in fact there is no evidence it can do so. Kisthardt and Fines are
absolutely correct in identifying this informational disconnect between lawyers and psychologists about the empirical
nature of the discipline and the limits of the underlying science. Simply, the psychology discipline needs to educate the
public, including the legal profession, about what it really has to offer--and reeducate those evaluators within its own ranks
who increasingly have departed from the empirical foundation of the profession.

We want to thank J udge Rotman (this issue) for her commentary. We agree that the specificity offered in custody
recommendations is a clear invasion of the realm of judicial responsibility, but it is important to underscore the fact that
because evidentiary doctrine vests in the trial court the gatekeeper role, all that is necessary to preclude potential usurpation
of the judicial function is the informed and vigilant exercise of that power to exclude such evidence. By carefully tailoring the
terms of the appointment order and thoughtfully framing the precise psycho-legal issues the evaluator is to address, the trial
judge can prohibit the expression of such empirically unsupportable conclusions from the outset. Further, the trial court may
also direct in the appointment order that the evaluator cite in the report to the empirical research upon which he or she relies
when asserting psychological risks and benefits of different custody arrangements at Level III. This would not only result in a
more elucidating work-product but would also have the salutary effect of reigning in those evaluators who are inclined to
remain detached from the scientific foundation of the psychology discipline.

*277 While we agree with J udge Rotman that a fair resolution of a custody matter may be in a childs best interests, the
broader issue is the essentiality of a reasoned and predictable adjudicative process, that is, observance of procedural due
process. J udge Rotmans comments suggest she would not be in disagreement with this assertion.

J udge Rotmans reference to dedicated family courts points to an important innovation. However, unless judges receive
intense training in behavioral sciences, the scientific method, and the nuances attendant upon proper exercise of their
gatekeeping role, even specialized and intelligent judges will be relegated to on-the-job training, to learning these things on a
case-by-case basis, and then only if the lawyers are acting competently in their roles. It may, in fact, be the lack of
meaningful judicial education on divorce-related behavioral science that has fueled, in part, excessive reliance on the mental
health profession.

We also agree that attorneys all too frequently cross-examine mental health witnesses in a manner suggesting inadequate
preparation and lack of thoroughness, thereby failing to do what J udge Rotman suggests is an important part of the trial
process: educating the judge. Even skilled trial lawyers, who may do quite well at pointing out logical flaws or inferential
gaps, leave the potentially more fertile ground of empirical basis untilled simply because they lack the training to pursue this
issue.

Hunters response (this issue) focuses on the important area of the development of model standards. She notes that it would
be disconcerting if evaluators recommendations were routinely rubber-stamped into final judicial determinations. We can
only note that we regularly hear from attorneys that this is the case in an unfortunately high number of family courts
nationally. Because the recommendation often comes to court wrapped in impressive credentials and shrouded in
scientific-sounding jargon, an uninformed court may well give it a greater weight than it warrants.

We are pleased that AFCC is developing new model standards for custody evaluations. We would urge that the new
standards stress the importance of empirical research as the foundation for all forensic conclusions expressed in reports or
testimony and that they adopt our suggestion that the practice of making specific recommendations cease.

The problem in establishing standards is in carefully defining the goal. If the new standards are built upon the assumption
that the end result of the evaluation process should be the expression of an ultimate best interest opinion or a specific
recommendation of custodial arrangement, success is likely to remain elusive. Until such time as there is a scientific basis to
define the optimal child outcome--which is highly unlikely given the implicit nonpsychological, personal, moral, and social
value judgments implicit in such definition--and an empirical research base exists that facilitates valid prediction of outcomes
in the face of multitudinous parenting and other variables--efforts to define best practices aimed at such prediction will be
circular and futile. Such an effort is akin to attempting to map out a navigational route without knowing either the destination
or which roads are passable, or more to the point, attempting to map a route to a destination that does not exist. Only if the
drafters carefully define the goal of evaluation as the generation of empirically supported statements that can inform--without

13 cv 1944 CAB BLM

attempting to decide--the best interest issue, can meaningful progress be made.

Martins response (this issue) to our article references public sector evaluations and the use of qualitative approaches to
scientific inquiry. Her outline of the use of prolonged engagement, PRD briefing and support, triangulation, member
checking, and auditing resonate with one of the essential notions behind our theoretical argument: solo practitioners often
have little reality check and few self-corrective mechanisms to help them spot instances where they may be getting it
wrong. Martins description of her approach *278 suggests that the processes her staff uses may include at least some
self-corrective mechanisms, a procedure that is quite welcome. However, it is important to emphasize that a forensic opinion
can be checked and cross-checked, appear internally consistent, logical, and intuitively sensible, yet rest entirely on a lack of
adequate scientific basis, which is the critical underpinning of whether or not an opinion should be admissible under Daubert.
A discipline presenting itself as scientific should be able to show that what it does in its forensic procedures has been
scientifically determined to be doable to an acceptable level of validity and reliability.

We note that Martins qualitative process appears largely to involve the gathering of Level I information for the court and we
believe that it is at Level I that forensic evaluators often have the greatest amount to offer the fact-finding process. However,
we disagree with Martins assertion that higher-level abstractions that overreach the empirical research should be taken by the
court for what they are worth. Such postulation grossly underestimates the dramatic extent to which specific
recommendations can affect the trajectory of a childs life without adequate empirical foundation. The issues raised in the
lead article and further discussed here and by the responders cannot be answered with a facile Ill take it for what its worth
ruling. Such a response has no more validity here than it has when invoked to dispose of a cogent hearsay objection. On its
face and in fact, the for what its worth rubric reflects a measure of weight, for example, what weight will the trier of fact
give to the admissible evidence. It is not a standard of admissibility and, when invoked as such, negates the existence of any
threshold assessment of evidentiary reliability.

J udge Dessau (this issue) poses the question of whether a judge is better placed than a social scientist to decide the best
interest issue. In addition to our comments above in response to the related question posed by Kisthardt and Fines, we would
suggest that on a purely legal basis the judge is the only one who is placed to make the best interest decision.

This is a broad public policy issue of legislative, if not constitutional, dimension (Shuman, 2002). For better or worse, that
issue has been determined in the investiture of authority in a duly elected or appointed judge. Were the issue to be debated
within the context of either legislative or constitutional amendment, that is, be it resolved that henceforth the best interest
custodial decision will be made by social scientists/mental health professionals, the question would be, what evidence is
there that they are better placed than a judge to decide the custodial issue? For that answer, we would need to look to
empirical research. As already noted above, to date, there is no evidence that the social scientist is any better placed to
accurately and validly answer that question than is a judge or, for that matter, than is a mail carrier.

J udge Dessau reminds us that the traditional evidentiary bar on ultimate issue testimony has been widely abandoned. This
is true. However, as we point out in the lead article, the ultimate issue evidentiary rule does not frame the relevant question.
At common law, even a well grounded, empirically based opinion would not be allowed with respect to the ultimate issue
because it was considered to intrude upon the exclusive province of the fact-finder. In the name of liberalization, the ultimate
issue rule was abandoned. However, the requirement that all expert opinion must be based on the scientific knowledge base
of the experts profession was not abandoned. Thus, for example, FRE 704, in lifting the common law ban on testimony that
speaks to the ultimate issue, expressly provides: testimony in the form of an opinion or inference otherwise admissible is not
objectionable because it embraces an ultimate issue to be decided by the trier of fact [italics added]. Clearly, an expert
opinion that runs afoul of FRE 702s requirement that it have a reliable basis would not be otherwise admissible. Thus, the
principal evidentiary issue that we have raised *279 in the lead article is a basis issue, not the more formulaic, semantic
question of ultimate issue testimony.

J udge Dessau also raises the concern that the more conservative empirical model we propose might result in the loss of
potentially valuable information. We would note that nowhere in the lead article or elsewhere in our writings do we
advocate expelling the baby with the bath water. Our model is designed as much to ensure that empirically valid testimony is
received as to ensure that pseudoscience and intuitive judgments are excluded. We propound simply that information in the
form of inferences, conclusions, and opinions is only as valuable as it is empirically valid. Opinions are valuable--indeed they
qualify as expert as opposed to personal opinions--only because they are drawn from or are tightly tied to the scientific

13 cv 1944 CAB BLM

knowledge base of the psychology discipline, not because the individual witness has particular credentials or magical powers
of intuition.

The concern expressed by J udge Dessau with respect to counselor-decided outcomes is a valid one and any experienced
family law practitioner can attest to the fact that many cases are settled not because the compromising parent receives the
evaluators conclusions as an epiphany of parenthood or even sees the report as valid or true but, rather, because an adverse
report is perceived--wrongly in our view--as an insurmountable obstacle to a positive litigation outcome. All too often, the
attorney for the disfavored parent is insufficiently schooled in the content of empirical behavioral research and the strictures
of scientific method to appreciate the enormous opportunities that exist therein to dispute an evaluators conclusions through
effective cross-examination. Thus, they recommend that their clients throw in the towel simply because they do not know
how to fight the fight or, even worse, they do not realize that there is a legitimate battle to be fought.

This comports further with J udge Dessaus concern that expert testimony must be scrupulously scrutinized: On what
material did he or she rely? What is the precise nature of the experts expertise and experience? Is he or she from an agency
which, for example, has an overriding philosophy that a purported victim of sexual abuse is always to be believed, or one
sympathetic to alleged perpetrators? How does the report sit with the rest of the evidence? Is it based on factual findings that
the court can make? Has the expert been given the opportunity to reconsider in the light of the evidence in court? Is the expert
defensive or partisan?

These and many other areas of inquiry are important considerations to be sure but they focus only on the minor premises of
the experts opinions, to wit, they ask the question, in effect, Did this expert do it right? The more fundamental question,
and the one that is the focus of the lead article, is, Can any expert do it at all? For all of the reasons stated in our lead
article, if the issue purportedly being answered by the expert is the best interest question, then the answer is surely no, at
least at this point in time given the state of the research. Until such time as there is empirical evidence that such answer does
repose within the scientific knowledge base of the behavioral science discipline, the answer will continue to be no and
issues of methodology will be secondary.

This is a mistake many lawyers make in their cross-examinations. They may cross-examine effectively with respect to
pointing out flaws in logic or deviations in methodologies from one or more of the published protocols and guidelines. But
until such time as there is empirical basis for defining what best interest really means and deriving a methodology to
predict it reliably and accurately--in terms of psychological construct as opposed to the socio-moral terms of best interest
as a legal standard--there is no valid major premise to support such ultimate opinions, a point that will receive ample and
penetrating attention if the cross-examination is effective.

*280 Finally, J udge Dessau raises the question whether the adversarial system is appropriate in custody litigation where the
court is attempting to fashion the future as opposed to most other types of lawsuits where the emphasis is on reconstructing
past events. The related question is posed as to whether the European model may present a better alternative than the
Anglo-American common law tradition.

Yes, the custody court is looking to the future but, as in most endeavors of planning or predicting the future, we are informed
by history. Napoleon once said that [h]istory is the version of past events that people have decided to agree upon. In
custody litigation, when the parties are unable to agree, they proffer competing histories, competing versions of the truth, and
the one that is believed then informs the courts decision with respect to the future. Custody litigants, no less than any others,
deserve the opportunity to present their competing versions in a jurisprudential setting that is reasoned, predictable, and
protective of due process. Thus, the European model, which, as J udge Dessau presents it, seems to meld settlement endeavors
with the adjudicative function, and appears to give the court carte blanche as to what procedures will be employed and what
evidence admitted, is precisely what ought not be done. It conflates the adjudicative process with the settlement process and,
thus, weakens the certainty and due process that comes from stringent observation of procedural protocols and vigilant
application of the rules of evidence. In our view, quite the opposite approach has more merit. A dual track approach would
allow the parties to negotiate, mediate, and consider opinions for what they are worth to achieve settlement. But, if in the
end all such settlement efforts prove futile, then the case ought to be adjudicated upon plenary courtroom trial, one which
employs strict adherence to evidentiary doctrine and common law trial procedure.

Stahl (this issue) emphasizes, as did Kelly and J ohnston, the rich twenty-five-year history of research related to a variety of

13 cv 1944 CAB BLM

divorce-relevant factors that he suggests supports the notion that recommendations can be made. He further asserts that he
believes competent evaluators can make Level III analyses and Level IV recommendations. Unfortunately, in a court of law
where fundamental rights are at risk, believing it to be so is not sufficient. The existence of divorce-related research in no way
establishes the capacity of an individual evaluator to make accurate custodial predictions within an acceptable margin of
error. It is with regard to the weighting of the complex set of psychological variables, parental assets and deficits, child needs,
and so on, and the related production of a prediction and recommendation, that we have inadequate empirical support.

We are in agreement with Stahl that mental health professionals can provide specialized guidance with regard to treatment
needs, conflict resolution recommendations, domestic violence interventions, and so forth. However, Stahls assertion that
judges should consider recommendations to be advisory only ignores decades of judgmental heuristics research (e.g., Garb,
1998; Turk & Salovey, 1988) that makes clear the powerful and highly directive nature of early impressions. That is, once
having an evaluators recommendation lodged in ones brain as a judge, the likelihood of dislodging it may be quite low
and the entire case, according to the judgment research, is likely to be viewed through the lenses of the recommendation that
appears on the judges desk from that point forward, possibly affecting how subsequent information is gathered and
interpreted.

Stahl mentions that we suggest that lay people can provide the same observational data that expert evaluators can provide,
suggesting that we have little more to offer the court. A careful reading of our article will make clear that we never
indicated this. In fact, we said quite the opposite: Evaluators bring the same sense organs as all other observers, but they also
bring to the task a set of observational tools and techniques that help define the status *281 of expert. We do note,
however, that at Level I evaluators are often making observations that lay people can also make, for example, father slapped
his daughter in the waiting room, mother yelled at father across the room, and so on. This was mentioned with regard to
evidentiary doctrine to draw a clear distinction between factual testimony and inferential expert testimony.

Stahl expresses significant concern about our willingness to refer to the making of custody recommendations as unethical and
he mentions the fact that neither the licensing boards nor the APA has made such a determination. He is concerned about the
increased burden that our ethical assessment will place on practitioners. We would respond that the burden should be
dramatically increased because of what is at stake. To wait for licensing boards or the APA to render ethical decisions about
this specific practice surrenders ethical decision-making responsibility to authorities rather than recognizing that ethical
documents are meant to be interpreted and applied in the community by individual practitioners on a day-to-day,
case-by-case, procedure-by-procedure basis, hopefully with the guidance of fellow practitioners and ethical experts, and
informed by dialogues such as that contained in this issue. In short, it is the practitioners responsibility to make decisions
about what is ethical and what is not, using the APAs document as a template. If it is not unethical to imply, via a
recommendation to a court, that ones instruments and methods can do something--a life-changing something--that they
have not been shown to be able to do, this is a very worrisome state of affairs for our profession.

Stahls understandable concern about suddenly having a common professional practice recast as potentially unethical is quite
understandable because of the bind in which evaluators would be placed. However, once the paradigm shift we propose is
supported by the legal profession, custody evaluators may well find it quite freeing to no longer imply, via concrete and
specific recommendations, that they possess a level of predictive power that has never been shown to exist. If parents or
children find themselves before a court of law and their life is about to be substantially redirected by the words uttered by a
mental health professional, they have every right to ask, Doctor, since you have been so willing to very specifically tell the
court which custodial plan you believe is best, would you kindly cite those empirical studies that establish that your discipline
can accurately make such a specific designation? Given the current state of the research, the silence should be deafening.

CONCLUSION
Direct parent dialogue, parenting education, attorney negotiation, and/or mediation blessedly produce relatively peaceful,
agreed upon resolutions in the vast majority of custody disputes. But when such efforts fail, families turn to the adversary
system for adjudication. When they do, no less than any other litigant, they are entitled to due process, i.e., a reasoned and
predictable procedure for determining facts and rendering a binding disposition.

Because there is no source of perfect knowledge, the adversarial system sets up what is supposed to be a relatively even
contest between two versions of reality. Each party submits to the court, subject to highly specific, controlled, and predictable

13 cv 1944 CAB BLM

procedural precepts and evidentiary rules designed to safeguard the fundamental liberties of the litigants, his or her version of
reality. In the end, one of those versions, or a third, blended version, emerges as the more credible of the two and is the basis
for the courts decision. As noted in the lead article and here again, as well as throughout the thoughtful responses by others
in this issue of FCR, it is acknowledged that the psychology discipline can offer substantial empirically *282 derived
information that can inform the custody court as it grapples with the best interest determination.

It is here, in the crucible of the contested litigation, that the principles of empirical proof must be vigilantly observed. It is
also here, that the principles of law and the tenets of the psychology discipline find a point of convergence in the empirical
standard. Psychology did not become a distinct and valid discipline, standing apart from philosophical speculation about the
nature of human behavior and mental processes, until it embraced the empirical standard in the mid to late nineteenth century
(Schultz & Schultz, 2004; Stanovich, 2004). Likewise, the trial court did not become a valid gatekeeper of evidentiary
reliability in its own right, as opposed to a mere delegator of that responsibility, until the U.S. Supreme Court rejected the
Frye test in Daubert and enunciated an empirical standard for the admissibility of expert testimony predicated upon science
(Tippins, 2004; Faigman et al., 2002). Thus, in very real and practical terms, the empirical standard represents a crucial point
of convergence of the two professions.

Thus, it is here, in this legal milieu of contested custody litigation, where mental health professionals must cleave most
faithfully to their empirical foundation and where judges must exert extraordinary vigilance as gatekeepers. Only through
such fidelity and vigilance can we ensure that only those opinions predicated upon empirically established knowledge are
allowed to impinge upon the fundamental rights and human liberties that are at stake when the issue before the court is the
custody of a child.

REFERENCES
Daubert v. Merrell Dow Pharmaceuticals, Inc., 43 F.3d 1311, 1319 (9th Cir. 1995).

Faigman, D. L., Kaye, D. H., Saks, M. J ., & Sanders, H. (2002). Modern scientific evidence: The law and science of expert
testimony. St. Paul, MN: West.

Frye v. United States, 293 F. 1013, 1014 (D.C. Cir. 1923).

Garb, H. N. (1998). Studying the clinician. Washington, DC: American Psychological Association.

Goldstein, J ., Freud, A., & Solnit, A. J . (1979). Beyond the best interests of the child. New York: Free Press.

ODonohue, W., & Bradley, A. R. (1999). Conceptual and empirial issues in child custody evaluations, Clinical Psychology:
Science and Practice, 6, 313.

Schultz, D. P., & Schultz, S. E. (2004). A history of modern psychology (8th ed.). Fort Worth, TX: Harcourt.

Shuman, D. W. (2002). The role of mental health experts in custody decisions: Science, psychological tests, and clinical
judgment. Family Law Quarterly, 36, 135-162.

Stanovich, K. E. (2004). How to think straight about psychology (7th ed.). Boston: Allyn & Bacon.

Turk, D. C., & Salovey, P. (Eds.). (1988). Reasoning, inference, & judgment in clinical psychology. New York: Free Press.

Wise, D. (2004, October 15). Overuse of forensic experts criticized at matrimonial panel. New York Law Journal, p. 1, col. 3.

Footnotes


13 cv 1944 CAB BLM

a1

Author Note: We wish to thank FCR for publishing our four-level model for analyzing the custody evaluation process and our
critique of how the conclusions expressed by custody evaluators ought to be scrutinized in the courtroom in terms of evidentiary
admissibility.

43 FAMCR 270
End of Document



13 cv 1944 CAB BLM
MENTAL HEALTH EVALUATIONS IN CHILD CUSTODY..., 43 Fam. Ct. Rev. 187

Family Court Review
April, 2005
Editorial Notes
MENTAL HEALTH EVALUATIONS IN CHILD CUSTODY DISPUTES
Andrew Schepard
Hempstead, New York
Copyright 2005 by Association of Family and Conciliation Courts; Andrew Schepard
This issue focuses on the role of mental health evaluations (also called child custody assessments, or evaluations) in the child
custody dispute resolution process, a topic of great interest to the interdisciplinary membership of AFCC. Indeed, AFCC has
been a national leader in promoting the professionalization of child custody evaluations by drafting standards of practice,
conducting symposia, and developing publications such as this issue to help develop the field.

Court-appointed mental health experts are a relatively new--in legal system time--innovation in child custody disputes, a
reaction to the demise of the tender years doctrine in the late 1970s and the certainty of outcome it provided. Instead of
gender-based assumptions and moral concerns, post tender years courts make a multifactored inquiry into the childs
emotional best interests to determine custody when parents cant agree. Today, courts routinely appoint mental health experts
as court witnesses with investigative powers and the power to provide an evaluation of family functioning to help them
decide on a parenting plan.

Extensive use of court appointed neutral experts is one of the features of family law that distinguish it from other fields of
law, such as torts and contracts. In most civil cases, courts do not appoint their own expert, but allow conflicting testimony
from party-paid experts to be weighed by a jury. While judges, not juries, decide most custody disputes, mental health
testimony originally followed the party-controlled, adversarial model of the ordinary civil case. Party-paid mental health
testimony, however, proved itself to be highly unsatisfactory. Party-paid experts did not have access to the entire family or
the children, and could not base their testimony on reliable clinical observations. Many party-paid experts could not resist the
temptation to testify favorably to the party who paid them, placing opinions in court records based on hypothesis, not
empirical research.

Court appointment of neutral mental health evaluators in custody disputes is thus, in some ways, a healthy social
development because they are a step away from the adversarial model of child custody decision making. A court-appointed
neutral expert is a symbol that courts recognize that something more is at stake in a custody dispute than the grievances of
one party against another-- the welfare of the child.

Neutral evaluators generally perform a valuable service for the court, parents, and children. The court receives valuable
information from a credible, neutral source about what is in the childs best interests. The evaluators report serves as a check
on the biases and idiosyncrasies of the judge. The experts report is also a strong and generally positive influence toward
settlement of the custody dispute. The parents and their lawyers know that the evaluators report is, in effect, a preview of the
courts ruling. The parent who receives a less than favorable report will think seriously about settling rather than pushing the
dispute to trial, where the outcome could be even worse. The evaluators scrutiny motivates *188 parents to control their
most aggressive actions and impulses toward each other for fear that the expert will draw unfavorable inferences.

The rapid growth of neutral mental health evaluation in child custody disputes has led to problems well captured in the
headline of a recent front page article in the New York Times on the subject--For Arbiters in Custody Battles [meaning neutral
mental health evaluators], Wide Power and Little Scrutiny.
1
J udges are elected by the people of a state, or appointed by the

13 cv 1944 CAB BLM

governor and confirmed by the legislature, a process that gives their power to make decisions in so vital an area as
parent--child relationships social and political legitimacy. A mental health evaluator is not a state official, and is not
accountable to anyone but the appointing judge. There is a great danger that the appointment of a neutral mental health
evaluator is, in effect, an improper delegation of judicial authority, that instead of being an aid to a judges decision, the
evaluator becomes the de facto judge.

There has also been little systematic scrutiny and standardization of role and techniques of neutral mental health evaluators
which has been until recently, in effect, a cottage industry. In complicated cases, neutral mental health reports can run to
hundreds of pages, and cost thousands of dollars. Some evaluators believe in the importance of psychological testing; some
do not. Some meet with parents and try to encourage settlements; others do not. Some evaluators are confident enough in
their judgments to make recommendations to the court for a specific parenting plan; others believe they do not have a
sufficiently rigorous research base to make such judgments.

This issue addresses the role of the neutral mental health evaluator in child custody disputes in all of its complexity. It began
with a chance meeting with an old professional acquaintance, Tim Tippins, at the New York State Divorce Mediation
Council meeting in April 2004, where we were both speaking. Tim is one of New Yorks best known (and funniest) divorce
litigators. He has served in many leadership roles in the divorce bar during his long and distinguished career--President of the
New York Chapter of the American Academy of Matrimonial Lawyers, Chair of the New York State Bar Association Task
Force on Family Law, and Chair of the NYSBA Family Law Section. Tim has also displayed a scholarly and academic turn
of mind, authoring New York Matrimonial Law and Practice, a multivolume treatise published by West Group, and teaching
at Albany Law School since 1987.

Tim writes a regular Matrimonial Law developments column for the New York Law Journal, New Yorks daily newspaper
for the legal community. Around the time of our chance meeting, he had been focusing on the shortcomings of mental health
testimony in contested custody disputes in that forum. My own book had just been published, and we shared some thoughts
on our mutual passion (though from different means and frameworks) for improving the way the legal system treats child
custody disputes. Tim told me he and a psychologist coauthor, J eff Wittmann, had pulled their thoughts about child custody
evaluations together into an article questioning whether mental health child custody evaluators should ever make specific
recommendations to the court. I knew of J eff as a psychologist, family therapist, custody evaluator, and divorce mediator in
Albany, New York. His practice concentrates on services for divorcing families, including the Kids First After Divorce
educational program. When Tim asked me if FCR would be interested in publishing his and J effs article, I immediately
replied yes.

Tim and J effs piece was what I expected--thoughtful and scholarly, but longer than the typical FCR article. I suspected it
would generate controversy and healthy debate in the AFCC and child custody evaluation community.

*189 I discussed the articles publication with the FCR Editorial Board at our meeting in San Antonio at the AFCC Annual
Conference in May 2004. The Board unanimously agreed that the article was an important contribution to the field that FCR
would be honored to publish. The Board also felt we should invite judges, lawyers, academics, researchers, and child custody
evaluators to comment on the piece and give Tim and J eff an opportunity to reply. Peter Salem, AFCCs Executive Director,
also invited Tim and J eff to present their views at the AFCC Child Custody Symposium in Nashville, Tennessee in October
2004 as part of an extensive program on whether child custody evaluators should make specific recommendations to the court
for a parenting plan.

FCR then recruited experts from diverse backgrounds--law professors, mental health experts, researchers, judges, child
custody evaluators in different practice settings and in different countries--to comment on Tim and J effs article. The
comments are led off by an introduction by Leslye Hunter, AFCCs President and a working child custody evaluator.
Comments from Thomas Grisso, Mary Kay Kisthardt and Barbara Glesner Fines, J oan Kelly and J anet J ohnston, Arline
Rotman, Lorraine Martin, J onathan Gould and David Martindale, Philip Stahl, and Linda Dessau follow. It would take me a
good deal of space to recite the credentials and outlook of each of these thoughtful and able people. I leave it to readers to
look at the end of each article for a brief biography of our distinguished authors. It suffices for me to simply thank each of
them, and Tim and J eff again, for their willingness to share their thoughts with FCR readers and move the ongoing dialogue
about the role of mental health evaluations in child custody disputes forward.


13 cv 1944 CAB BLM

OTHER RICHES IN THIS ISSUE
One of the most difficult problems facing the family court is the evaluation of allegations of child abuse, neglect, and family
violence in child custody disputes between divorcing parents. The best way to clarify and address this emotionally charged
problem is with data, not polemics. FCRs Associate Editor, J anet J ohnston, and her colleagues Soyoung Lee, Nancy Olesen,
and Majorie Walters describe a study of 120 divorced families referred for child custody evaluations. Multiple allegations of
child abuse, neglect, and family violence were raised in the majority of cases, and about half of the alleged abuse was
substantiated in some degree. One fourth of the substantiated allegations involved both mother and father within the same
family. One conclusion the authors draw from their findings is that allegations of abuse, neglect, and violence in a custody
dispute cannot be dismissed as a litigation tactic; nor can the court assume that there is one good and one bad parent. The
only way that the complex problems of these high-conflict, violent families can be effectively addressed is by coordinated
case management protocols that continuously focus all concerned on the welfare of the children.

Barry Nobel contributes an imaginative article in which he asks and answers the question of what meditation can contribute
to mediation. Readers will, I think, be drawn to Barrys wide-ranging discussion of the sources of meditative wisdom and his
analysis of how meditation theory can enrich mediation practice.

FCRs law student staff contributes two notes on different facets of family mediation. Michael Lindstadt discusses how
mediation can address the social problem of school truancy. Melissa Schoffer discusses how children can be involved in
mediation in family disputes.

The issue concludes with two book reviews. J anet J ohnston reviews Bob Emerys new book on children and divorce and
Susan Zaidel reviews Linda Fisher and Micke Brandons new book on family mediation.

Footnotes

1

Leslie Eaton, For Arbiters in Custody Battles, Wide Power and Little Scrutiny, N.Y. TIMES, May 23, 2004, at 1, col. 1.

43 FAMCR 187
End of Document



13 cv 1944 CAB BLM
COMMENTARY ON TIPPINS AND WITTMANNS EMPIRICAL..., 43 Fam. Ct. Rev. 233

Family Court Review
April, 2005
Commentary to Tippins and Wittmann
COMMENTARY ON TIPPINS AND WITTMANNS EMPIRICAL AND ETHICAL PROBLEMS WITH
CUSTODY RECOMMENDATIONS: A CALL FOR CLINICAL HUMILITY AND JUDICIAL VIGILANCE
J oan B. Kelly
a1
J anet R. Johnston
a2

Copyright 2005 by Association of Family and Conciliation Courts; J oan B. Kelly, J anet R. Johnston
Although in substantial agreement with Tippins and Wittmanns analysis, their call for a moratorium on the practice of
custody evaluators making recommendations to the court does not solve the many problems that they have raised, and may
have unintended consequences which place families at even greater risk. This commentary reflects our agreement with some
of the authors major points of contention, focuses on several points of disagreement, and suggests alternative remedies for
the shortcomings and ethical problems described in child custody evaluations.

Keywords: child custody evaluation; custody recommendations; child adjustment and divorce research

Tippins and Wittmann (this issue) have written a thoughtful, well-researched, and provocative article that merits response.
We fully agree with them that the custody evaluation and related judicial decision has the potential to change the entire
course of childrens lives, including the extent and meaningfulness of their parent-child relationships, emotional and social
adjustment, school functioning, and economic well-being. While their article focuses on psychologists as custody evaluators,
in many family courts mediators, counselors, and guardians ad litem provide judges with opinions and recommendations as
to custody and access, sometimes after spending only one to two hours with the parents. Tippins and Wittmanns concern
and caution are warranted, and should serve as a clarion call to the entire family law field involved in custody and access
determinations.

The authors four-level conceptual model for stratifying data and clinical inferences provides a helpful framework for mental
health and legal professionals to examine the evaluation process, whether or not one agrees with their conclusions. Custody
evaluators without sufficient scientific training are unaware of the serious limitations of the data they collect, the validity of
the testing instruments they use, or the rigor needed to make inferences and draw conclusions from this information. Instead,
the authors contend, such custody evaluators are more likely to make inferences and recommendations from unsubstantiated
theory, personal values and experiences, and cultural and personal biases. Our own observations and reviews of evaluations
over several decades lead us to the same conclusion. Common examples include unexamined strong beliefs in the primacy of
mothers (or essentiality of fathers) regardless of the circumstances, biased perception of their clients derived from their own
negative marital and divorce experiences, or a conviction that joint physical custody benefits (or harms) all children.

Too few custody evaluators are well acquainted with, and make use of, the existing body of empirical literature on divorce,
parenting, child development, and childrens emotional *234 and social adjustment for purposes of formulating questions to
guide the collection of data, and for making inferences and recommendations. Too often simplistic research, replete with
outdated findings and formulations is used in the area of divorce and attachment theory and its effects on childrens
adjustment; overgeneralizations from empirical data are also especially common. Parental conflict, for example, is often
treated as an undifferentiated variable without awareness of the different effects on children; the intensity and content of
conflict; whether conflict is expressed in avoidance, angry words, hostility, or physical violence; whether parents protect or
expose their children to their differences and unresolved disputes, and whether buffers exist that ameliorate the potentially
negative effects of high conflict.

In their discussion of Level III inferences arising from Level I and II observations, Tippins and Wittmann are correct to infer
that too many evaluators state as facts certain conclusions without disclosing that some of these issues or conclusions are
the focus of considerable disagreement if not raging controversy in the field, or that these facts are not supported by

13 cv 1944 CAB BLM

well-designed, replicated, empirical research. A common example is the position that if overt alienating behaviors are
identified in one parent when a child resists visitation, custody should be shifted to the other parent. The controversy in the
field regarding overnight visits for very young children is another example. Without reference to any empirical data that
supported his or any other viewpoint, one evaluator stated, This five-year-old child will be irreparably damaged were she to
spend even one overnight away from her mother.

The immediate and obvious implication of Tippins and Wittmanns valid concerns is that more stringent ethical,
professional, and scientific standards of practice should be required of custody evaluators with respect to drawing clinical
inferences. This surely requires higher standards for initial training and certification to perform custody evaluations, as well
as ongoing professional development to update the evaluators knowledge base in social science and law at regular intervals.
It is striking that the AFCC Standards of Practice for Child Custody Evaluations (1994) do not mention knowledge of
relevant research as an important aspect of the process, nor does it raise issues of reliability or validity. Hopefully, this will be
remedied in a revision of the standards currently being prepared by the AFCC Task Force.

The contribution of evaluators involves collecting Level I data and these behavioral observations are especially valuable
when guided by key concepts and variables abstracted from relevant research. For example: With respect to this childs
ability to separate from one parent and transition to the other, it was noted that at each of the four interviews the child clung
to her mother and resisted entering the room where her father was waiting. Level II data-- conclusions about the psychology
of the parent, child, and family--can be made, provided that the custody evaluator can support these clinical inferences by
citing relevant research studies from the literature within the custody report. For example According to criteria developed by
Main, Kaplan, and Cassidy (1985), this six-year-old child has an ambivalent attachment to her mother (marked by clinging
dependency and intermittent hostility) and a disorganized attachment to her father (marked by a mix of fear, avoidance and
passive compliance).

POINTS OF DISAGREEMENT
Tippins and Wittmann take the position that few statements can be made at Level III that are within the requisite ethical and
scientific parameters. By this they mean, conclusions based on Level I and II data relevant to custody and access questions,
such as a particular *235 childs functioning and needs, or the fit between a childs needs and parental abilities. They
concede that clinicians can be useful to the court at Level III when they summarize important psychological risks or
advantages regarding certain parenting plans, but only if they provide clearly articulated qualifications, cautionary
statements to the court, and references to the limitations of the evaluation methods used (Tippins & Wittmann, this
volume).

In support of their argument, the authors cite the absence of predictive validity for such constructs as parent-child fit,
different parenting plans, and future child outcomes. They also cite the fact that empirical research related to Level III
conclusions is quite limited. We take issue with this latter point, and suggest that, provided the appropriate precautions are
taken with respect to Level III summaries and conclusions, the authors are more conservative than they need to be. There is
far more empirical research describing specific factors associated with risk and resiliency in children following divorce than
the authors appear to indicate. While it is true that most empirical research was designed to assess the various impacts of
separation and divorce on childrens adjustment, rather than test predictive statements about custody or access plans, the
results of three decades of increasingly sophisticated research has provided a more complex understanding of variables
associated with risk and resilience following divorce to guide the well-informed evaluator.

Well-designed empirical studies point to the negative impact of parental depression, anxiety, mental illness, and personality
disorders on child adjustment during marriage and after divorce. Diminished parenting following separation and the
importance of postdivorce parenting characterized by warmth, emotional support, adequate monitoring, authoritative
discipline, and age-appropriate expectations have been widely reported. Risk and protective factors associated with
attachment processes, loss of relationships, long-term parent-child relationships, academic functioning, school drop-out, and
higher education have been described. Access frequency, shared physical custody, quality of father-child relationships,
father-child closeness in relation to different access patterns, childrens resistance to visitation, childrens views of their
access patterns concurrently and retrospectively, and more recently, overnight visits for young children, all in relation to
childrens adjustment, have received substantial research attention. Multiple dimensions of the effects of high conflict and
parent violence on child adjustment, and protective buffers against conflict have been reported, as have the various impacts of

13 cv 1944 CAB BLM

remarriage and re-partnering. Articles and books reviewing the empirical literature on children and divorce contain
descriptions of and citations to such relevant variables and studies (Amato, 2000; Barber, 2002; Emery, 1999; Hetherington,
1999; Hetherington & Kelly, 2002; Kelly, 2000; Kelly & Emery, 2003; Maccoby & Mnookin, 1992; Pruett, Williams,
Insabella, & Little, 2003; Waxler, Duggal, & Gruber, 2002), and evaluators should search out updates regarding newer
empirical research on such controversial variables as overnight visits for young children (Pruett, Ebling, & Insabella, 2004),
child alienation (Johnston, 2003; Johnston & Kelly, 2004), and shared physical custody and conflict (Bausermann, 2002;
Lee, 2002). Research reviews of various clinical, mediation, and arbitration interventions can inform the evaluator as to the
effectiveness of the services they suggest (Johnston, 2000; Kelly, 2002, 2004; Pruett & Johnston, 2004).

Hence, we suggest that to the extent that available research can be cited that includes studies of variables pertaining to child
custody and access, circumscribed inferences at Level III might be drawn. Following our example of the six-year-old child
who had difficulty transitioning between parents, the evaluator could state:
Research studies show that hostility and unresolved parental conflict undermines parenting capacities and
negatively impacts the child (Cummings & Davies, 1994; Krishnakamur & *236 Beuhler, 2000; Tschann,
Johnston, Kline, & Wallerstein, 1989), and that following divorce, expressing anger in the presence of or
through the child is associated with childrens depression and anxiety (Buchanan, Maccoby, &
Dornbusch, 1991; Hetherington, 1999). In this family, the fathers expressed anger toward the mother, in
the presence of the child appears to contribute to the mothers distress and emotional unavailability to the
child as well as the childs fear of him. Without remedying this family dynamic, shared physical custody
arrangements that require parents to communicate will probably be detrimental to this child.

In more complex cases, the custody evaluator must draw upon a number of different studies and weigh the beneficial buffers
that are present (like a good parent-child relationship) with the risks that are posed (like ongoing parental conflict and
parental psychopathology) together with the resiliency of the child (age, temperament, etc.) in order to support Level III
conclusions regarding beneficial access patterns. Thus, for example, if the evaluator observes a close, supportive relationship
between a father and his nine-year-old son, but also notes a high level of interparental conflict, the following could be stated:
Traditional access of every other weekend is experienced as insufficient time and distressful for most
children and is likely to erode the childs relationship with the nonresidential parent over the long term
(Kelly & Emery, 2003). In contrast, more frequent access between children and fathers when the
relationship is positive is associated with better behavioral adjustment and academic performance in
children (Amato & Gilbreth, 1999), especially when fathers are more involved in childrens school
activities and projects and provide authoritative discipline and emotional support (Amato, 2000;
Menning, 2002; Nord, Brimhall, & West, 1997; Simons, 1996). Studies point to the benefits of
substantial time with both parents even in the presence of parental conflict provided that parent-child
relationships are positive, with diminishing benefits of shared residential arrangements in the presence of
very high and sustained conflict (Bausermann, 2002; Lee, 2002). In this family, the warmth and support
in both parent-child relationships appears to buffer this bright and adaptive boy against the potential
negative effects of his parents conflict (Emery, 1999; Kelly, 2000). This is a fairly recent separation and
the parents have no significant psychopathology. It is therefore unlikely that these parents will be among
the 8-15% of parents who continue in high conflict several years after divorce (King & Heard, 1999).

Such references to the research can form the basis for the evaluators comments at Level III that the customary jurisdictional
guideline of four days per month would be depriving and unsatisfactory for this particular child, and has the potential to
cause more negative adjustment problems.

Tippins and Wittmann have taken a strong position--based on the American Psychological Association Ethical Principles of
Psychologists and Code of Conduct (2003)--that it is unethical for evaluators to make recommendations to the court (Level
IV inferences) because of the limitations of the data collected, and the lack of empirical knowledge that links data to specific
parenting plans and outcomes for children. Moreover, they back up this position by a legal objection, arguing that
recommendations to court amount to the custody evaluator usurping the role of the judge as trier of fact.

In cases where trained and experienced evaluators have collected data systematically, used valid and reliable testing

13 cv 1944 CAB BLM

instruments, and linked their Level I and II observations and data to Level III conclusions, citing empirical research to
support these conclusions, then we believe it is ethical to make recommendations as to custody and access that provide the
trier of fact with some options for parenting plans that might benefit this particular child. The contribution at Level IV could
involve the custody evaluator offering a series of *237 alternative hypotheses, predictions about the future functioning of the
child under different custody and access scenarios, also backed by research findings. To return to our example of the
six-year-old girl, the evaluator could state:
According to Maccoby and Mnookin (1992) and Hetherington (1999), it is rather unlikely that highly
conflicted parents will develop a cooperative co-parenting relationship within the next few years. Hence,
primary residence with one parent and a specific, clearly structured access plan with a neutral place of
exchange is likely to be more supportive of this childs security. Alternatively, if parents, especially the
father, successfully complete psycho-educational parenting counseling, then a more shared parenting
arrangement might be warranted (Arbuthnot, Kramer, & Gordon, 1997).

More definitive recommendations at Level IV might be made in the case of the nine-year-old boy described above: While no
research exists to support one specific parenting plan, the research cited suggests this child would most likely benefit from
liberal access to the father during some part of each week, including a stable pattern of school days and weekend time.
Requiring the custody evaluator to document knowledge claims and to present clinical inferences and hypotheses that are
conditional would ensure more accountability. It would also allow appropriate challenge from competing facts, theories, and
research findings.

While we have indicated our agreement with many of the arguments made by Tippins and Wittmann, we disagree with the
authors call to place a moratorium on making recommendations because the alternatives are dismal, even destructive, and
may cause even more harm to families and parent-child relationships. We do not agree that judges should be left to make the
final decision without any input from custody evaluators or others about what is considered to be in the best interests of the
child. In the face of this vague legal mandate, judges are even less qualified in training and experience than are mental
health professionals to address this question without undue influence of their personal biases.

In the absence of recommendations by custody evaluators who have considered each child and family situation in great
depth, judges and legal advocates will probably rely more and more on prescriptive guidelines (like primary residence with
one parent and every-other-weekend with the other). It is even more likely that judges decisions will be governed by
presumptions that will increasingly be cast into statutes by political and professional interest groups with access to the state
legislatures (like the American Law Institutes approximation rule, or a primary parenting presumption advocated by
womens lobbyists, or joint parenting presumptions touted by fathers rights groups). These prescriptive rules and
presumptions are not research based and they do not consider the individual needs of children and variations in parent-child
relationships. Rather, they are simplistic answers, a one-size-fits-all substitute for the vexing question of what is in the best
interests of each child.

A promising alternative is the development of parenting plans that could provide judges (and parents) with a range of
possible alternative dispositions on custody and access. Such model parenting plans would be based on the empirical
literature to date, and offer choices for the court that address the needs of different kinds of family situations and
developmentally appropriate options for different age groups. Some states (like Arizona) have piloted this approach, using an
interdisciplinary task force, with promising results (Arizona Supreme Court, 2001). It is indeed true that custody
arrangements are based on a mix of tradition, law, science, untested theory, and prevailing cultural values about child rearing.
For this reason, local task forces of community members made up of *238 interested citizens, and mental health and legal
professionals, could participate in the development and updating of these parenting plan models for the use of judicial and
parental decision making.

THE OVERUSE OF CUSTODY EVALUATIONS
Acknowledging the serious deficits in custody evaluations, particularly the flimsy grounds (ethically, empirically, and
legally) for making recommendations on the ultimate issue, leads one to question the appropriateness of this tool for
developing clarity and dispute resolution for many cases in family court. Clearly, evaluations can be more solidly grounded

13 cv 1944 CAB BLM

when they are investigating serious allegations of physical abuse, sexual abuse, and neglect of the child as well as mental
illness, substance abuse, and domestic violence on the part of parents. In these domains, community standards and values are
more clearly defined and the empirical research literature is more extensive and robust in its findings of what is not in
childrens best interests.

In the absence of such serious concerns about family abuse and neglect, mental illness, or substance abuse, too often custody
evaluations must focus on who is and who is not more emotionally healthy and the better parent. Personality testing is
undertaken with no solid basis for concluding how the findings might impact parenting (Brodzinsky, 1993; Roseby, 1995;
Tippins & Wittmann, this volume). Particularly in those cases where angry, hurt, but good enough parents are contesting
custody or the allocation of time sharing, there is generally no basis in psychology or law for choosing between parents.
Evaluators split hairs to make a case for one parent or the other, and the evaluators personal values and cultural biases are
likely to be more prominent in this decision making. This may also be the case when both parents have demonstrated
significant character or psychological problems and parenting deficits, and it is impossible to argue for a preference between
parents without relying on subjective reactions and biases.

In this quest, custody evaluations may have inadvertently produced de facto double standards, where those held up for
parents in family courts are far more stringent than those faced by parents in dependency courts. The result is that custody
evaluators are now producing exhaustive, intrusive, negatively biased assessments, psychological testing, and written reports
in which separating parents are scrutinized and held to a higher standard of accountability than those in nondisputing divorces
and intact families. This seems unfair, unnecessarily stressful for already vulnerable families, and may even constitute
grounds for claiming violation of parents civil rights. It is in the search for the elusive better or best parent that personal
values and cultural beliefs are likely to infiltrate and contaminate what is supposed to be a scientifically defensible
investigative process and report.

ALTERNATIVE REMEDIES
A better policy would be for forensic custody evaluations to be reserved for serious allegations of child abuse, neglect, and
molestation, as well as contested claims of parental psychopathology, substance abuse, or domestic violence, where standards
for parental behavior in family court would be more on a par with those in dependency court. Where parents have extremely
discrepant views of their childs needs, difficulty making decisions together in a timely manner, and co-parenting
disagreements that do not rise to the level of abuse allegations, the use of extended interventions such as confidential
child-inclusive mediation and therapeutic mediation (Kelly, 2002; Pruett & Johnston, 2004; Sanchez & *239
Kibler-Sanchez, 2004) and child-focused psychological and family assessment might suffice (Johnston, 2000; Roseby,
1995). In these confidential assessments of the childs needs, and the parents relative capacities to meet those needs, more
attention could be paid to prescribing how the family can resolve its impasse, the ways in which children can have access to
the positive contributions of each parent, and how the childrens development can be protected, which can be the basis for
further mediation, counseling, or a recommended settlement. It is important to note, however, that this information from a
mental health professional could not be admitted as evidence in litigation because it would not meet the higher standards of
expert testimony required in court. Instead, these kinds of assessments are educational and advisory tools to be used in
alternative conflict resolution forums and to promote change in parents.

It is important to note that it is quite feasible to undertake the kind of research that would address many of the predictive
dilemmas in custody recommendations, including those relevant to Level IV clinical inferences. However this will involve
commitment to longitudinal studies. The virtual absence of long-term outcomes of custody decisions made by family courts
is indeed an embarrassment, if not a scandal, in the field. Recent surveys of psychologists who undertake custody evaluations
show that data are being collected in a fairly standardized manner, a trend that will increase if standards of practice are
mandated (Bow & Quinnell, 2002). It would be relatively straightforward to create central databanks to store coded data,
ensuring confidentiality and appropriate protections for human subjects, thereby creating a database for follow-up studies that
could eventually make predictions based on actuarial data. In jurisdictions where judges choose among a range of parenting
plans for families, long-term research on the various outcomes for children of different plans could be initiated, providing
data for the next generation of better trained evaluators.

In summary, our proposals for dealing with the serious problems described by Tippins and Wittmann are: first, for better
training of forensic custody evaluators and adoption of more specific guidelines for practice from professional organizations

13 cv 1944 CAB BLM

like APA and AFCC, ensuring more stringent standards with respect to making clinical inferences. This includes evaluators
delineating the limitations of their data and conclusions, documenting their knowledge claims by citing relevant research, and
critiquing the pros and cons of different custody and access options, rather than recommending specific arrangements.
Second, the greater use of parenting plan options would serve to guide judges in making custody decisions and obviate the
need for them to rely upon simple presumptive rules that do not consider the developmental needs of children and specific
family situations. Third, forensic custody evaluations should be reserved for serious allegations of family abuse, domestic
violence, substance abuse, mental illness, and severe character pathology. When parental disputes do not rise to this level, the
use of advisory confidential assessments and feedback to parents would aid in settlement through alternative dispute
resolution forums of mediation and negotiation. Finally, we advocate for research on long-term outcomes of child custody
decisions, building databases in order to make predictions on what is in the best interests of the child that are based on
actuarial data.

REFERENCES
Amato, P. (2000). The consequences of divorce for adults and children. Journal of Marriage and Family, 62, 1269-1287.

Amato, P., & Gilbreth, J . (1999). Nonresident fathers and childrens well-being: A meta-analysis. Journal of Marriage and
the Family, 61, 557-573.

*240 Arbuthnot, J ., Kramer, K. M., & Gordon, D. A. (1997). Patterns of relitigation following divorce education. Family &
Conciliation Courts Review, 35, 269-279.

Arizona Supreme Court. (2001). Model parenting time plans for parent/child access. Phoenix, AZ: Administrative Office of
the Courts, Family Law Unit. Retrieved November 8, 2004, from www.supreme.state.az.us.

Association of Family & Conciliation Courts. (1994). Model standards of practice for child custody evaluations. In P.
Bushard & D. Howard (Eds.), Resource guide for custody evaluators. Retrieved November 8, 2004, from
www.afccnet.org/pdfs/Child_Model_Standards.pdf.

Barber, B. K. (Ed.). (2002). Intrusive parenting: How psychological control affects children and adolescents. Washington,
DC: American Psychological Association.

Bausermann, R. (2002). Child adjustment in joint-custody versus sole-custody arrangements: A meta-analytic review.
Journal of Family Psychology, 16, 91-102.

Bow, J . M., & Quinnell, F. A. (2002). A critical review of child custody evaluation reports. Family Court Review, 40(2),
164-176.

Brodzinsky, D. M. (1993). On the use and misuse of psychological tests in child custody evaluations. Professional
Psychology: Research and Practice, 24, 213-219.

Buchanan, C., Maccoby, E., & Dornbusch, S. (1991). Caught between parents: Adolescents experience in divorced homes.
Child Development, 62(5), 1008-1029.

Cummings, E., & Davies, P. (1994). Children and marital conflict: The impact of family dispute and resolution. New York:
Guilford Press.

Emery, R. E. (1999). Marriage, divorce, and childrens adjustment (2nd ed.). Thousand Oaks, CA: Sage.

Hetherington, E. M. (1999). Should we stay together for the sake of the children? In E. M. Hetherington (Ed.), Coping with
divorce, single parenting, and remarriage (pp. 93-116). Mahwah, NJ : Erlbaum.

Hetherington, E. M., & Kelly, J . (2002). For better or for worse: Divorce reconsidered. New York: W. W. Norton & Co.


13 cv 1944 CAB BLM

Johnston, J . R. (2000). Building multidisciplinary professional partnerships with the court on behalf of high-conflict
divorcing families and their children: Who needs what kind of help? University of Arkansas at Little Rock Law Review, 22,
453-479.

Johnston, J . R. (2003). Parental alignments and rejection: An empirical study of alienation in children of divorce. Journal of
the American Academy of Psychiatry & Law, 31(2), 158-170.

Johnston, J . R., & Kelly, J . B. (2004). Rejoinder to Gardners Commentary on Kelly and Johnstons The alienated child:
A reformulation of parental alienation syndrome. Family Court Review, 42(4), 622-628.

Kelly, J . B. (2000). Childrens adjustment in conflicted marriage and divorce: A decade review of research. Journal of Child
and Adolescent Psychiatry, 39, 963-973.

Kelly, J . B. (2002). Psychological and legal interventions for parents and children in custody and access disputes: Current
research and practice. Virginia Journal of Social Policy and Law, 10, 129-163.

Kelly, J . B. (2004). Family mediation research: Is there empirical support for the field? Conflict Resolution Quarterly,
22(1/2), 3-35.

Kelly, J . B., & Emery, R. E. (2003). Childrens adjustment following divorce: Risk and resilience perspectives. Family
Relations, 52, 352-362.

King, V., & Heard, H. E. (1999). Nonresident father visitation, parental conflict, and mothers satisfaction: Whats best for
child well-being? Journal of Marriage and Family, 61, 385-396.

Krishnakamur, A., & Buehler, C. (2000). Interparental conflict and parenting behaviors: A meta-analytic review. Family
Relations, 49, 25-44.

Lee, M-Y. (2002). A model of childrens postdivorce behavioral adjustment in maternal and dual-residence arrangements.
Journal of Family Issues, 23, 672-697.

Maccoby, E., & Mnookin, R. (1992). Dividing the child. Cambridge, MA: Harvard University Press.

Main, M., Kaplan, N., & Cassidy, J . (1985). Security in infancy, childhood, and adulthood: A move to the level of
representation. In I. Bretherton & E. Waters (Eds.), Growing points of attachment theory and research. Monograph of the
Society for Research on Child Development (Vol. 50(1/2); pp. 66-104). San Francisco: J ossey-Bass.

Menning, C. L. (2002). Absent parents are more than money: The joint effects of activities and financial support on youths
educational attainment. Journal of Family Issues, 23, 648-671.

Nord, C. W., Brimhall, D., & West, J . (1997). Fathers involvement in their childrens schools (NCES 98-091). Washington,
DC: National Center for Education Statistics.

Pruett, M. K., & Johnston, J . R. (2004). Therapeutic mediation with high conflict parents: Effective models and strategies. In
J . Folberg, A. Milne, & P. Salem (Eds.), Divorce and family mediation: Models, techniques, and applications (pp. 92-111).
New York: Guilford Press.

*241 Pruett, M. K., Ebling, R., & Insabella, G. (2004). Critical aspects of parenting plans for young children. Family Court
Review, 42(1), 39-59.

Pruett, M. K., Williams, T. Y., Insabella, G., & Little, T. D. (2003). Family and legal indicators of child adjustment to
divorce among families with young children. Journal of Family Psychology, 17, 169-180.

Roseby, V. (1995). Uses of psychological testing in a child-focused approach to child custody evaluations. Family Law

13 cv 1944 CAB BLM

Quarterly, 29, 97-110.

Sanchez, E. A., & Kibler-Sanchez, S. (2004). Empowering children in mediation. Family Court Review, 42(3), 554-575.

Simons, R. L. (Ed.). (1996). Understanding differences between divorced and intact families: Stress, interaction, and child
outcome. Thousand Oaks, CA: Sage.

Tippins, T. M., & Wittmann, J . P. (2005). Empirical and ethical problems with custody recommendations: A call for clinical
humility and judicial vigilance. Family Court Review, 43, 193-222.

Tschann, J ., Johnston, J ., Kline, M., & Wallerstein, J . S. (1989). Family process and childrens functioning during divorce.
Journal of Marriage and Family, 51, 431-444.

Waxler, Z. W., Duggal, S., & Gruber, R. (2002). Parenting and psychopathology. In M. H. Borstein (Ed.), Handbook of
parenting (Vol. 4; pp. 295-327). Mahwah, NJ : Erlbaum.

Footnotes

a1

Joan B. Kelly, Ph.D., is a consultant and lecturer and formerly Director of the Northern California Mediation Center and forensic
psychologist. Her research and practice focused on childrens adjustment to divorce, child development research and parenting
plans. Parenting Coordination with high conflict parents, and mediation interventions. jbkellyphd@mindspring.com

a2

Janet R. Johnston Ph.D. is a professor in the Department of Justice Studies, San Jose State University and formerly research
director of the Judith Wallerstein Center for the Family in Transition. She is a specialist in high conflict and violent divorcing
families and co-author of the books Impasses of Divorce and In the Name of the Child.

43 FAMCR 233
End of Document



13 cv 1944 CAB BLM
A SECOND CALL FOR CLINICAL HUMILITY AND JUDICIAL..., 43 Fam. Ct. Rev. 253

Family Court Review
April, 2005
A SECOND CALL FOR CLINICAL HUMILITY AND JUDICIAL VIGILANCE: COMMENTS ON TIPPINS
AND WITTMANN (2005)
J onathan W. Gould
a1
David A. Martindale
Copyright 2005 by Association of Family and Conciliation Courts; J onathan W. Gould, David A. Martindale
Offering commentary on Empirical and Ethical Problems with Custody Recommendations: A Call for Clinical Humility
and J udicial Vigilance, we express our agreement with Tippins and Wittmann. We opine that professional activity
stimulates inquiry and it is through inquiry that our knowledge base will be expanded and our skills will be improved. In the
interim, we should be mindful of our limitations, and actively articulate the ways in which the incompleteness of our
knowledge base and the deficiencies of our assessment tools limit the confidence that can be placed in the recommendations
that we offer.

Keywords: child custody; ultimate issue testimony; rules of evidence; clinical judgment; scientifically informed

We have been invited to comment on Empirical and Ethical Problems with Custody Recommendations: A Call for Clinical
Humility and J udicial Vigilance (Tippins & Wittmann, 2005). With a reasonable degree of clinical humility, we opine that
Tippins and Wittmann have articulated several noteworthy concerns relating to the certainty with which child custody
evaluators offer opinions to the court. We share their concerns about overreaching by mental health professionals in offering
to the court specific residential placement recommendations that purport to address the best psychological interests of a child
when such recommendations have little, if any, scientifically informed foundation. Our concerns, previously expressed both
individually and jointly, have been cited by Tippins and Wittmann.

Gould (1998) has cautioned about the certainty with which mental health professionals present recommendations to the
court. Specifically, he has opined:
Our tools are often not valid for custodial assessment. Our models are often rationally, not empirically,
derived. And our opinions are more educated guesses than truth. We need to be careful in how we present
our data and opinions to the court so as not to mislead (p. 38).

Martindale (2001) has written:
The defining attributes of an expert opinion relate ... to the procedures that were employed in formulating
the opinion and the body of knowledge that forms the foundation upon which those procedures were
developed. If the accumulated knowledge of the experts field was not utilized, the opinion expressed is
not an expert opinion. It is a personal opinion, albeit one being expressed by an expert (p. 503).

We believe the four-level system of analysis presented by Tippins and Wittmann provides a useful conceptual framework for
understanding the relationship between direct observation and level of certainty. We understand their model to indicate that
as the level of abstraction *254 from direct observational data increases, the level of certainty about the reliability of the
conclusion decreases.

We agree with Tippins and Wittmann that in discussions of the value (or lack thereof) of advisory reports from custody
evaluators, reliability and relevance should be the focus. When assessment tools are unreliable, the information obtained is
untrustworthy. When information is untrustworthy, we cannot know its relevance.

13 cv 1944 CAB BLM

Scientific expert testimony is based on empirical findings derived through systematic research in which variables under study
can be controlled and outcomes can be measured with reasonable precision. Research into the variables that are considered in
formulating parenting plans always will be constrained by the inability to use true experimental designs ... (Otto,
Buffington-Vollum, & Edens, 2003, p. 203). By utilizing reliable procedures and formulating opinions based on the
accumulated knowledge of their fields, mental health professionals can still meet the helpfulness standard (Daubert v.
Merrell Dow Pharmaceuticals, Inc., 1993). Gross and Mnookin (2003) discuss the problem of assessing reliability with
respect to nonscientific expert testimony. They write:
Evaluating the reliability of knowledge not produced through formal methods thus raises especially
difficult questions. As important as an examination of method, however, and much less noted, is another
dimension: the degree of certainty that the expert posits in what she offers. One of the central problems
with much expert testimony introduced in court--both scientific and non-scientific alike--is that experts
claim as matters of fact or probability opinions that should be couched in more cautious terms, as
possibilities or hypotheses (p. 143).

The most fundamental question to be asked by the court about expert information is whether it lies within a valid category of
expertise. We need to ask whether custodial fitness evaluation (as a subspecialty within the field of psychology) is based on a
sufficiently established body of knowledge and whether practitioners of this subspecialty utilize sufficiently reliable methods
and instruments in obtaining case-specific information. Only if our work rests on a sound foundation, and only if the methods
employed by us are reliable, can we provide the assistance that courts seek.

In the search for a better foundation, some psychologists have endeavored to construct instruments that would directly assess
parenting skills and parent-- child relationships. LaFortune and Carpenter (1998) evaluated the seven instruments most
frequently used and concluded: [T]he validity of these measures is unestablished at best and seriously flawed at worst (p.
222). There is much to be done. Existing instruments must be improved upon and new instruments must be developed. If, as
a profession, we retreat from custody work, the necessary research will not be undertaken.

Expert testimony is admissible only if it is the product of reliable principles and methods (Fed. R. Evid. 702). One of the
most difficult tasks for a judge is to distinguish between acceptable and unacceptable practices by qualified members of a
legitimate discipline (Gross & Mnookin, 2003). We are somewhat discouraged by the fact that many of our colleagues
continue to utilize indisputably unreliable assessment techniques. Harm comes to families when, even under
cross-examination, methodological deficiencies are not brought to light, untrustworthy data are not identified as such, and the
opinions based on these flawed data are utilized in constructing parenting plans.

Courts do not generally have difficulties with expert testimony that sticks close to observable facts, or Level I in the Tippins
and Wittmann model. It is Level IV opinions that cause the court--or should cause the court--great concern. Often, Level IV
opinions do *255 not raise concerns for judges because, [m]ore often, courts admit unreliable generalizations because they
are accustomed to evidence of that sort ... (Gross & Mnookin, 2003, p. 174). That is, specific custody recommendations are
often allowed into testimony because that is what is done within the field or within the jurisdiction. Opinions that are
expressed with a high level of confidence, and that outline specific custodial arrangements for particular children, are not
grounded in science. We concur with Tippins and Wittmann in their call for professional humility. The problem is with
labeling and delivery, rather than with substance. An opinion formulated on the basis of professional experience may be of
assistance to the court, but it is deceptive to offer such an opinion as though it had empirical support in the published
literature.

PROS AND CONS OF OFFERING ULTIMATE ISSUE TESTIMONY
We find many compelling elements in the Tippins and Wittmann position about the limitations of mental health
professionals providing specific recommendations to the court. In offering recommendations to the courts, custody evaluators
should take great care to include explanations of the limitations inherent in those recommendations. Too often, custody
evaluators do not remind consumers of their advisory reports of the methodological shortcomings of our research and the
ways in which the research deficiencies limit the certainty with which we can delineate the advantages of specific custodial

13 cv 1944 CAB BLM

arrangements or the disadvantages of others.

We find useful Myers (1991) distinction between the ultimate legal issue and the ultimate factual issue. The ultimate legal
issue refers to the matter about which the trier of fact must reach a decision. In the context of child custody work, an ultimate
legal issue might be who is awarded custody of the minor children. As indicated above, mental health professionals are
limited in their testimony to the best psychological interests of the child and not to the best interests of the child. The latter
concept includes psychological factors but also includes factors not within a psychologists sphere of expertise and, therefore,
not to be considered in a psychologists evaluation. The ultimate legal issue is based both on the law, that is outside the
competence of most mental health professionals, and on common-sense judgments, that are properly reserved for the judge
(Melton & Limber, 1991). As Tippins and Wittmann state: The best interests standard is a legal and socio-moral construct,
not a psychological construct (p. 215).

An opinion on an ultimate factual issue might rest upon a defensible foundation in the published empirical research. As
Myers (1991a) points out,
Many forms of expert testimony by psychologists are not opinions on ultimate issues at all. Although an expert opinion that a
particular child was sexually abused may well be an opinion on an ultimate factual issue, the rules of evidence in force in
most states expressly permit experts to offer opinions on ultimate factual issues (Federal Rules of Evidence, Rule 704) ... (p.
82).

In the context of child custody work, evaluators are permitted to offer opinions about ultimate factual issues that have been
the focus of the investigation, when the evaluators have adequate informational foundations to support their opinions.

We agree with much of Tippins and Wittmanns argument about the need to limit the offering of ultimate opinion testimony
in child custody work. As Melton and Limber (1991) point out,
*256 Ultimate-issue opinions by mental health professionals do not assist the fact-finder, and they constitute legal opinions
by definition outside the specialized knowledge of mental health professionals. Therefore, they do not meet the standard for
admissibility of expert opinions under Rule 702 (Federal Rules of Evidence) .... As a matter of ethics, experts should not
offer opinions as if they were based on specialized knowledge when they are not (p. 83).

On the other hand, we believe that most judges want a specific recommendation as to custodial placement, and will avoid
using an evaluator who refuses to offer such an opinion. In many jurisdictions, the failure to offer such an opinion would
throw a monkey wrench into the traditional method of designating responsibility for experts fees. In many jurisdictions, by
rules of the court, the party wishing to introduce the experts report into evidence must pay the experts court fees. In some
jurisdictions, rules of the court dictate that a report prepared by a court-appointed expert will be introduced into evidence and
that the party wishing to contest the experts findings must pay the experts fees in order to have her available for purposes of
cross-examination. (See State v. Kim, 1982).

Wisely, the APA assisted psychologists in defining the scope and limitations of expert testimony. APAs use of the term
psychological, when used in conjunction with best interests, clearly circumscribes the area of our expertise. We can
provide expert testimony about ultimate factual issues and we can provide testimony about the best psychological interests of
the child, provided the opinion has an adequate basis in data, which supports it.

EMPIRICISTS VERSUS ROMANTICS
We agree with Tippins and Wittmann that psychologists have an ethical responsibility to our profession and a moral
obligation to those whom we serve to use the best informed scientific research to guide our choice of assessment tools and to
guide our recommendations about parenting plans. Elsewhere (Martindale & Gould, 2004), in defining the forensic model,
we noted that particular attention is given to the established reliability and validity of instruments under consideration (p.
3). We have an ethical responsibility to limit our recommendations to those for which there is empirical support. When there
are no research data on an issue, psychologists are ethically bound to explain that there are no data to support an expressed
opinion. We agree with Tippins and Wittmann that it is unethical for mental health professionals testifying as experts to offer
opinions based merely upon clinical hunches.

13 cv 1944 CAB BLM

There is a debate among some child custody evaluators about the role of clinical judgment in child custody evaluations. In
our opinion, clinical judgment is involved in most steps of the custody evaluation process. Custody evaluators use clinical
judgment in deciding what follow-up questions to ask a parent. They use clinical judgment in deciding what assessment
battery will provide relevant and reliable data. Evaluators use clinical judgment in deciding what to memorialize in note
taking.

Calloway (2002) has opined that [w]hen forensic rigor is overly applied ... clinical intuition, hunches, expertise, methods,
and techniques are castigated as inherently inferior (p. 216). The Tippins and Wittmann analysis clarifies the relationship
between level of abstraction from observational data and level of testimonial certainty. Use of their model may assist judges
in deciding if clinical hunches should be admissible and, if so, what weight should be assigned to them.

We urge forensic psychologists to maintain an active interest in current relevant literature. We are concerned that the
sentiment expressed by David Barlow (1981), more than twenty *257 years ago, about clinical practice may be applicable to
forensic practice today--that research has little or no influence on practice.

The transition from a clinical scientist-practitioner model to a forensic model need not initiate another debate about the
relative contribution of actuarial versus scientifically informed methods and clinical judgment methods. Yet, despite
empirical studies showing that certain assessment tools lack reliability and validity, some evaluators insist that, in the right
hands (theirs), these tools can yield data that are sufficiently sound to form the basis for expert opinion testimony. The debate
should easily be resolved by looking at federal and state rules of evidence that guide the admissibility of information and
opinions contained in custody advisory reports. Specifically, such information and opinions must be based on relevant and
reliable methods and procedures.

The continued support by some custody evaluators for basing opinions to the court on clinical judgment and hunches should
be particularly distressing to a discipline (psychology) and a specialty area (forensic psychology) whose goals over the last
several decades have been to produce professionals who would integrate the methods of science into forensic practice, to
produce new knowledge, and to use already existing knowledge in a manner that scientifically informs assessment methods
and scientifically informs psychological opinions presented to the courts.

Tippins and Wittman are not the first to publicly register some strong criticisms of the procedures employed by custody
evaluators or the opinions expressed by them. In 1987, in their much-cited work Psychological Evaluations for the Courts,
Melton and his colleagues (Melton, Petrila, Poythress, & Slobogin, 1987), commenting on our assessments of comparative
custodial fitness, declared: [T]here is probably no forensic question on which overreaching by mental health professionals
has been so common and so egregious (p. 330).

The Honorable Stephen Hjelt (2000) has opined that the mental health disciplines, more than any others, are relied upon by
the courts. When the advisory input provided by mental health professionals is flawed and when courts rely upon such input,
the results can be regrettable. Our justice systems hunger for guidance on mental health issues can be clearly seen if we look
beyond custody disputes to a death penalty case--Barefoot v. Estelle (1983)--in which a prediction of future dangerousness
was a key element. When informed by the American Psychiatric Association (in an amicus brief) that predictions of future
dangerousness were unreliable, the United States Supreme Court disregarded the Associations cautionary advice, declaring:
Neither petitioner nor the Association suggests that psychiatrists are always wrong with respect to future dangerousness,
only most of the time (Barefoot v. Estelle, 1983).

The Bow and Quinnell (2001) survey data, cited by Tippins and Wittmann, suggest that the vast majority of custody
evaluators are offering opinions that lack empirical support in the published literature. It seems reasonable to hypothesize that
most custody evaluators believe that no change is needed, either in our methods or in the manner in which we communicate
our findings and opinions to courts. We disagree with those who see no problem and concur in Tippins and Wittmanns view
that the elephant in the room must be discussed ... if not shot.

Certain intractable human dynamics and powerful economic forces must be considered as decisions are made concerning
how best to address the problems alluded to by Tippins and Wittmann. First, when judges seek answers, they are likely to be
more receptive to hearing from the expert who proclaims s/he has the answers than from the expert who opines that the

13 cv 1944 CAB BLM

answers are elusive. Particularly where the issues are complex, the expert *258 who offers facile solutions is perceived as
being helpful. The expert who declines to speculate may be perceived as uncooperative. Many practitioners, appreciating the
cogency of the Tippins and Wittmann perspective, may contemplate a change of course with apprehension. It may be felt by
some that if responsible practitioners refuse to offer the answers that are sought, less qualified practitioners will quickly fill
the void and no benefit will be derived. As is true elsewhere in our society, the services offered by mental health
professionals and the methods employed in providing those services are, to a large extent, consumer driven. Tippins and
Wittmann have called for clinical humility and judicial vigilance. At the risk of appearing immature by stating, You go
first, it seems to us that when judges and attorneys become more vigilant, clinical humility from mental health professionals
will quickly follow. If judges and attorneys (as the consumers of forensic mental health services) encourage evaluators to
ignore the limits of an empirically established specialized knowledge base, only the atypically conscientious will resist the
economic incentive to provide that which has been sought.

In our view, there are three possible positions that evaluators, as a group, can take: (1) We can conclude that in order for
evaluators to be of meaningful (as opposed to illusory) assistance to triers of fact, all evidence offered by us must be
scientifically derived; that evaluators who offer evidence relating to custody matters are, therefore, doing a disservice to the
judicial system, to parents and children, and to the mental health professions, and, that we should stop offering this service.
(2) We can encourage those who interact with us in custody-related matters to play the dont ask, dont tell game; that is,
cross-examining attorneys will not inquire about the empirical bases for our opinions and we wont volunteer information
concerning the known deficiencies in our current procedures. This tactic is in violation of our obligation to inform others of
the known limitations of our procedures. (3) We can be diligent in acknowledging the known limitations of our procedures
and dedicate ourselves to improving them.

Knowledge is derived through inquiry, and professional activity stimulates inquiry. If mental health professionals stop
performing custody evaluations, the interactions that catalyze empirical research will not occur and no progress will be seen.
If we are to provide the assistance that the legal system seeks, we must persevere, but as we do so, we must be mindful of our
limitations and must respond to the call for clinical humility by articulating those limitations.

REFERENCES
Barefoot v. Estelle, 463 U.S. 880 (1983).

Barlow, D. H. (1981). On the relation of clinical research to clinical practice: Current issues, new directions. Journal of
Consulting & Clinical Psychology, 49, 147-155.

Bow, J . N., & Quinnell, F. A. (2001), Psychologists current practices and procedures in child custody evaluations: Five years
after the American Psychological Association guidelines. Professional Psychology: Research & Practice, 32(3), 261-268.

Calloway, G. (2002). When forensic rigor is overly applied. In Proceedings of the fifth international symposium on child
custody evaluations and the fifth international congress on parent education and access programs (pp. 213-219). Madison,
WI.: Association of Family & Conciliation Courts.

Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579 (1993).

Gould, J . W. (1998). Conducting scientifically crafted child custody evaluations. Thousand Oaks, CA: Sage.

Gross, S. R., & Mnookin, J . L. (2003). Expert information and expert evidence: A preliminary taxonomy. Seton Hall Law
Review, 34, 141-189.

Hjelt, S. (2000). Professional psychology: A view from the bench. Register Report, 26(1), 8-13.

*259 LaFortune, K. A., & Carpenter, B. N. (1998). Custody evaluations: A survey of mental health professionals. Behavioral
Sciences & the Law, 16, 207-224.

Martindale, D. A. (2001). Cross-examining mental health experts in child custody litigation. Journal of Psychiatry and Law,

13 cv 1944 CAB BLM

29(4), 483-511.

Martindale, D. A., & Gould, J . W. (2004). The forensic model: Ethics and scientific methodology applied to custody
evaluations. Journal of Child Custody, 1(2), 1-22.

Melton, G. B., & Limber, S. (1989). Psychologists involvement in cases of child maltreatment: Limits of role and expertise.
American Psychologist, 44, 1225-1233.

Melton, G. B., & Limber, S. (1991). Caution in child maltreatment cases. American Psychologist, 46, 82-84.

Melton, G. B., Petrila, J ., Poythress, M. G., & Slobin, C. (1987). Psychological evaluations for the courts: A handbook for
mental health professionals and lawyers. New York: Guilford Press.

Myers, J . E. B. (1991a). Comment on Melton and Limber. American Psychologist, 46, 81-82.

Myers, J . E. B. (1991b). Evidence in child abuse and neglect cases (3rd ed.). New York: Aspen Publishers.

Otto, R. K., Buffington-Vollum, J . K., & Edens, J . F. (2003). Child custody evaluations. In A. M. Goldstein (Ed.), Handbook
of psychology: Volume 11: Forensic psychology (pp. 179-208). New York: Wiley.

State v. Kim, 645 P.2d. 1330 (Haw. 1982).


Footnotes

a1

Jonathan W. Gould, Ph.D., is a principle in the national consulting firm of Child Custody Consultants. He practices forensic
psychology, focusing on attorney consultations, review and critique services, litigation support, and pretrial consultation. He is
author of the book, Conducting Scientifically Crafted Child Custody Evaluations (2nd ed.), published by Professional Resource
Press and author or coauthor of more than thirty articles about child custody related activities.

43 FAMCR 253
End of Document



13 cv 1944 CAB BLM
COMMENTARY ON EMPIRICAL AND ETHICAL PROBLEMS..., 43 Fam. Ct. Rev. 223

Family Court Review
April, 2005
COMMENTARY ON EMPIRICAL AND ETHICAL PROBLEMS WITH CUSTODY RECOMMENDATIONS:
WHAT NOW?
Thomas Grisso
a1

Copyright 2005 by Association of Family and Conciliation Courts; Thomas Grisso
Replying to the paper by Tippins and Wittmann, this commentary notes that the problems they identify have been recognized
for many years, yet this has resulted in little change in the practice of child custody evaluations. Three underlying reasons are
offered for the stalemate that frustrates the implementation of standards for an empirically based child custody evaluation
practice: (a) the economics of child custody evaluation practice; (b) inconsistencies between proposals to restrict testimony in
this area and the lack of similar restrictions in most other areas of forensic practice; and (c) inadequate motivation for
researchers who might contribute an empirical base for child custody evaluations. Directions for breaking the stalemate are
offered for each of these problems.

Keywords: child custody; evaluations; practice standards

In their article on issues in expert custody evaluations and recommendations, Tippins and Wittmann (2005) (a) provide a
helpful structure for thinking about levels of inference in offering information to the courts, (b) discuss serious limits to what
clinicians can offer with scientific integrity at the higher levels of inference, and (c) call for several policies and actions that
would restrict the boundaries of expert opinion practice in child custody. They recommend abolishing the practice of asking
mental health experts to offer opinions about specific custody arrangements.

By and large their arguments are well made and their recommendations, while I do not agree with all of them, are consistent
with the concerns that they raise. Others have raised similar points over the past three decades. But that does not detract from
Tippins and Wittmanns restructured statement of the issues, because experts practice in child custody evaluations
apparently has not responded to earlier appeals.

Yet repeated haranguing of practitioners has only limited value, especially when its tone widens the gap between
practitioners and their academic critics, thus reducing their collaborative potential to solve the problems. The fields
resistance to improvement suggests that there are underlying issues that are impeding progress toward improvements in
practice and that need attention before proposals for reform can make a difference. This commentary cannot offer a thorough
analysis of those dynamics, but it can at least identify some matters that need attention in order to break the stalemate that
seems to have gripped this field for some time.

A HISTORICAL PERSPECTIVE
The complaint that clinicians often exceed their scientific basis for child custody testimony was first heard about thirty years
ago (Otto & Edens, 2002) in the wake of Goldstein, Freud, and Solnits (1973) publication, Beyond the Best Interests of the
Child. Critics focused on the lack of evidence for the validity of clinicians predictions concerning custody arrangements.
Some recommended solutions more sweeping than those of Tippins and Wittmann, including banning clinicians from child
custody cases altogether. Nothing happened.

*224 A second wave of cautionary statements arose in the 1980s with the publication of the first set of comprehensive
textbooks on forensic psychological evaluations (e.g., Grisso, 1986; Melton, Petrila, Poythress, & Slobogin, 1987; also, on
child custody evaluations specifically: Schutz, Dixon, Lindenberger, & Ruther, 1989). Typically they recommended that

13 cv 1944 CAB BLM

clinicians should not offer an opinion on the ultimate question of placement of the child. This time something did happen. In
the early 1990s, national professional organizations published the first practice guidelines for custody evaluations. Finally,
Tippins and Wittmann described a third phase of criticisms during the 1990s, especially noting that the various professional
guidelines, while well intentioned, were neither bold enough nor translated into practice in ways that resolved the problems.

Arguments have differed little across those decades. For example, returning to an article I wrote for this journal fifteen years
ago on Evolving Guidelines for Divorce/Custody Evaluations (Grisso, 1990), I rediscovered my effort to describe what
clinicians could and could not do in custody evaluations within the limits of a scientific practice:
... we can describe parents abilities ... we can accurately explain their strengths and deficits ... we can
often estimate the risks of various future placements ... But [that] is not the same as testifying that those
negative consequences will occur ... that type of predictive opinion rarely has any support from our
profession or our science (p. 38).

Tippins and Wittmann describe their levels of clinical inference better and more completely, but my four points would fit
fairly well into their four-level hierarchical scheme.

So, after thirty years of calls for reform, one must believe that it will take something more than arguments, harangues, and
published guidelines--even the best of them--to bring about a change in practice. The real world will not yield to logic until
some practical dilemmas are faced and resolved. Let me briefly describe some of them, focusing especially on
recommendations at the end of Tippins and Wittmanns article.

PRACTICE GUIDELINES AND THE DILEMMA OF COMPETITION
One obstacle to change mentioned by Tippins and Wittmann is the financial gain associated with expert services to clients.
The client wants the most definitive and far-reaching argument that an expert can make, and obeying cautions to remain at
the lower levels of inference inhibits the experts ability to provide what the client wants. Thus there is significant economic
pressure to venture into the higher levels of inference.

Similarly, when working in the interests of its practitioners, organizations like the American Psychological Association face a
dilemma when they develop practice guidelines. Restrictions focused on the narrower boundaries of empirically based
practice are good for practitioners in that they preserve the integrity of the profession, thereby strengthening its credibility.
Moreover, constructing ethical sanctions (rather than mere voluntary guidelines) can produce a level playing field within the
profession, in that all must abide by the rules rather than allowing some to seek an advantage by catering to the client. On the
other hand, ethical sanctions that restrict practice run the risk of impairing psychologists livelihoods, because organizations
that represent psychiatrists and social workers--who also provide child custody expertise--may have less restrictive ethical
standards. Thus, a professions formal restrictions that attempt to maintain the integrity of its practitioners can end up putting
them out of business.

*225 As a consequence, perhaps guidelines cannot have a serious impact on practice until they can be developed through
interdisciplinary collaboration. This is not an encouraging thought, because the primary national organizations for
psychologists, psychiatrists, and social workers do not have a history of collaboration on anything. However, the forensic
specialty groups for psychiatry (American Academy of Psychiatry and Law) and psychology (American Psychology-Law
Society) have a more congenial posture toward each other. Unfortunately, these two organizations have not focused
sufficiently on child custody issues in the past, so that many child custody examiners do not consider these organizations
their home. The Association of Family and Conciliation Courts fits that designation better. Therefore, the development of a
three-way relation between these groups aimed at joint standards might begin to resolve the competitive dilemma for various
professions providing custody evaluation services.

BANISHING CONCLUSORY TESTIMONY AND THE DILEMMA OF LEGAL INCONSISTENCY
Looking to the law to solve the problem of conclusory (fourth-level) testimony is not likely to work for two reasons. First,
there is a fair amount of evidence that judges favor the availability of expert opinions about the ultimate legal question

13 cv 1944 CAB BLM

(which requires a fourth-level inference; for a discussion, see Chapter 18 in Melton et al., 1997). This need not always be so.
Many of us know judges who will not allow experts to testify to the ultimate legal question. Yet, the weight of judicial
opinion on average appears not to produce fertile ground for sowing the seeds of reform that would prohibit fourth-level
expert testimony.

The second reason that legal circumstances are not favorable for banishing level-four testimony in custody cases is its general
acceptance across almost all other areas of law in which psychological testimony is involved. For example, these include
fourth-level opinions on criminal responsibility, competence to stand trial, sex offender recidivism, and a host of other
forensic issues for which opinions about the bottom line are expected and admitted as evidence in most jurisdictions. To
prohibit it in one area of law-- child custody cases--but to allow it in others would require some compelling argument that
Level IV opinions in child custody cases have far less empirical support (and thus less probative value or more prejudicial
impact) than in these other areas of forensic expertise.

But this is not the case. Tippins and Wittmanns statement that there is not one piece of research ... that supports the notion
that clinicians can reliably and validly ... [make] recommendations that are correlated with positive child adjustment
(Tippins & Wittmann, 2005) could be repeated for many other areas of forensic evaluation. For example, there is also not
one piece of research that supports the notion that clinicians can reliably and validly determine whether defendants could
appreciate the consequences of their acts or conform their behaviors to the requirements of law at the time of their
offenses, which are the Level IV inferences to which forensic psychologists and psychiatrists routinely testify in criminal
cases involving the insanity defense. J ust as there has been no groundswell of judicial scrutiny regarding the reliability and
validity of Level IV inferences in child custody cases, there also has been no serious legal challenge regarding their testimony
about the bottom line in many other types of forensic cases.

The message here is that the field of child custody evaluations cannot go forward alone to resolve its legal and ethical
dilemmas. What the law will accept or disallow in this area may depend on legal precedent in the use of forensic
psychological experts in other areas *226 of law. Perhaps the battle cannot be won without strategic collaboration with the
broader fields of forensic psychology, psychiatry, or social work.

REQUIRING FORENSIC SPECIALIZATION AND THE TRAINING DILEMMA
Tippins and Wittmanns (2005) suggestion that only clinicians with adequate forensic training should be allowed to
function as expert evaluators in child custody cases is good in principle but difficult to imagine from a practical perspective.
What does adequate forensic training mean, and where are the forensic training centers that will provide the nation with the
large number of child custody examiners that are needed to assist the courts in custody deliberations? What emerging
economic or ideological motivation will cause academic programs across the country to develop new child custody forensic
training programs? If this does not happen, then banning all but forensically trained examiners will leave the courts in many
jurisdictions without even the first-, second-, and allowable third-level inferences that are important for helping them resolve
custody cases in ways that further the welfare of children. Certainly, the training programs need to be developed. But their
positive impact will be very long-range and does not warrant a current rule that bars practice to all but those who receive that
training.

EMPIRICAL PRACTICE AND THE RESEARCH DILEMMA
Tippins and Wittmanns call for more research is undeniably important, but it echoes cries that have been heard for three
decades. There has been some response, but why has it been relatively meager? Only academics can reasonably be expected
to produce the research that we need, and they seem to have been intent on using their position in academic towers more as a
safe place for hurling stones at the empirically starved practitioners below than for producing the scientific data that would
nourish them.

Part of the problem is funding. It is difficult to find federal or foundation grant agencies that have child-custody-related
research among their main priorities. In addition, it may be that few researchers care about (or are aware of) the kinds of
research questions that need to be studied. Changing those two conditions presents an enormous challenge, requiring some
sort of coalition of scientists and practitioners, or the efforts of individuals who are able to play both roles. The potential for

13 cv 1944 CAB BLM

such coalitions is found almost exclusively in clinical academic settings, where research is encouraged and where there are
clinicians to generate the relevant research questions.

The field of child custody practice should not imagine that this is impossible, because other subfields of forensic psychology
and psychiatry have successfully forged such coalitions. In recent years, research academicians in the field of law and
psychology--most of whom do no forensic clinical practice--have produced an enormous quantity of research designed
specifically to improve their practitioner colleagues forensic evaluations for competence to stand trial, predictions of
dangerousness, and a number of evaluations pertaining to children in court (e.g., children as witnesses, adolescents
capacities to waive rights). I do not know specifically what it would take to create academic conditions that would produce
similar benefits for child custody examiners. All that is clear is that scientist/practitioner child custody examiners must seek
opportunities to stimulate their research colleagues, and academicians who claim that general practice in child custody
evaluations is bad must begin to look for (perhaps) the minority of examiners who value science as potential colleagues to
further their research.

*227 This is not to say that such efforts, even if successful, will produce research that will resolve the problems of testimony
that Tippins and Wittmann describe. As they point out, many commentators believe that testimony on the ultimate legal
question in any forensic evaluation should be prohibited not merely because of questions about empirical evidence, but
because all ultimate legal questions require a moral judgment that no empirical data can address (Grisso, 2003; Melton et al.,
1997; Tillbrook, Mumley, & Grisso, 2003).

Moreover, anticipating data that will resolve problems of complex Level III and Level IV inferences may be like waiting for
the discovery of the Holy Grail. Tippins and Wittmanns appeal for research to discover the multivariate combinations of
individual, family, and access plan variables may be more than a daunting task; it may be impossible. For example, thirty
years of intensive scientific research on the prediction of violence--involving dozens of multisite projects of the highest
quality and costing several hundred millions of dollars--has finally achieved a level of actuarial certainty that is as high as it
is ever likely to become. Clinicians using methods derived from these studies can now estimate the degree of risk of violence
fairly accurately, but the methods only rarely allow them to conclude that anyone is more likely than not to be violent. If
we are waiting for the formulas that will tell us-- across the wide variety of people and circumstances involved in divorce--
whether a child is more likely than not to be harmed by a particular access arrangement then we can give up now, or we
can reconsider our objectives.

Presuming that we do not want to give up on the possibility of a scientific practice, adjusting our objectives is likely to
require a more balanced perspective regarding the value of theory in child custody practice. I refer to theory not as a matter of
intuition (as in I have a theory about this case), but as a scientific tool (e.g., attachment theory) involving a set of coherent
principles and constructs for making sense of certain psychological and social phenomena. The foundation of science is
theory, not empirical prediction. The value of empirical research is to test a theorys ability to produce hypotheses that make
sense of what we see around us. A theory is strengthened by multiple empirical validations of hypotheses that it has
generated, which in turn allows us to use that theory to guide our judgment in clinical work. Science depends on theory
development, because no amount of empirical research can ever test all of the relationships that arise in complex physical and
social processes. In this perspective, ethical practice does not require empirical proof that a particular child will not thrive if
left in the custody of a particular parent. It requires a sound, empirically validated theory to explain the logic on which ones
opinion is based.

This perspective, of course, does not change the need for empirical research; it takes a lot of it to refine and validate a
scientific theory. As Tippins and Wittmann urge, clinicians still must bind themselves to the scientific method. Using a
theory is far different, and far more demanding, than using clinical intuition. But, it allows us to proceed with something
while we wait for better empirical verification of our opinions, perhaps allowing us to move further into the third level of
inference than Tippins and Wittmann would suggest while still remaining true to our scientific principles.

REFERENCES
Goldstein, J ., Freud, A., & Solnit, A. (1973). Beyond the best interests of the child. New York: Free Press.

Grisso, T. (1986). Evaluating competencies: Forensic assessments and instruments. New York: Plenum Press.

13 cv 1944 CAB BLM

Grisso, T. (1990). Evolving guidelines for divorce/custody evaluations. Family & Conciliation Courts Review, 28, 35-41.

Grisso, T. (2003). Evaluating competencies: Forensic assessments and instruments (2nd ed.). New York: Kluwer
Academic/Plenum Press.

*228 Melton, G. B., Petrila, J ., Poythress, N. G., & Slobogin, C. (1987). Psychological evaluations for the courts. New York:
Guilford Press.

Otto, R., & Edens, J . (2002). Parenting capacity. In T. Grisso, R. K. Otto, R. Borum, J . F. Edens, & J . Moye, Evaluating
competencies: Forensic assessments and instruments (2nd ed., pp. 229-307). New York: Kluwer Academic/Plenum Press.

Schutz, B. M., Dixon, E., Lindenherger, J . C., & Ruther, N. J . (1989). Solomons sword: A practical guide to conducting
child custody evaluations. San Francisco: J ossey-Bass.

Tillbrook, C., Mumley, D., & Grisso, T. (2003). Avoiding testimony on the ultimate legal question. Journal of Forensic
Psychology Practice, 3, 77-78.


Footnotes

a1

Thomas Grisso, Ph.D., a clinical psychologist, is Professor of Psychiatry at the University of Massachusetts Medical School,
where he is Director of Psychology and Coordinator of the Law and Psychiatry Program. His research, teaching, and practice
focus on the improvement of mental health and forensic evaluations and services for adults and children in the criminal and
juvenile justice systems.

43 FAMCR 223
End of Document



13 cv 1944 CAB BLM
MAKING A PLACE AT THE TABLE: RECONCEPTUALIZING..., 43 Fam. Ct. Rev. 229

Family Court Review
April, 2005
MAKING A PLACE AT THE TABLE: RECONCEPTUALIZING THE ROLE OF THE CUSTODY EVALUATOR IN
CHILD CUSTODY DISPUTES
Mary Kay Kisthardt
1
Barbara Glesner Fines
2

Copyright 2005 by Association of Family and Conciliation Courts; Mary Kay Kisthardt, Barbara Glesner Fines
This response to Timothy Tippins and J effrey Wittmanns article Empirical and Ethical Problems with Custody
Recommendations: A Call for Clinical Humility and J udicial Vigilance builds upon the authors conclusions that custody
evaluations cannot and should not be a substitute for the socio-legal judgment of the best interests of the child. Recognizing
that clinical humility and judicial vigilance may not be sufficient to restrain the misuse of psychological evaluation, we offer
for consideration three structural changes that would provide for a more appropriate use of the skills and talents custody
evaluators bring to legal decisions. We suggest using custody evaluators in the less adversarial setting of preparing parenting
plans, revising the procedures by which custody evaluations are elicited in litigation, and adopting the approximation
standard for child custody determinations.

Tippins and Wittmann provide much food for thought about why we have come to the present situation and what we can do
about it. In the first instance, they make a very persuasive case for the exclusion of forensic psychological assessments on the
ultimate questions presented in disputed child custody cases. Their conclusions appear to be based on two general theories:
first, there is simply no reliable scientific basis for making such determinations, and second, the best interest standard
currently used by the courts is not a psychological construct but rather a legal and socio-moral one. As to the first point, they
make a compelling case for the lack of empirical research sufficient to make these judgments and we are not prepared to
dispute those findings. Rather, as lawyers and law professors, it is the second basis for their conclusion that raises more
interesting questions.

To describe the best interest standard as both a legal and socio-moral concept foreshadows the reason why it has proved to
be so difficult to apply. Unlike more determinative standards, the best interest standard invites the most subjective of
analyses. It might be suggested that judges are in no better position to determine the best interest of the child than are
psychologists. Their unease in doing so has, no doubt, led to the excessive reliance on psychological assessments reported by
the authors. The authors exhort custody evaluators to be restrained in their testimony while encouraging judges to take
greater care in evaluating the admissibility of this testimony. While important short-term considerations, these cautions are
unlikely to significantly improve the use of mental health professionals in crafting custody decisions. Moreover, the
adversarial pressure on attorneys, particularly in high-conflict divorces, is one for which there exists few checks. While the
rules of professional conduct have substituted standards of competence and diligence for the older duty of zealous
representation, the adversary system remains structurally designed for opposition, not cooperation. Attorneys have a duty to
present the clients case with persuasive force.
3
In doing so, attorneys are unlikely to be able to effectively evaluate when
an expert has moved from Level II testimony to Level III testimony because they have every incentive, and even a duty, to
move the expert in a direction that will provide the most persuasive *230 testimony possible. Courts as well, hoping to have
an objective grounding that will provide efficient disposition of their ever-growing family law dockets, have little incentive to
police the boundaries of expert testimony. If custody evaluators increasingly pull their punches by significantly hedging their
observations and refusing to provide opinions on the ultimate question in custody disputes, attorneys and judges alike may
conclude that they have little value in the adversarial contest.

Rather than writing off the entire profession, which brings important and useful knowledge and skills to the custody
determination, the legal system must find better ways to use those skills and talents. For that to happen, we believe more
substantial structural changes are necessary. We suggest three such changes: using custody evaluators in the less adversarial
setting of preparing parenting plans, revising the procedures by which custody evaluations are elicited in litigation, and

13 cv 1944 CAB BLM

adopting the approximation standard for child custody determinations.

One way to better use the special expertise of child custody evaluators would be to bring these professionals in at a much
earlier stage, before the parties have polarized their positions and the evaluator is viewed as choosing sides. Currently, when
custody evaluators are used, the rate of relitigation of the resulting stipulations or orders is dramatically high.
4
With the trend
toward requiring the use of parenting plans in all child custody cases comes the opportunity to engage child custody
evaluation experts to assist in crafting these plans. A parenting plan is a detailed description of the manner in which parents
intend to continue caring for their children after divorce. The development of these plans benefits children by requiring their
parents to focus on the specific needs of the children and to anticipate and address expected changes in their lives. Hopefully
this forethought, as well as an agreement on how to resolve disputes that arise, will reduce future conflict that is so
detrimental to childrens well-being. Making these plans mandatory requires parents to sit down and think very specifically
about what the childs life is going to look like after divorce. It does not permit parents to simply and very generally portray
the other parent as inadequate thereby placing themselves in the primary parenting position.

In this context, it would seem that psychologists might play a very valuable role. To begin with, we know that childrens
reaction to change and separation from a parent is based not just on individual differences but also on their developmental
stage. Psychologists have much to offer parents in terms of suggestions related to developmental processes. Furthermore, by
the very nature of its subject matter, a parenting plan will have to respond to change. Again, alerting parents of the need to be
flexible as their children age would be most useful in avoiding future conflicts.

Additionally, interviewing the child in order to ascertain his/her individual differences would allow a psychologist to work in
the best interest of that child by assisting the parents themselves in making good decisions. Rather than being put in a
position of having to assist (in a very direct way, if the authors are correct) a judge who relies so heavily on the report
because of a lack of knowledge of the child or the family, the psychologist would be in the position of assisting the
appropriate decision makers, the parents who are in a superior position to judge what is in their childs best interest.
Obviously, all the cautions the authors provide regarding restraint in providing opinions on ultimate plans apply here as well.
However, this is made easier by the absence of the adversarial pressure that may come later should the parties fail to reach an
agreement.

When adversarial litigation does result, there are structural reforms that the legal system might consider to improve the ability
of lawyers and judges to understand the limits of custody evaluations. There is an informational disconnect between lawyers
and psychologists about the work they do and the limits of their science. Mental health professionals are *231 often
surprised to learn that there is not a definitive answer to the question of what is the law in this situation?, in much the same
way that lawyers will be surprised to learn that a psychologist is not qualified to make a recommendation on what child
custody plan would be best for a child. The authors recommendation that there be a better effort made in cross-disciplinary
understanding is a welcome one indeed.

J udges and lawyers might better understand the limits of science if they heard from more than one custody evaluator. Many
commentators suggest that, to reduce cost and acrimony, courts should appoint custody evaluators, who can then provide
objective and neutral testimony.
5
However, the authors provide many examples of how evaluator testimony may be less
than objective-- even without the weight of adversarial bias. It may be that the practice of each side engaging a custody
evaluator may result in greater, rather than less, objectivity as the differences in their observational standpoints and
theoretical orientations may then be disclosed. The federal courts of Australia have adopted a number of innovations in using
multiple experts, including pretrial conferencing between experts to prepare a joint report and the so-called hot tub rule for
testimony. Seeking to provide expert testimony in dialogue between the professionals rather than in adversarial testimony,
experts are called as a panel, rather than individually, and are allowed to ask one another questions, rather than routing all
their testimony through the attorneys.
6

Obviously, increasing the number of custody evaluators or bringing them in at an earlier stage may increase the cost of
custody determinations. More fundamental structural reforms may be required before the skills and knowledge of mental
health professionals can effectively and ethically guide the custody decision. The difficulties identified by the authors and the
obvious reluctance of both sets of professionals to make these decisions should lead us to a closer examination of the broader
issues: questioning the advisability of the best interest standard as the overriding criteria for making these decisions and
determining an appropriate role for mental health professionals to play. Criticisms of the best interest standard are

13 cv 1944 CAB BLM

long-standing
7
and will no doubt be renewed by the authors recommendations. In response to these criticisms several
alternatives have been proposed including a change in the standard courts should use and the creation of parenting plans in all
child custody cases. If the appropriate use of psychologists could enhance either of these alternatives, they are likely to be
viewed more favorably.

For making child custody determinations in contested cases, the American Law Institute has proposed that the proportion of
custodial responsibility allocated each parent should approximate the proportion of caretaking functions each parent
exercised prior to the separation. This standard, known as the approximation standard, is designed to reduce litigation by
making the outcome of child custody cases more predictable. Several concerns raised by the authors may be addressed by the
adoption of this standard.

First, who are the families for whom child custody evaluations are currently being ordered? A vast majority of child custody
cases are settled by the parents, often with the assistance of their lawyers. Cases that proceed to the point of needing a third
party decision maker are generally disputes between fairly contentious parents.
8
This is not to say that there are not many
meritorious reasons why a parent would need to resort to litigation, but for many families the dispute is between fairly
equally good parents or fairly equal not so good parents who are for whatever reason intent on continuing the fight. For
these families the use of the best interest standard provides a fertile battleground. Proponents of the approximation standard
suggest that replacing the best interest standard with one that makes the outcome more predictable will reduce the
opportunity for legally sanctioned combat.

*232 Another rationale presented for the standard is that it removes much of the speculation about future conduct as a basis
for making a decision. This aspect is particularly attractive to those who argue that the adversarial system is not particularly
well adapted to that task. Most legal decisions are made by determining what has occurred in the past and assigning the
appropriate consequence to it. Predicting future human behavior is far more difficult than assessing what happened in the
past. Furthermore, proponents of the approximation standard argue that, with respect to parenting, past behavior is predictive
of future parent-child interactions, citing the Mnookin and Maccoby study which showed that over time, parents tend to
revert to their predivorce parenting patterns.
9

As proposed, the standard does have exceptions and this is perhaps where the psychologist might be most helpful. The most
relevant exception refers to a situation in which the presumptive allocation would harm the child because of a strong disparity
in the quality of the emotional attachment between each parent and the child, or in each parents demonstrated ability or
availability to meet the childs needs. It would seem that the psychologists are on much firmer ground here in evaluating not
the future but the demonstrated ability of the parent to meet the childs needs. This assessment could be done at a Level I or
II, thereby meeting our concerns. Furthermore, an additional exception permits accommodation of a firm and reasonable
preference of a child who has reached a specific age. It would seem that psychologists could also be of assistance in this
regard. Interviewing children is not generally a highly developed skill for lawyers or judges.

The custody decision is one of the most fundamental decisions courts are called upon to make. The need and desire to inform
(or even shift) that decision through the expertise of mental health professionals is understandable and perhaps even
unavoidable. We provide a great service in identifying the risks inherent in involving mental health professionals in that
decision beyond the boundaries of their expertise and ability. The challenge for law is to find appropriate, helpful, and
efficient methods to reduce family conflict and help parents craft appropriate custody arrangements. All professionals and
policy makers involved in these decisions must give much more careful consideration to the appropriate stage at which
mental health professionals can contribute, the optimum procedural structure for their participation, and the standards for
decision making that can be best informed by their expertise.

Footnotes

1

Associate Professor of Law, University of Missouri, Kansas City.

2

Professor of Law, University of Missouri, Kansas City.


13 cv 1944 CAB BLM

3

MODEL RULES OF PROFL CONDUCT R. 3.3 cmt. 2 (2002).

4

J anet R. J ohnston, Building Multidisciplinary Professional Partnerships with the Court on Behalf of High-Conflict Divorcing
Families and Their Children: Who Needs What Kind of Help?, 22 U. ARK. LITTLE ROCK L. REV. 453, 472-73 (2000).

5

Linda D. Elrod, Reforming the System to Protect Children in High Conflict Custody Cases, 28 WM. MITCHELL L. REV. 495
(2001).

6

J ustice Garry Downes AM, Concurrent Expert Evidence in the Administrative Appeals Tribunal: The New South Wales
Experience, Paper presented at the Australasian Conference of Planning and Environment Courts and Tribunals in Hobart February
27, 2004, available at http:// www.aat.gov.au/CorporatePublications/speeches/downes/concurrent.htm(last visited September 28,
2004).

7

Robert H. Mnookin, Child-Custody Adjudication: Judicial Functions in the Face of Indeterminacy, 39 L. & CONTEMP. PROBS.
226, 251-52 (1975).

8

Elrod, supra note 5.

9

Robert H. Mnookin & Eleanor Maccoby, Facing the Dilemmas of Child Custody, 10 VA. J . SOC. POLY & L. 54 (2002).

43 FAMCR 229
End of Document



13 cv 1944 CAB BLM

Family Court Review
April, 2005
COMMENTARY ON EMPIRICAL AND ETHICAL PROBLEMS WITH CUSTODY RECOMMENDATIONS: A
CALL FOR NEW FAMILY COURT PRIORITIES
Hon. Arline S. Rotman
a1

Copyright 2005 by Association of Family and Conciliation Courts; Hon. Arline S. Rotman
Whether or not custody evaluators, testifying as expert witnesses, continue to make specific custody recommendations,
custody decision making will continue to be based upon inadequate and untested evidence unless and until we reform the
family court system in American courtrooms. J udges and lawyers must have specialized knowledge and training about such
things as the developmental needs of children, the effects of divorce on children, domestic violence, and child safety issues;
lawyers must develop litigation as well as mediation and negotiation skills; specialized family courts utilizing individual
calendars and case management techniques must be established; and the complexities and intellectual challenges of family
law cases needs to be recognized.

Keywords: custody evaluations; specialized family courts; expert witness testimony; case management

It would be difficult for me to quarrel with Tippins and Wittmanns basic premise that custody evaluators should not be
allowed to testify on the ultimate issue, or with any of their recommendations. It is generally agreed within the legal
profession that custody evaluators wield too much influence in a contested custody case. Whether or not specific
recommendations regarding access plans should be included in their report is a subject of much debate, although it has long
been my opinion that such specificity is a clear invasion of judicial responsibility.
1

I would suggest, however, that Tippins and Wittmann do not go far enough in their call for reform. Even if we were to adopt
all eight of their recommendations, including barring Level IV testimony on the ultimate issue, we would still be faced with
the problem of custody decision making based on inadequate and untested evidence.

Without focusing on the limitations of scientific knowledge regarding the effects of various custodial arrangements on
childrens adjustment, I would like to suggest that the inordinate influence of custody evaluations and recommendations
result from systemic problems within the legal system.

Family law has long been at the bottom of the legal hierarchy, notwithstanding that more people interact with the judicial
system through family law cases than any other kind of legal action. Many reasons exist for this lack of regard.

ECONOMICS
Until the latter part of the twentieth century, there was relatively little of economic value to litigate in a divorce case other
than alimony. Community property states divided property according to law. Most states were title states with no judicial
authority to transfer property from one spouse to the other. Child support was minimal and available only for children born of
a marriage. Before no-fault divorce, matrimonial cases often consisted of trumped-up evidence presented to prove grounds
for divorce. Fathers were rarely granted custody unless a mother was proven unfit, and in fact, fathers rarely requested
custody. Thus, it *243 did not take great legal skill to represent a client in a divorce, nor did it require significant intellectual
acumen to adjudicate a divorce case. Although the nature of family law cases has changed and can involve significant,
complicated, and intellectually challenging issues, a significant portion of the organized bar still maintains a basic disdain for
divorce court.

13 cv 1944 CAB BLM

Litigation is costly. In most complex cases outside of family law, significant economic resources are available to fund a case.
In many cases, there are deep pockets that can be reached to provide significant recoveries. Lawyers are often paid from these
recoveries in contingent fee cases, or by corporations who are willing and able to fund litigation without a significant impact
on their profit. In other cases, insurance companies underwriting malpractice policies pay for the costs of litigation. In serious
criminal cases, indigent defendants are provided legal counsel funded by the government.

Notwithstanding the stakes in complex family litigation, there is no third party to fund litigation. The costs of preparing and
properly trying a domestic case are beyond the reach of most litigants, and therefore, cases can be inadequately prepared, and
often settled as a matter of economy.

SOCIOLOGY
Family law, and in particular, cases involving disputes over children, is a highly emotional arena. Lawyers are trained in legal
analysis, a relatively emotion-free, analytical exercise. They pride themselves on clarity of thought and expect to achieve a
clear-cut decision. The client is either guilty or innocent; the verdict is either for the plaintiff or the defendant. Family law
cases do not lend themselves to such black and white clarity. Lawyers working in this area deal with highly emotional clients,
who are fearful of potential economic dislocation, loss of their children, and death of their dreams. Except in rare cases,
litigants and lawyers alike understand that no one can honestly talk about justice being done or winners or losers. The best
that can be hoped for is a fair resolution as defined by the law, with no family member being badly hurt. Given the significant
social issues involved with domestic cases, the emotional overlay interfacing with the intellectually challenging issues, and
the lack of economic resources to fund the litigation, it is not surprising that many lawyers and judges alike do not choose to
work in this area. They consider that domestic law is something akin to social work and not real law.

RECOMMENDATIONS FOR CHANGE
Specialized family courts
Family law cases are complex and require a broad range of knowledge and expertise. J udges hearing custody cases must be
aware of such things as the dynamics of family systems, the developmental needs of children, the research relative to the
effects of divorce on children, the tensions between a childs need and right to a relationship with a parent, and the need to
keep that child safe when there are issues of sexual abuse, chemical dependency, domestic violence, or mental illness.

J udges who sit in courts of general jurisdiction should not be expected to hear a murder case one day, a product liability case
the next day, and a divorce case on the third day. Beyond the fact that more time is allowed on the docket for the first two
types of cases and generally more money is available to fund those cases, the knowledge base required to oversee and decide
custody cases will not ordinarily be acquired by the generalist judge. It is an interesting commentary on our social values that
specialized courts with specialized *244 judges are now being established for complex business litigation while the call for
such specialization for custody cases that involve the future of our children goes unheard.

Courts with jurisdiction over only family law cases are in a much better position to understand and manage the complexities
of custody cases. Dedicated family courts are able to implement individual calendaring, with each case assigned to a single
judge who will hear and manage the case from beginning to end. That judge is able to discern the nature of a custody contest,
and then make a referral with well-defined questions to a forensic custody evaluator tailored to the needs of that case. An
overbroad generic request for a custody evaluation to determine who should have custody or to recommend a parenting plan
encourages evaluators to utilize a generic format for all evaluations.

A more appropriate and useful reference to a custody evaluator should include a detailed and tailored request for relevant
information, with a specific referral question to be answered. For example, if there are allegations of sexual abuse, a custody
evaluator with specific expertise in sexual abuse should be chosen. The evaluator should be asked to describe factors
increasing or decreasing the probability that such abuse has taken place and the potential psychological effect on the child of
allowing supervised contact, unsupervised contact, or no contact with that parent. The evaluators report will then go back to

13 cv 1944 CAB BLM

the judge who ordered it and the case can be managed with hearings scheduled in an orderly and timely fashion.

It is the task of the judge who hears all the evidence in the case to integrate the evaluators opinion about the effect of contact
with the parent with all the other evidence in the case. If the report goes beyond the referral question and information
requested, it would be the attorneys task to file a motion in limine to redact the overreaching portion of the report. With case
management techniques in place, the family court would have the ability to schedule timely pretrial motions to deal with
evidentiary issues.

It should be noted that a legitimate and important function of expert witness testimony is to educate the judge hearing the
case. A careful explanation of the appropriate interview protocols, indicia of the reliability of a childs reporting or
subsequent recanting, and behaviors that could be consistent with sexual abuse will serve to educate judges for subsequent
cases with the same issue. Lawyers should take advantage of the cases where there are adequate resources to bring in expert
witnesses to educate the judges.

TRAINING FOR LAWYERS
Given the stakes involved for children and families in dissolution cases, it is neither surprising nor inappropriate that lawyers
invest their energies and skill development in resolving disputes through mediation, negotiation, or collaborative law.
Unfortunately, when cases cannot or should not be settled outside the courtroom, domestic relations lawyers too often lack
the necessary trial skills to challenge the testimony of the custody evaluator. It is this unchallenged testimony that allows for
their excessive influence.

Tippins and Wittmann decry the fact that judges allow expert testimony to be admitted that is not grounded in empirical
research and therefore does not meet the threshold of reliability and relevance. Unfortunately, there is little a judge can do
when evidence comes in unchallenged by a lawyer. If, for example, a psychologist administers a battery of psychological
tests not relevant to the issue of the case, that portion of the report and any opinions linked to the results of the test, should
not be admitted. For this to happen, lawyers must be educated about the appropriate use of each of the tests administered;
they then must learn the right way to challenge the evidence before it is admitted at trial. It is painful to preside over a case
where evidence that should be excluded is allowed in without objection. *245 It is even more painful to watch a lawyer
attempt to challenge evidence in a less than competent manner. Until lawyers master necessary trial skills, psychologists will
be free to opine on matters beyond their expertise. This mastery is made even more difficult when one realizes how few cases
ever go to trial.

If lawyers learned to effectively challenge the procedures, reports, or oral testimony of evaluators, then evaluators might be
more careful about what they put in their reports. Too many evaluators become accustomed to their testimony going
unchallenged. This puts an undue burden on the evaluator, who is often urged to make the ultimate custody recommendation.

JUDICIAL EDUCATION
J udges who elect to serve in dedicated family courts must master a whole new range of knowledge associated with the mental
health profession. In addition to becoming conversant with psychological knowledge and research about child development
and family systems, aberrant behavior such as uncontrolled anger, domestic violence, and substance abuse, and the diagnostic
criteria of mental illnesses set out in the Diagnostic Statistical Manual (DSM-IV) published by the American Psychiatric
Association, a family court judge must be an expert in courtroom and case management. Much of this education would work
best in an interdisciplinary setting such as provided by the Association of Family and Conciliation Courts.

In addition to providing detailed instructions to a custody evaluator in every order of appointment, there are other case
management techniques that help to rein in overexpansive expert testimony. For example, rather than allowing a blanket
qualification of expert witnesses prior to their substantive testimony, a judge may defer such qualification until the
examination is underway. It is difficult to foresee where a line of questioning might lead. Qualifying a witness as an expert in
child psychology runs the risk that the questioning may go well beyond the evaluators expertise. All child psychologists are
not necessarily experts in substance abuse or parenting implications of obsessive compulsive disorder. A better practice may
be to refrain from qualifying experts until they are asked a question that seems to be beyond their expertise and then rule on a

13 cv 1944 CAB BLM

timely, specific objection.

In short, if all professionals involved in family dissolution matters played their role well, evaluators as expert witnesses
would be only one cog in the wheel. Lawyers would be educated about the substance of psychological testimony, remain
up-to-date on significant research, and be skilled in litigation techniques. J udges would be family law specialists skilled in
case and courtroom management, and family courts would be given the resources to provide dispute resolution services to
litigants and adequate trial time to families that cannot resolve their disputes between themselves.

Footnotes

a1

Hon Arline S. Rotman (ret.) was an associate justice in the Massachusetts Probate and Family Court from 1988-2000. She is a
past president of the Association and Family and Conciliation Courts, a fellow of the American Academy of Matrimonial Lawyers,
and a member of the Family Law Sections of the Massachusetts and Boston Bar Associations. She currently provides Dispute
Resolution Services in all family law matters.

1

Specific recommendations, as I understand the term, involves setting out access plans detailing the number and identity of days
each week that a child should be with each parent, as well as holiday and vacation planning.

43 FAMCR 242
End of Document



13 cv 1944 CAB BLM
TO RECOMMEND OR NOT TO RECOMMEND: THAT IS..., 43 Fam. Ct. Rev. 246
Family Court Review
April, 2005
TO RECOMMEND OR NOT TO RECOMMEND: THAT IS NOT THE QUESTION
A Response to Tippins and Wittmann's Article Empirical and Ethical Problems with
Custody Recommendations: A Call for Clinical Humility and Judicial Vigilance
Lorraine Martin
a1
Copyright 2005 by Association of Family and Conciliation Courts; Lorraine Martin
Tippins and Wittmann provide a cogent argument for custody evaluators not to make recommendations to the court. From
their forensic and scientific perspectives, they have identified some important issues, which will certainly stimulate interesting
discussion among custody evaluators. In response to their article, it is my view that public sector custody evaluations offer a
philosophical and procedural alternative to forensic evaluations. This article proposes that recommendations should be viewed
as part of the process of evaluations rather than the outcome. Based on a qualitative and interpretive model, recommendations
are judged based on their applicability, transferability, and transparency. Recommendations viewed in qualitative terms provide
parents with opportunities to step out of litigation and provide guidance for parents' ongoing roles postseparation.
Keywords: child custody; custody evaluation; divorce; public services
INTRODUCTION
This article is in response to the Tippins and Wittmann (2005) article entitled Empirical and Ethical Problems with Custody
Recommendations: A Call for Clinical Humility and Judicial Vigilance. This article identifies some important issues and will
stimulate interesting discussion among custody evaluators. Clinicians and researchers in public services involved in custody
evaluations have generally remained peripheral to the debates regarding the efficacy of custody evaluations. It is my view
that public sector custody evaluations offer a philosophical and procedural alternative to forensic evaluations. In an approach
resembling qualitative research, recommendations can be evaluated based on applicability and transferability rather than
generalizability and predictability.
THE DEBATE
Debate regarding the efficacy of child custody evaluations and expert testimony has recently emerged in the field of custody
evaluations. Tippins and Wittmann have joined *247 with other forensic scholars (Karras & Berry, 1985; Melton, Petrila,
Poythress, & Slobogian, 1997) in suggesting that psychologists should avoid making recommendations to the courts in cases
of custody and access.
They suggest that recommendations and expert opinions are flawed by the lack of methodological validity and reliability in the
standardized instruments they use, by the lack of empirical foundation for making recommendations, and by the presence of
evidentiary and jurisprudential concerns related to custody recommendations. In summary, they argue that recommendations
in custody evaluations lack the generalizability and prediction needed to provide the court with a reliable opinion that is
both ethical and empirical in nature. These notions of generalizability and prediction are based on a scientific assumption
that human behavior can be statistically measured. Within this assumption, they argue that evaluators should not make specific
13 cv 1944 CAB BLM
recommendations to the court because recommendations are limited by the lack of the evaluator's capacity to engage in an
extremely complex multivariate prediction process.
In contrast, others have contended that psychologists should participate in child custody proceedings and should present
expert testimony concerning an appropriate child custody placement, because this psychological information can incrementally
increase the validity of judicial opinions (Ackerman & Ackerman, 1996; Clark, 1995).
PUBLIC SECTOR CUSTODY EVALUATIONS
An exhaustive review of the literature reveals that this debate is situated and limited to the field of psychology. Other mental
health professionals, such as social workers, routinely conduct custody evaluations for publicly funded services for the courts,
but discussions regarding these evaluations are limited (see Bala, 2004; Birnbaum & Radovanovic, 1999; Horvath, Logan, &
Walker, 2002). Hetherington, Stanley-Hagan, and Anderson (1989) note that 90% of custody matters are settled before the
parents even reach the courts. When parents are at an impasse, they look to the courts to help them resolve their differences.
Given that many families are unable to afford the cost of private custody evaluations, judicial systems employ mental health
professionals who provide public sector services to the court (Horvath et al., 2002) to help these families develop child-focused
parenting plans.
Publicly funded evaluations are generally fact based (Bala, 2004), less costly than a typical private evaluation (Horvath et al.,
2002), follow standards for conducting evaluations (e.g., see Association of Family and Conciliation Courts, 1994), and adhere
to ethical standards established by the evaluator's professional bodies (e.g., see Board of Registration for Social Workers in
British Columbia, 2002; Saskatchewan Association of Social Workers, 2001).
The Office of the Children's Lawyer is an example of a publicly funded service that delivers programs in the administration
of justice on behalf of children under the age of eighteen with respect to their personal and property rights. Lawyers within
the office represent children in various areas of law including child custody and access disputes, child protection proceedings,
estate matters, and civil litigation. Clinical investigators prepare reports for the court in custody and access proceedings and
may assist lawyers who are representing children in such matters. Legislation provides the basis for clinical investigators to
investigate and make recommendations to the court on all matters concerning custody of or access to the child.
*248 NATURALISTIC INQUIRY IN PUBLIC SECTOR CUSTODY EVALUATIONS
Bala (2004) states that there is no single accepted methodology for conducting assessments. Mental health professionals will
vary their methods and responses to the demands of each particular situation. Kelly (1997) points out that the best interest of
the child represents a willingness on the part of the court and the laws to consider the children on a case-by-case basis rather
than adjudicating children as a class or homogeneous grouping with identical needs.
This case-based focus of mental health professional practice has been historically and more recently connected to an emphasis
on the process of qualitative understanding of an individual's problem and crisis within a social context (Denzin, 2002; Sherman
& Reid, 1994). Denzin (2002) argues, for example, that within the field of social work, there has always been a desire to
capture past experiences from the perspective of those who have lived them. Social work views human behavior as relative to
circumstance instead of a more absolutist and objective view of human behavior (Vayda & Satterfield, 1997). This approach
emphasizes the importance of the interrelatedness and relationships of the family members involved in custody and access
disputes rather than focusing on the pathologies of each member.
ESTABLISHING CREDIBILITY AND RIGOR
13 cv 1944 CAB BLM
Tippins and Wittmann (2005) have highlighted that in Daubert v. Merrell Dow Pharmaceuticals (1993) the U.S. Supreme
Court established a new standard for assessing the admissibility of expert scientific testimony where the court directly assesses
whether the science has been subjected to the rigors of peer review and publication and whether the principles or methods have
gained general acceptance within the relevant scientific community.
The 1994 Supreme Court of Canada's decision in R. v. Mohan is the leading precedent on the law of expert evidence in Canada
(Bala, 2004). This decision states that the relevant criteria in considering whether to admit expert opinion are: relevance,
necessity in assisting the trier of fact, the absence of an exclusionary rule, and a properly qualified expert. Given the impact
of these decisions on expert testimony, child custody evaluators need to demonstrate that their methodology has credibility,
rigor, and relevance to the scientific community.
Qualitative approaches to scientific inquiry have, in the last twenty years, achieved status and visibility in the social sciences and
helping profession and can be found in peer journals in social work, nursing, counseling, family relations, health, community
services, education, and even medicine (Creswell, 1998).
Sound custody evaluations require a systematic and rigorous approach to the design and implementation of the investigation,
the collection of data, data analysis, and interpretation and reporting. Central to the quality of the evaluation is whether the
family members in the dispute have been authentically represented in the investigation and the interpretations made from the
information gathered are coherent in the sense that they fit the data and social context from which they were derived (Fossey,
Harvey, McDermott, & Davidson, 2002).
I contend that publicly funded evaluations generally use qualitative methodological procedures to establish credibility and
rigor that include: prolonged engagement, triangulation, peer debriefing and support, member checking, and auditing (Lincoln
& Guba, 1985). For example, credible data collection in public service evaluations includes the evaluator interviewing both
parents individually, conducting home observation visits with the parents and the children, interviewing the children in private
locations, conducting *249 additional interviews with the parents as needed, and contacting collateral sources (Bala, 2004).
To perform these tasks, the evaluator is involved with the family over time, usually between two to four months. This prolonged
engagement provides the evaluator with a rich description of the family situation.
Collateral sources can come from objective reports, such as physicians, school contacts, the police, and child welfare agencies or
from individuals who have been in the position to observe the parents and children, such as grandparents, babysitters, and new
partners (Austin, 2002). This procedure of triangulation, or the reliance on multiple sources of information, helps the evaluator
to reveal inconsistencies, fabrication, and inaccuracies in or confirm the primary data obtained from family members (Austin,
2002). Triangulation, therefore, yields clearer and deeper descriptions of the families involved in the evaluation and provides
credibility to the public sector evaluation.
Member checking, or verification of the evaluator's interpretations of the data collected during the evaluation process, is
performed at various times during the investigation, including at the end of the investigation to ensure the evaluator has
accurately interpreted the parents' positions. Member checking is also formally completed during a disclosure meeting prior
to filing a report with the court, which provides the evaluator with additional data to ensure that the evaluator has accurately
accounted for the family's information.
Publicly funded custody evaluations are performed within the shadow of the courts and, therefore, data is collected with the
expectation that it can be made available to the parties and the courts during the court process. Leaving a trail, or audit trail, of
case notes based on telephone calls, interviews, home observations, and collateral contacts provides a transparent methodology
where others can confirm or dispute the evaluator's findings and recommendations. Leaving an audit trail means adopting a
spirit of openness and documenting each step in data collection and analysis.
13 cv 1944 CAB BLM
The public sector's emphasis on public accountability provides the public sector evaluation with an additional procedure of
credibility. Supervision, or peer debriefing and support, is a mechanism for debriefing that guards against bias and keeps the
evaluator in-check throughout the evaluation process. Supervisors approve reports and recommendations before the evaluator
submits the report to the court, which provides the opportunity for any bias found in the reports to be vetted out. Publicly
funded custody evaluations are also subject to a hierarchical accountability built into the publicly funded system. For example,
the evaluators are directly accountable to their supervisors, but also to the Clinical Coordinator and The Children's Lawyer
of Ontario.
RECOMMENDATIONS IN PUBLICLY FUNDED EVALUATIONS
STEPPING OUT OF LITIGATION
The parents' inability to resolve conflict following divorce has many harmful effects, including higher rates of psychological
problems for children (Emery, 1999) and higher rates of postdivorce litigation (Emery, 1994). This conflict is often perpetuated
by the lack of clear boundaries and ambiguity in parenting arrangements following divorce (Madden-Derdich, Leonard, &
Christopher, 1999). Kelly (1997) suggests that the best interest of the child during the divorce process should not be an
adversarial term, but should be looked at in psychological and developmental terms. Providing recommendations to the parents
regarding the best interests of the child has the potential to empower parents to find ways to *250 create parenting plans
within a set structure without future litigation, which can create further tension and animosity between the parents (Kruk, 1992;
Birnbaum & Radovanovic, 1999).
PROCESS VERSUS OUTCOME
While Tippins and Wittmann view recommendations as an outcome derived from scientific testing, this model views
recommendations as a logical set of choices for the family to consider. Most evaluations (forensic, investigative, etc.) contain
an educative element and offer assistance to parents to divert focus away from the conflict to the needs of the children.
Recommendations are also helpful because they signal an end to the dispute. Those of us who work with this population know
that the conflict takes on a life of its own, and the parents and energy of the conflict resists closure.
INTERPRETIVE PROCESS
Unlike positivistic approaches of moving toward inferential analysis of the data, the naturalistic evaluator uses an interpretive
process beginning with making sense of what is heard and observed during the data-gathering process and then building
an understanding of the meaning of the data through data analysis of the evidence and known research in the field. In this
interpretive process, the parties' accounts and positions in the custody and access dispute, as well as the views and preferences of
the children and the collateral information, are all visible in the report and these descriptions are synthesized into interpretations
and recommendations. In this approach, the recommendations do not claim to have predictive value but rather are presented to
the parents as possibilities for effective parenting plans and to the court for what they are worth.
TRANSPARENCY IN THE RECOMMENDATIONS
By detailing the description of the methods and explaining both the manner in which the evaluation was conducted and the
investigator's reasoning, the evaluator addresses issues of congruence, as well as the appropriateness and adequacy of the
recommendations. The report is a transparent mechanism that provides the court and the parties with the ways in which the data
was gathered and analyzed. This transparency provides the parties and their lawyers with the available information to dispute
the evaluation in accordance with legislative procedures or local practice.
13 cv 1944 CAB BLM
APPLICABILITY OF RECOMMENDATIONS
A qualitative evaluation should make no claim of the generalizability of its findings nor make claims on the probabilistic or
predictive functions of the recommendations. Rather, these recommendations should be tested for their applicability of the
findings. The onus is on the evaluator to provide a well articulated rationale as to the applicability of the recommendations.
RECOMMENDATIONS AS TOOLS FOR PARENTAL ROLES
Recommendations are often presented to the parents in the form of a disclosure meeting before the publicly funded evaluator
submits a final report to the court. The disclosure meeting is an educational exercise, which informs the parents about the
methods used to *251 gather the information, the evidence collected during the investigation, the process that the evaluator
used to analyze the information, and the recommendations that flow from the analysis. Recommendations made during the
disclosure meeting provide valuable information for the parents, who are the individuals who ultimately will have the most
influence on the child's well-being (Bala, 2004).
CONCLUSION
Tippins and Wittmann present a cogent argument based on their philosophical position. I concur that there are limitations
in attempting to predict human behavior in custody cases. From a qualitative perspective, recommendations are considered a
natural extension of the evaluation process because of its interpretive process, transparency in data gathering, and notions of
applicability. Publicly funded custody evaluations have received preliminary support for their methods of responding to the
needs of the families involved in custody disputes (Birnbaum & Radovanovic, 1999). The role of the public sector custody
evaluator is to provide the parents with a credible and transparent evaluation with recommendations to help the parents adapt
to their new living arrangements following divorce and separation.
REFERENCES
Ackerman, M. J., & Ackerman, M. (1996). Child custody evaluations practices: A 1996 survey of psychologists. Family Law
Quarterly, 30, 565-586.
Association of Family & Conciliation Courts. (1994). Model standards of practice for child custody evaluations. Retrieved
September 2, 2004, from http://www.afccnet.org/pdfs/Child_Model_Standards.pdf.
Austin, W. G. (2002). Guidelines for utilizing collateral sources of information in child custody evaluations. Family Court
Review, 40(2), 177-184.
Bala, N. (2004). Assessments for postseparation parenting disputes in Canada. Family Court Review, 42(3), 485-510.
Birnbaum, R., & Radovanovic, H. (1999). Brief intervention model for access-based postseparation disputes. Family &
Conciliation Courts Review, 37, 504-513.
Board of Registration for Social Workers in British Columbia. (2002). Standards of practice child custody and access
assessments. Retrieved September 2, 2004, from http://www.brsw.bc.ca/pdf/CustodyAcessStd.pdf.
Clark, B. (1995). Acting in the best interest of the child: Essential components of a child custody evaluation. Family Law
Quarterly, 29, 19-37.
13 cv 1944 CAB BLM
Creswell, J. W. (1998). Qualitative inquiry and research design: Choosing among five traditions. Thousand Oaks, CA: Sage.
Denzin, N. K. (2002). Social work in the seventh moment. Qualitative Social Work, 1(1), 25-38.
Emery, R. E. (1994). Renegotiating family relationships: Divorce, child custody, and mediation. New York: Guilford Press.
Emery, R. E. (1999). Changing the rules for determining child custody in divorce cases. Clinical Psychology: Science &
Practice, 6(3), 323-327.
Fossey, P., Harvey, C., McDermott, F., & Davidson, L. (2002). Understanding and evaluating qualitative research. Australian
and New Zealand Journal of Psychiatry, 36, 717-732.
Hetherington, E. M., Stanley-Hagan, M., & Anderson, E. R (1989). Marital transitions: A child's perspective. American
Psychologist, 44(2), 303-312.
Horvath, L. S., Logan, T. K., & Walker, R. (2002). Child custody cases: A content analysis of evaluations in practice.
Professional Psychology: Research and Practice, 33(6), 557-565.
Karras, D., & Berry, K. K. (1985). Custody evaluations: A critical review. Professional Psychology: Research and Practice,
16, 76-85.
Kelly, J. B. (1997). The best interests of the child: A concept in search of meaning. Family & Conciliation Courts Review,
35(4), 377-387.
*252 Kruk, E. (1992). Psychological and structural factors contributing to the disengagement of non-custodial fathers after
divorce. Family & Consiliation Courts Review, 30(1), 81-101.
Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry. Beverly Hills, CA: Sage.
Madden-Derdich, D. A., Leonard, S. S., & Christopher, F. S. (1999). Boundary ambiguity and coparental conflict after divorce:
An empirical test of a family systems model of the divorce process. Journal of Marriage and the Family, 61, 588-598.
Melton, G., Petrilia, J., Poythress, N., & Slobogin, C. (1997). Psychological evaluations for the courts: A handbook for mental
health professionals and lawyers. New York: Guilford Press.
Saskatchewan Association of Social Workers. (2001). Standards in custody/access for registered social workers in
Saskatchewan. Retrieved September 2, 2004, from http://www.sasw.ca/standards/Custody%20_% 20AccessStandards.pdf.
Sherman, E., & Reid, W. (1994). Qualitative research in social work. New York: Columbia University Press.
Tippins, T. M., & Wittmann, J. P. (2005). Empirical and ethical problems with custody recommendations: A call for clinical
Vayda, E. J., & Satterfield, M. T. (1997). Law for social workers: A Canadian guide (3rd ed.). Scarborough, ON: Carswell
Thompson Professional Publishing.
Footnotes
13 cv 1944 CAB BLM
a1
Author's Note: I would like to thank the staff at the Office of the Children's Lawyer for their participation in this article. Special
thanks go to Clare Burns, The Children's Lawyer of Ontario and to my research assistants, Michael Saini and Camille Cooper, both
Clinical Investigators for the Office of the Children's Lawyer. I would also like to acknowledge the editorial comments provided by
Dr. Rachel Birnbaum of King's University College, University of Western Ontario in preparation for this response article.
Lorraine Martin is Clinical Coordinator of the Office of the Children's Lawyer, Ministry of the Attorney General of Ontario.
43 FAMCR 246
End of Document 2014 Thomson Reuters. No claim to original U.S. Government Works.
13 cv 1944 CAB BLM
A SHORT COMMENTARY ON TIMOTHY M. TIPPINS AND..., 43 Fam. Ct. Rev. 266
Family Court Review
April, 2005
A SHORT COMMENTARY ON TIMOTHY M. TIPPINS AND JEFFREY P.
WITTMANN'S EMPIRICAL AND ETHICAL PROBLEMS WITH CUSTODY
RECOMMENDATIONS A CALL FOR CLINICAL HUMILITY AND JUDICIAL VIGILANCE
Hon. Linda Dessau
a1
Copyright 2005 by Association of Family and Conciliation Courts; Hon. Linda Dessau
In this commentary, the call for clinical humility and judicial vigilance in custody recommendations is confirmed as valid and
the Australian experience, where the child custody report writer has for some years been permitted to express an opinion on
the ultimate issue, is considered. The inherent risks are briefly discussed, and the question of who of the judge and the social
scientist might be better placed to decide the exquisitely difficult children's issues after family breakdown is touched upon. It
suggests that a combination of the expert's opinion and judicial fact finding probably produces a result that is as good as it
gets. But a greater danger is highlighted. It is the impact of the adversary system, and whether it is suitable in any event to
these sensitive court decisions.
Keywords: expert opinions on the ultimate issue in child custody cases; Family Court of Australia; less adversarial proceedings
The call by the authors for clinical humility and judicial vigilance is valid, but I respectfully suggest that it begs a fundamental
question. Why is it that a judge, an expert in the law, is better placed than a social scientist to decide the exquisitely difficult
issues thrown up in the aftermath of family breakdown? The quick answer might be because he or she hears all the evidence,
and can make findings of fact, but that begs a further question. It suggests that there is an ultimate truth, and that the adversary
system, designed for traditional civil disputes about contracts or injuries, is able to cater to family law's complexities and
nuances, and the delicate projections as to what might be best for young children.
I make two points. The first is that a combination of the expert's opinion and judicial fact finding is probably as good as it gets
in the present system. The second is that, arguably, the answer lies elsewhere altogether, in a reassessment of the adversary
system, and the recognition that it is not the optimal forum for children's cases.
In Australia, the Commonwealth Evidence Act was passed in 1995. It codified the law in relation to evidence in the federal
courts, including the Family Court of Australia. The Commonwealth Evidence Act altered the long-standing rule precluding
opinion evidence as to the ultimate issue.
1
Since then, the Family Court has received reports and evidence from expert witnesses
as to the orders the court might make, including where children should live, what contact they should have with the other parent,
whether a parent should be able to relocate them, or as to specific issues, such as the children's school or religion.
The Family Court of Australia is a free-standing, specialist court. Its primary jurisdiction is in relation to divorce, property
settlement, spousal maintenance, and all aspects of parenting following family breakdown. There are forty-four judges. There
are fifty-seven legally qualified registrars, engaged in case management and mediation. There are ninety-two counselors--
psychologists or social workers--engaged in the counseling of parties, mediation, and most important for this commentary, the
preparation of court-ordered family reports.
13 cv 1944 CAB BLM
*267 In Australian family law, the children's best interests are the paramount consideration.
2
The court must bear in mind
the object and the principles of the Family Law Act,
3
in particular the parents' responsibilities for the care, welfare, and
development of the children, and the children's right to know and be cared for by both parents. The matters to be taken into
account in determining a child's best interests are set out in the Act.
4
They include the child's wishes, the nature of the
relationship with the parents and other people, the parents' capacities to provide for the needs of the child, the need to protect
the child from harm or violence, any practical difficulties of contact, and the child's maturity, sex, and background.
Most cases are resolved by negotiation or mediation. Only about 5% of custody cases are heard by a judge. A Family Report is
prepared in most of those cases. The framework for expert evidence is contained in the Family Law Rules 2004.
5
The thrust
is toward the preparation of a single expert's report, to avoid the preponderance of various reports from hired guns, with the
risk that a wealthier party can afford the most experts, and a particular risk for the children, who could be exposed to numerous
interviews as the parents shop around for a favorable expert opinion.
In the sort of cases requiring judicial determination, the complexity still frequently means that a number of experts are involved.
In addition to the Family Report writer, there may be psychiatrists who have assessed the parents, psychologists or counselors
who have counseled the children, protective workers who have investigated allegations of abuse, or previous report writers, if
for one reason or another a report was obtained from a different source on an earlier occasion. This is not infrequent in difficult
cases, where there are, for example, allegations that a child has been sexually abused.
Family Reports are essentially designed to assess family relationships. In the course of preparing a report, a counselor generally
reads affidavit material, other expert reports, and material submitted by the legal representatives for the parents, or the legal
representative for the children. He or she interviews the parents, the children, and any other adults or relatives relevant to the
child's family life. The counselor also speaks with teachers, doctors, or other counselors involved with the family. It is common
for a recommendation to be made on the ultimate issue, although frequently, alternative recommendations are given, to take
into account the possibility of alternative factual findings.
It is interesting to consider the effects in Australia after nearly ten years in which experts have been able to express an opinion
on the ultimate issue. I emphasize that my observations are personal, not empirical. I also emphasize that as a judge, I have
the blinkered perspective of the cases that come before me. But experience and anecdote inform me of some broader effects,
including the impact on cases in the pretrial phase.
First, parties negotiate, not only as it is traditionally said, in the shadow of the law, but very much in the shadow of an
expert Family Report. As noted by the Australian Law Reform Commission (ALRC),
6
reports provide valuable information to
parents, and may be the only source of information as to the children's best interests. In view of the distress and self-absorption
of some family litigants, the Family Report can serve as a concrete reminder of the views and interests of children.
7
However,
the ALRC also noted a concern as to the possibility of counselor-decided outcomes,
8
where untested reports can be used as
leverage for settlement. Although an agreement is considered a good thing, as parents and children must continue to deal with
each other into the future and to agree and cooperate, many parents complain that they feel pressured into settlement, and an
unfavorable report can be a source of pressure to settle, and a source of subsequent disquiet.
*268 Second, in Australia, with a limited legal aid dollar, a parent generally must qualify for public legal assistance according
not only to means, but also to merit. The authors express misgivings about judges relying too heavily on a clinician's opinion.
Such concerns are compounded when a legal aid officer, usually without the experience of a judge, and certainly without the
benefit of the evidence, predetermines the merit of a party's position, according to the Family Report. It is inarguable that it
is a danger.
The big question is, how dangerous is it for a judge to receive the expert's opinion as to how the case should ultimately be
decided? The answer, in my view, is that it does have its risks. The value of expert opinion of course depends on the experience
13 cv 1944 CAB BLM
and expertise of the writer. It depends on the information received. It depends on the cooperation of the parties, and the
interaction between the parties and the particular counselor. It depends on interpretation. In these respects, the risks associated
with the counselor's evidence are similar to the risks associated with much of the evidence, which needs to be sifted and sorted,
compared, analyzed, and carefully evaluated by the judge.
The risks associated with an expert expressing a view on the ultimate issue are probably in direct proportion to the skills of
the counselor, and the skills of the judge. It makes for a complex equation. The requisite skills of the expert are well analyzed
by the authors in their article.
Let me deal briefly with the requisite skills of the judge. The judge must not abdicate responsibility, no matter how tempting it
may be, to simply rely on a report to provide the magical answer to an excruciatingly difficult question. The expert's evidence
must be scrutinized. On what material did he or she rely? What is the precise nature of the expert's expertise and experience?
Is he or she from an agency that, for example, has an overriding philosophy that a purported victim of sexual abuse is always
to be believed, or one sympathetic to alleged perpetrators? How does the report sit with the rest of the evidence? Is it based
on factual findings that the court can make? Has the expert been given the opportunity to reconsider in light of the evidence in
court? Is the expert defensive or partisan? These are just examples of the complex considerations for the judge.
I am satisfied that the dangers implicit in an expert report as to the ultimate issue are outweighed by the value of the expert
perspective, in turn weighed in the balance with all of the evidence. The greater danger is presented by a more fundamental
problem--the suitability of the adversary system in cases in relation to children.
The adversary system may work well enough when two parties have different versions as to how a particular contract arose, or
whether or not a snail found its way into a drink bottle, and who was responsible. But in deciding the future of young children,
it risks leaving them out of the process, and the judge may be forced to make decisions without all the relevant evidence, as the
parties engage in their own forensic conflict, even when there is a lawyer in the middle trying to keep an eye out for the children.
In this respect, countries like Australia and the United States of America have much to learn from our friends in those parts of
Europe where the judge takes a leading role in the conduct of the hearing. He or she decides the issues to be determined, the
evidence to be called, the way the evidence is given, and the manner in which the hearing is conducted. He or she can shift
between determining contentious material and issues, and using mediation techniques to reach a conclusion. The approach is
focused on the interests of the child and the parties' proposals for the future, rather than the history of the parties' relationship.
The Family Court of Australia is conducting a trial of less adversarial proceedings. The combined skills of the judge and the
counselor are integral to it, although the judge still remains the independent decision maker.
Footnotes
a1
Justice Dessau has been a judge in the Family Court of Australia since 1995. Prior to that she was a magistrate in the Children's
Court, Coroner's Court, and Melbourne Magistrates' Court. She pursues interests in judicial education, case management, child
abuse cases, and mediation. In addition, she is involved in various community boards and councils including school, hospital, opera,
and football organisations.
1
Commonwealth Evidence Act, 1995, s. 79-80 (Austl.)
2
Family Law Act, 1975, s. 65E (Austl.)
3
Family Law Act, 1975, s. 60B (Austl.)
4
Family Law Act, 1975, s. 68F(2) (Austl.)
5
Family Law Rules, 2004, Part 15.5 (Austl.)
13 cv 1944 CAB BLM
6
Report 89: Managing Justice: A Review of the Federal Civil Justice System, 2000
7
Paragraph 8.178
8
Paragraph 8.182
43 FAMCR 266
13 cv 1944 CAB BLM
THE BENEFITS AND RISKS OF CHILD CUSTODY..., 43 Fam. Ct. Rev. 260

Family Court Review
April, 2005
THE BENEFITS AND RISKS OF CHILD CUSTODY EVALUATORS MAKING RECOMMENDATIONS TO THE
COURT: A RESPONSE TO TIPPINS AND WITTMANN
Philip M. Stahl
a1

Copyright 2005 by Association of Family and Conciliation Courts; Philip M. Stahl
In their article, Tippins and Wittmann articulate a strong position against custody evaluators making specific custody
recommendations to the court. This response will focus on my thoughts about their concepts and my belief that
recommendations can be useful and appropriate some of the time, even though there may be significant problems with certain
types of recommendations and problems with some evaluators.

In their article, Empirical and Ethical Problems with Custody Recommendations: A Call for Clinical Humility and J udicial
Vigilance, Tippins and Wittmann strongly discourage custody evaluators from making recommendations related to child
custody and the ultimate issue to the court. They describe four levels of data and inferences that evaluators gather and work
with. They articulate that evaluators can collect Level I and Level II data and make conclusions, and they believe that those
observations and conclusions are appropriate to share with the court. They acknowledge that Level III inferences about
custody-specific variables are potentially relevant, as long as the evaluator limits him/herself to statements of risk associated
with those variables. They are adamant that the evaluator should not make recommendations about Level IV conclusions,
which they refer to as custody-related shoulds. They argue that this is unethical, given the lack of empirical evidence to
support that evaluators are capable of making such Level IV conclusions. While I agree with some of their concerns, I do not
support their conclusion about avoiding recommendations or their statement that such conclusions are unethical.

Tippins and Wittmann articulate that there is a lack of scientific research sufficient to allow custody evaluators to make such
recommendations. They say that a lay person can provide the same observational data that the expert evaluator can provide,
suggesting that the expert evaluator has little more to offer the court. In contrast, because there is a twenty-five-year history
of research related to the effects of divorce on children, the effects of high conflict on children, parenting skills and styles that
promote healthier outcomes in children, domestic violence, custody and access, attachment, childrens resiliency, and
childrens coping, I believe there is sufficient research to allow evaluators to make such recommendations. This research
exists in both the divorce and child development literature.
1
Lay persons would not know about or be able to integrate that
knowledge base into their understanding of a particular family being studied. Experts who know this research and who are
trained in doing child custody evaluations will understand how to integrate that research into their analysis and
recommendations. This is certainly consistent for what Tippins and Wittmann refer to as Level III statements and I believe it
gives much greater latitude than they do for well-trained child custody evaluators to make such analyses. Of course, the
psychological research related to these issues is in its infancy, sometimes controversial, and continues to evolve over time.
Hence, it is critical for custody evaluators to stay up-to-date and current with the research in the field.

*261 In addition, over the last ten or more years, there has been a growth in the literature about how to conduct child custody
evaluations. There have been several books written on conducting child custody evaluations. Family Court Review, the new
Journal of Child Custody, and various psychology journals have taken the lead in publishing relevant articles to assist child
custody evaluators in understanding relevant issues related to performing child custody evaluations.
2
There has been a
growing emphasis on both the science and the art in this work. While Tippins and Wittmann articulate that there is no place
for the art of child custody evaluations, I disagree. The art of doing an evaluation refers to the ability to utilize a scientific
style in gathering data while understanding how to integrate that data into sensible and well-articulated recommendations to
the family and the court. Child custody evaluation literature has raised the standard for custody evaluators to use a scientific

13 cv 1944 CAB BLM

approach, gather relevant data from multiple sources, and integrate that data into the analysis and recommendations. I believe
that this allows well-trained and competent evaluators to make Level III analyses and Level IV recommendations to the court.

Before addressing when it is appropriate for a custody evaluator to make recommendations, I want to point out that custody
evaluators have an obligation to inform the clients that recommendations will be made to the courts, and that the clients may
not agree with those recommendations. This is critical so that the client understands before agreeing to participate in the
evaluation that such recommendations will be made. Once that is clear, I believe that there are many reasons why it is
appropriate for well-trained child custody evaluators to make recommendations to the court. Reasons for this include:
The evaluator is best able to understand all dynamics from a neutral, unbiased position;

The evaluator is the person most trained to understand the needs and functioning of all family members;

The evaluator is the person who is best able to think conceptually and abstractly and then apply that thinking concretely to
the children and their psychological best interests;

The evaluator is best able to put childrens needs above those of the adults and/or the courts;

The evaluator is in the best position to integrate the knowledge base from the literature into the specific needs of the
children and family;

Most judges prefer to have the input of the evaluators recommendations, and recommendations often help families settle
without a trial;

Recommendations help give the judge a starting point to consider various options for the family; and

Recommendations help everyone stay focused on the psychological best interests of the child.

Along with the above, while Tippins and Wittmann discuss what they believe to be problems with recommendations on the
ultimate custodial issue, there are certain spheres in addition to the ultimate issue in which recommendations can effectively
be made in working with families that are typically seen in child custody evaluations. These areas include:
For medium-conflict families:
A parenting plan that can include a recommended time share and schedule for parents and their children;

Suggested plans for holidays and vacations; and

Recommendations related to the primary focus of the dispute for a given family.

In high-conflict families, all of the above plus:
A decision-making hierarchy;

Conflict resolution recommendations;

Specialized issues such as drug testing or substance abuse counseling;

*262 Issues regarding a special needs child; and

Domestic violence interventions.

Other useful areas of recommendation can include:
Therapy, with specific goals for parents and/or children;


13 cv 1944 CAB BLM

Modifications as children get older and developmental needs are expected to change;

Educational needs of the children; and

The need for minors counsel and/or parent coordinator.

Having outlined the reasons why I believe recommendations are important and useful and the areas where I believe they
make sense, I agree with some of the concerns raised by Tippins and Wittmann. For example, I worry about poor quality
child custody evaluations and the lack of standards in the field. While California has set an excellent example with Rules of
Court governing procedures and ethics in child custody evaluations (CA Rule of Court 5.220
3
) and continuing education and
experience requirements for child custody evaluators (CA Rule of Court 5.225
4
), few states have any such rules or established
guidelines. Rather than take the evaluations that set a bad example and decide that all recommendations are unethical, I
would support a call for improved national standards (see, e.g., ABA Wingspread report
5
) as a way to reduce the poor quality
evaluations and recommendations based on limited information. AFCC has established a new task force designed to revisit
and update the Standards
6
drafted ten years ago, when the field was in relative infancy.

I also agree that the judge needs to exert judicial influence when making a determination. J udges should be encouraged to
consider recommendations as advisory only, recognizing the limitations of applicability of social science research to a
particular family being evaluated. Evaluators are urged to put such cautionary language in the body of their report. At the
same time, a neutrally appointed evaluator is the one person who knows the family best and the judge should be the one to
determine how much confidence to have in the evaluators report, analysis, and recommendations.

While there is a growing problem with the rotation of judges in and out of family court, and there is always a need for judges
to have adequate training in all relevant areas of family law, in particular child development, domestic violence, high-conflict
divorce, custody, and access, this does not mean that evaluators should not make recommendations. Rather, judicial officers
need to be better informed about the kinds of information they will be receiving and about what constitutes good and bad
evaluation data. It may be the poorly trained judge who benefits the most from informed recommendations, while he/she gets
the necessary judicial education. Two national organizations (National J udicial College and the National Council of J uvenile
and Family Court J udges) sponsor programs for judges in family law that include information on custody evaluations and
experts in the curriculum. Many state judicial education programs sponsor similar courses and workshops.

I also agree that many custody evaluators have significant problems with their recommendations. Reasons for such
difficulties can include:
The evaluator does a poor or limited job of gathering data. If the data is unreliable, the recommendations are also likely to
be unreliable;

The evaluator goes beyond the limits of knowledge gained in the evaluation;

The evaluator bases his/her recommendations on personal biases or philosophy rather than on the family data and the
research;

The evaluator makes recommendations about anyone not seen as part of the evaluation;

The evaluator makes recommendations that are inconsistent with the data gathered;

The evaluator does not explain how the evaluation data leads to particular recommendations;

*263 The evaluator does not report on data which does not support his/her recommendations;

Recommendations appear to be form rather than tailored to the specific family and data;

Recommendations provide little room for negotiation and settlement; and

13 cv 1944 CAB BLM

The evaluator becomes rigid and does not consider data that would support different recommendations.

Along with this, I believe that certain issues are more likely to lend themselves to specific types of recommendations, or
perhaps no recommendation at all. For example, in cases where there are allegations of domestic violence, safety issues and
perhaps anger management or batterers treatment will be paramount factors in the recommendations. With substance abuse
cases, treatment will be important. In relocation cases, where one parent wishes to move with the child and the other parent
wants the child to remain, it will be important to fully understand statutory or case law before making a recommendation.
Depending on state law, the evaluator may need to help the court in assessing whether or not there is a primary parent and
whether or not the move will prejudice the welfare or best interests of the child. Such cases may need multiple
recommendations, depending on whether or not the move with the child is or is not allowed by the court.

Finally, I also believe that there are times when an evaluator is wise to avoid making any recommendations at all. These may
include:
When the data does not support making any recommendations;

When there is no psychological research to support potential recommendations or when the research is quite controversial;

When the data between choices is relatively even;

When the risks/benefits of various options are relatively even; and

When the court does not want specific recommendations.

In these instances, I would recommend doing what Tippins and Wittmann recommend in all evaluations, that is, take the
courage to refrain from making recommendations, even if the court is asking for specific recommendations.

In conclusion, I believe that the position that Tippins and Wittmann take is like throwing the proverbial baby out with the
bath water. There are problems with untrained and inexperienced custody evaluators making recommendations, especially if
they do a poor job and receive inadequate consultation and/or supervision. There are problems when custody evaluators use
psychological tests as the basis for their custody recommendations, rather than as information to help in the generation of
hypotheses. There are certainly problems when custody evaluators use their biases or personal agendas to take custody away
from a parent due to alleged parental alienation or domestic violence and recommend supervision without good data. There
are problems when evaluators use few rather than multiple sources of data and do not have a scientific approach to the
process. There are also problems when judges are either too rushed due to a heavy caseload or too new to family law and do
not understand the issues of child custody. It may be difficult for them to make informed decisions based on the data of the
report. Instead, they simply rubber-stamp a custody evaluators recommendations without the necessary scrutiny to determine
if it qualifies as an expert opinion.

I agree with the call for humility, for both judges and custody evaluators. I agree with the call for caution in
recommendations and for evaluators to make an affirmative statement in the report suggesting that the court and the parents
need to understand the limitations of an evaluators recommendations. In my workshops I have called for increased care in
our *264 analysis and in the presentation of all relevant data, including data that does not support the conclusions reached.
However, while these problems warrant improvements in the manner in which child custody evaluators do their job, and
improvements in the way judges utilize custody evaluator recommendations, it does not warrant the cessation of evaluator
recommendations.

Furthermore, I am concerned about Tippins and Wittmanns statement that making recommendations is unethical. While I
have no problem with anyone challenging the appropriateness of making custody recommendations, to call it unethical when
neither the American Psychological Association nor any licensing board has ever done so is potentially quite damaging. In
my opinion, such statements will only increase the burden that psychologists face when asked to perform child custody work
and could inadvertently lead litigious and angry clients to file frivolous licensing and/or ethics complaints against any

13 cv 1944 CAB BLM

psychologist who makes recommendations about the ultimate issue in child custody work. Unless and until the American
Psychological Association or a licensing board makes an affirmative statement that the giving of recommendations about the
ultimate issue is unethical, I am against practitioners or researchers, regardless of their stature and beliefs, making such a
claim, preferring that they be as humble as they want custody evaluators to be.

Finally, I am in agreement with the Florida statute
7
that states [i]t is a conflict of interest for a psychologist who has treated a
minor or any of the adults involved in a custody or visitation action to perform a forensic evaluation for the purpose of
recommending with which adult the minor should reside, which adult should have custody, or what visitation should be
allowed. While therapists for children and parents might have valuable information to add to the courts understanding of a
family, they should not be making custody recommendations, because they have not evaluated all parties to the case.

Footnotes

a1

Philip M. Stahl, Ph.D. a Board member of AFCC, is a psychologist in private practice in Danville, California specializing in
custody evaluations. He is a frequent presenter at meetings of AFCC and judicial education programs throughout the country. He
has done extensive writing in the field of high-conflict divorce and his books include Conducting Child Custody Evaluations: A
Comprehensive Guide (Sage, 1994), Complex Issues in Child Custody Evaluations (Sage, 1999) and Parenting After Divorce
(Impact Publishers, 2000).

1

See, e.g., Kelly, J ., & Emery, R. (2003). Childrens adjustment following divorce: Risk and resilience perspectives. Family
Relations: Interdisciplinary Journal of Applied Family Studies, 52(4), 352-362; Center for the Future of Children (1999). The
future of children: Children and divorce. Los Altos, CA: The David and Lucille Packard Foundation; and Kelly, J ., & Lamb, M.
(2000). Using child development research to make appropriate custody and access decisions for young children. Family &
Conciliation Courts Review, 38(3), 297-311.

2

See, e.g., Stahl, P. (1994). Conducting child custody evaluations. Thousand Oaks, CA: Sage; Stahl, P. (1999). Complex issues in
child custody evaluations. Thousand Oaks, CA: Sage; Gould, J . (1998). Conducting scientifically crafted child custody
evaluations. Thousand Oaks, CA: Sage; Bow, J ., & Quinnell, M. (2001). Psychologists current practices and procedures in child
custody evaluations five years after american psychological association guidelines. Professional Psychology: Research and
Practice, 32(3), 261-268; Gould, J ., & Stahl, P. (2000). The art and science of child custody evaluations: Integrating clinical and
forensic mental health models. Family & Conciliation Courts Review, 38(3), 392-414.

3

Court-Ordered Child Custody Evaluations. California Rules of Court, Rule 5.220 (2004), available from http://
www.courtinfo.ca.gov/rules/titlefive.

4

Education, Experience, and Training Standards for Court-appointed Child Custody Investigators and Evaluators. California Rules
of Court, Rule 5.225 (2004), available fromhttp://www.courtinfo.ca.gov/rules/titlefive.

5

Wingspread Conferees. (2001). High-conflict custody cases: Reforming the systemfor children--Conference report and action
plan. Family Law Quarterly, 34, 589.

6

Association of Family & Conciliation Courts. (1994). Model standards of practice for child custody evaluations. Madison, WI:
Author.

7

Florida Business and Professions Code, 64B19--18.007. Requirements for Forensic Psychological Evaluations of Minors for the
Purpose of Addressing Custody, Residence or Visitation Disputes.

43 FAMCR 260
End of Document



13 cv 1944 CAB BLM


13 cv 1944 CAB BLM
TIPPINS AND WITTMANN ASKED THE WRONG QUESTION:..., 43 Fam. Ct. Rev. 554
Family Court Review
October, 2005
Reader Commentary
TIPPINS AND WITTMANN ASKED THE WRONG QUESTION: EVALUATORS
MAY NOT BE EXPERTS, BUT THEY CAN EXPRESS BEST INTERESTS OPINIONS
Nicholas Bala
a1
Copyright 2005 by Association of Family and Conciliation Courts; Nicholas Bala
Tippins and Wittmann (2005) provide an important analysis of the limitations of child custody evaluations, but they are wrong
to propose that court-appointed evaluators should be precluded from making recommendations about best interests decisions.
While some of the evidence of evaluators may fail to meet the high standard of reliability expected for expert evidence, the
role of court-appointed evaluators in child-related cases is not the same as the role of party-retained experts in other types of
litigation, and the legal basis for their involvement in the family law dispute resolution process is very different. The family
courts should not apply the expert evidence standard when deciding how to use the evidence of a court-appointed evaluator,
but rather should use a more flexible standard that takes account of the family law context. If the Tippins and Wittmann proposal
is adopted, it will have negative implications for the resolution of family law cases, including making settlements less common,
thereby deleteriously affecting children.
Keywords: child custody evaluation; expert opinion evidence; best interests of the child
THE PROBLEM WITH THE TIPPINS AND WITTMANN QUESTION
Lawyer Timothy Tippins and psychologist Jeffrey Wittmann provide an important critique of the child custody evaluations
in their recently published article in Family Court Review,
1
but I believe that their most controversial proposals about the
appropriate scope of testimony for court-appointed evaluators are highly problematic. Their proposals would significantly limit
the scope of the testimony of evaluators, precluding recommendations about best interests decisions. If their proposals are
adopted by the family courts, there is no assurance that there will be better resolution of cases and there will be significantly
greater financial and human costs involved in resolving family law disputes, with negative effects for the court system, parents,
and children.
In my view, Tippins and Wittmann are arguing that there is one central question that must be asked in dealing with the role
of mental health professionals in family law cases:
What parts of the testimony that evaluators give in child-related cases meet the high standards for the
admission of party-retained expert testimony that have been articulated by the Supreme Court of the United
States in Daubert
2
(and by the Supreme Court of Canada in its similar decision in R v Mohan
3
)?
*555 They correctly point out that significant parts of the testimony that evaluators commonly give do not satisfy this standard.
Accordingly, they argue that evaluators should avoid making statements about their conclusions or inferences about custody
relevant constructs, and should not make recommendations about best interests decisions.
13 cv 1944 CAB BLM
I believe that Tippins and Wittmann have asked the wrong question. I argue that the role of a court-appointed evaluator in
child-related family law cases is legally and functionally different from the role of the party-retained expert, and hence Daubert
(and Mohan in Canada) should not be applied when dealing with child-related court-appointed evaluations. Rather family law
practitioners--both judges and lawyers--should adopt an approach that takes account of the context of the case and the standards
of the evaluation field when dealing with four related but distinct questions:
(1) Should a court order an evaluation and, if so, on what terms?
(2) What use should the parties make of the evaluation in helping to settle the case?
(3) What use should a judge make of the evidence, opinions, and recommendations of a court-appointed evaluator?
(4) How should a judge deal with evidence of mental health professionals who have not been appointed by the court, but are
retained by one party to present evidence, for example, to critique or rebut the evidence of a court-appointed evaluator?
Much of the analysis that Tippins and Wittmann provide about the limitations of social science research about children and
parental separation is sound. There are very significant difficulties in applying the research that has been conducted about the
effects of different types of child care arrangements to any specific case. The complex nature of the reality of the lives of
children and their parents often makes it impossible to use the existing research to make highly accurate predictions about
specific cases. Further, and perhaps most profoundly, as Tippins and Wittmann point out, the best interests standard is a legal
and socio-moral construct, not a psychological construct [so] .... there is no empirically supportable method or principle by
which an evaluator can come to a conclusion with respect to best interests entirely by resort to the knowledge base of the mental
health profession.
4
Until now most experienced family law judges and lawyers have favored a broader approach, with a clear majority of
practitioners wanting court-appointed evaluators to provide recommendations about the best interests decisions regarding
custody, visitation, and child care.
5
It is, however, clear that the critique of Tippins and Wittmann presents a serious challenge
to past practices. The analysis offered by Tippins and Wittmann, and other scholars with similar views,
6
may be starting to
influence the family courts in the United States
7
(and Canada
8
), with some judges citing Daubert (and Mohan) to restrict
the admissibility of evidence of mental health professionals. Their analysis deserves attention, but, in my view, some of their
central recommendations should not be followed.
THE LEGAL BASIS FOR A COURT TO APPOINT AN EVALUATOR
From a legal perspective, a useful place to begin a critique of the Tippins and Wittmann analysis is to recognize that court-
appointed child custody evaluators become involved in family litigation process on a very different legal basis from party-
retained experts in other litigation.
The law about the admissibility at trial of the testimony of party-retained experts is one of a number of laws which govern the
involvement of professionals as witnesses in the litigation process. These laws were largely developed in the context of non-
family law *556 litigation. While there is some variation between jurisdictions, there are significant similarities in the laws
about party-retained experts. There are professional and ethical rules that govern how lawyers should retain and instruct experts,
but ultimately a party-retained expert is only going to become involved, and stay involved, in the litigation process if one party
believes that the expert will support his or her case. It is not uncommon for one party to initially consult with an expert and
then discover that the expert's opinion may not support his or her position, and decide not to involve that expert in the case. For
ordinary civil litigation, there are laws that govern the preparation and disclosure to other parties of a report prepared by a party-
retained expert and pretrial examination of an opposing party's expert, and there are rules to limit the number of experts that a
13 cv 1944 CAB BLM
party may call as witnesses. In cases where the mental or physical health of a party may be at issue, there are generally rules to
govern whether and how one party's expert may examine the other party. In criminal cases, somewhat different laws may apply
to the pretrial involvement of experts, but, as in ordinary civil litigation, an expert will only remain involved in the process and
be called as a witness if one side believes that the party-retained expert will have evidence that the court should hear.
9
A different set of laws apply to the involvement of court-appointed evaluators in child-related family law disputes. Most
jurisdictions have a statute or rules of court that allow a judge to appoint an evaluator (in some jurisdictions called an assessor).
There is significant variation in the qualifications and experience of evaluators, the methodologies they employ, and the time
they can spend on each case. What is significant is that the evaluator is expected to be a delegated factfinder and opinion former
for the court, though both the factual findings and the opinions are subject to full challenge at a later hearing.
Judges may act on their own motion to appoint an evaluator, though usually in family law cases the judge will only make such
an appointment with the consent of the parties, and often the order is made at the specific request of the parties. A central point
is that the child evaluator is appointed and involved in the litigation process under the auspices of the court or with the consent
of both parties, and not as a result of being selected and engaged by one party.
At least in some measure, the party-retained expert is involved in the litigation process by one party in the belief that the expert
will advance the position of that party. While it is unfair to conclude that party-retained experts are biased because they are
being paid by just one side, there is likely to be some expectation by a party who retains an expert that the expert will have
an opinion favorable to that party. Indeed, if an expert's opinion does not support the case of a party, that expert will not be
involved by that party in the proceedings. By way of contrast, a court-appointed evaluator is selected in the expectation that
he or she is an unbiased professional.
The laws about the appointment and involvement of court-appointed evaluators in family law litigation were enacted for a
different purpose than the laws about party-retained experts. The special laws about the appointment of an evaluator by the
court reflect a recognition that the justice system has a unique responsibility when making a best interests decision about a
child's future. By having a court-appointed expert, a best interests case is different not only from non-family litigation, but also
from a family law case (or portion of a case) that deals with the economic issues and which is governed by the ordinary rules
of civil litigation and expert evidence.
It is possible, and sometimes appropriate, for a court to impose restrictions or give directions to an evaluator at the time of
appointment. In some cases, for example, there may be financial or time constraints that require a relatively narrow focus of
inquiry for the evaluator. *557 Or the judge, based on the submissions of the parties, may determine that it would only be
appropriate to have the evaluator do only one thing, such as interview the children and communicate their express wishes to
court, or assess whether a child was sexually abused. If the terms of the appointment limit the scope of an evaluator's role, then
that professional should not exceed the original mandate by expressing opinions about other matters. However, if there are no
restrictions placed on the evaluator at the time of appointment, it generally will be appropriate for that professional to express
an opinion about the child's best interests.
FACILITATION OF SETTLEMENT
Family lawyers and judges generally recognize that the court-mandated evaluation process can often help produce a settlement
of a child-related dispute which could not be reached through mediation or lawyer-assisted negotiation, and this is often a factor
when deciding whether to appoint an evaluator. The settlement promotion function of the court-appointed evaluator has no
direct analogue with the function of the party-retained expert in other types of litigation. The settlement promoting function of
evaluation may be significantly reduced if the evaluation is limited and does not offer a best interests recommendation. The
facilitation of settlements is a valuable function for the overburdened judicial system. Further, while an evaluation can be costly
for parents (though in some jurisdictions it may be paid for by the government), it is inevitably less expensive than a trial; if an
evaluation results in a settlement, the cost of the evaluation is money well spent by the parties. Perhaps most importantly, unless
13 cv 1944 CAB BLM
there are abuse or violence issues that have been improperly dealt with, a settlement that the parents are prepared to accept is
almost always preferable for a child than the emotional trauma resulting from continuing litigation.
Tippins and Wittmann argue that the reason an evaluation often results in a settlement that is similar to the recommendation
of the evaluator is that parents are told by their lawyers that a judge is likely to follow the recommendations of the evaluator,
and taking the matter to trial is unlikely to result in a different outcome.
10
There is some validity to this argument, at least in
some cases, though in many respects this is merely another way of saying that judges tend to respect the opinions of evaluators.
However, in many cases, it is not simply the expectation that a judge will follow the recommendation that leads to a settlement,
but rather it is the evaluation process and the information which it provides to parents and their lawyers that produces the
settlement. As a result of the evaluation, parents may have a sense that they and their children were heard by a neutral party who
has their children's interests at heart, and they may be ready to accept a settlement based on the recommendation of that person.
The evaluation process may also produce information about their children (such as their needs, perspectives, and wishes) of
which parents were not fully aware, and this may help promote a settlement. Good evaluators often develop a rapport with
parents; they make an effort to write their report with sensitivity to each parent's strengths and weaknesses, and they are prepared
to meet with the parents to explain their recommendations. It is understandable that many parents are prepared to accept these
recommendations as the basis for a settlement.
The settlement promotion function of evaluation reflects the fact that in some respects the evaluator may be regarded as
an initial, legally mandated decision maker for resolving child-related disputes, and attempting to challenge the evaluator's
recommendations has some of the characteristics of an appeal. While judges consistently say that they are not bound by the
recommendations of an evaluator and there are many reported cases in which the recommendations are not followed, in most
cases the ultimate decision at a trial is *558 broadly consistent with the recommendations of an evaluator. In practice, the
party seeking to challenge an evaluator's recommendations does have an onus to show why the evaluator was wrong, and this
may be analogized to the onus that a party has on an appeal. Indeed, some of the ways of challenging an evaluation before
a judge may be analogized to the standard grounds for an appeal, such as demonstrating unfairness by the original decision
maker. Challenging an evaluator is, however, quite different from an appeal of a trial judgment in two critical respects. First, a
party challenging the evaluator has complete latitude to establish that the evaluator made a factual error which was central to the
recommendations proffered, and, hence, that the recommendations should be rejected. Second, the evaluator can be called as
a witness who is subject to cross-examination in court, and the weaknesses and limitations of the report and recommendations
can be fully explored before the trial judge.
THE UNIQUE NATURE OF BEST INTERESTS DECISIONS AND THE ADMISSIBILITY OF EVIDENCE
The legal rules that restrict the admissibility of expert opinion evidence at trials developed in both the United States and Canada
primarily to restrict the use of potentially unreliable evidence in tort
11
and criminal cases.
12
There are some differences in
the articulation of the rules about the admissibility of expert evidence in Canada and the United States, and indeed within the
United States there are two different rules that are applied at the state court level at which family law cases are dealt with, the
Frye standard and the Daubert standard.
13
However, for present purposes, there is a broad similarity in these legal rules, in
that they are relatively restrictive. The context in which these rules developed is very different from a child-related family law
dispute, and accordingly it is submitted that the general legal rules that have been developed for the admission of the testimony
of the party-retained expert should not apply to child-related family cases, and in particular should not apply to the testimony
of the court-appointed evaluator. By enacting laws to allow for court-appointed evaluators, the legislators have made important
modifications to the ordinary rules governing the litigation process, and this requires different treatment of the court-appointed
evaluator's evidence in a trial.
It is not the qualifications or expertise of the court-appointed evaluator or even the nature of the matters in dispute that require a
difference in approach. Rather, it is the nature of the proceedings and the role of the evaluator that require a different approach to
this type of evidence. Some of the issues that arise in family law cases, for example, related to child abuse or domestic violence,
13 cv 1944 CAB BLM
may also be the subject of criminal or tort proceedings. A judge in the family law case may receive and rely on evidence from a
court-appointed child custody evaluator that might not be admissible if the same person were called as a party-retained expert
witness in the criminal or tort case dealing with the similar factual issues. Indeed, as discussed more fully below, if a mental
health professional who regularly appears as a court-appointed evaluator is retained by one party in a best interests case, the
issue of the admissibility of the proposed testimony of that expert may be subjected to a higher standard of scrutiny than when
that expert is a court-appointed expert.
It is inevitable that judges will make use of their own understanding of human behavior and child development in dealing with
family law cases, as well as knowledge that they may acquire from judicial education programs or their own reading. This can
be problematic to the extent that a judge may be unconsciously basing decisions on misconceptions, stereotypes, or personal
experiences that are very different from those of the individuals in the *559 case before the court. Having an evaluation
prepared for the court is often the best way to educate the judge about relevant research in the social sciences, and relate it
to the specific case before the court. An independent evaluation can provide a court with relatively objective and insightful
information about the parents and children. While there may be limitations in an evaluator's knowledge or disputes about the
research, these may be tested through cross-examination or even by calling other expert evidence. If the evaluator does not
establish a social science knowledge context for a case, it may be impossible for the parties to know or challenge the basis of
the judge's understanding of the social context of their case.
One significant feature of the context in which the general legal rules about the admissibility of expert evidence developed is
the possibility, or likelihood, that the case will be determined by a jury. There is a concern that jurors, in comparison to judges,
may be more easily misled by an expert witness, and may unduly rely upon his or her testimony. Child-related family law cases
are never decided by a jury (except in Texas), and hence there may be less need for restrictive rules about the admission of
expert evidence in these cases.
The focus of inquiry in most types of court cases is retrospective-- determining what happened in the past. At least in theory,
these are objective inquiries into factual issues. A child-related family law case centers on the question of what arrangements
for the care of this child will be in the child's best interests. Inevitably much of the evidence to make this decision will relate
to past conduct, but the primary focus of inquiry is prospective. Although various types of objective evidence may be relevant
to a best interests case, the central question has an inherently subjective and value-based element. In some cases there may be
clear and objective criteria for establishing harmful or bad parental conduct that can be linked to specific negative outcomes
for children. But as Tippins and Wittmann point out, in many cases there is no clear criterion for establishing that one type of
parental conduct is better for a child than another. Even if a clear link can be established between particular parental conduct
or characteristics and likely child outcomes, there is often not an objective basis for establishing whether one type of outcome
is best for a child.
Given the inherent indeterminacy of child-related family law cases, it is appropriate for judges to not unduly restrict the
opinions that a court-appointed evaluator may express. This is not, however, to argue that there should be no judicially imposed
restrictions on court-appointed evaluators. Judges do have a gatekeeping function with these witnesses, if for no other reason
than to limit the time and expense of the proceedings, both for the parties and the court system.
The first, and perhaps most useful, point at which to exercise a judicial gatekeeping function is at the time of appointment of
an evaluator. As discussed above, in some cases the judge appointing the evaluator, based on the submissions of the parties or
the judge's own reading of the materials filed, should limit the scope of the evaluation to one or more specific issues. Courts
should only be appointing evaluators with appropriate education and training for dealing with the issues raised by a case, and
in some localities this may not be possible. While in some jurisdictions there is government funding to retain psychologists or
social workers to perform evaluations, in many places there are few professionals who are willing and able to do this type of
work, or the parents may not be able to afford to have an evaluation.
13 cv 1944 CAB BLM
There may also be cases in which an evaluator has prepared a report which is consistent with the terms of the appointment, but
the court should nevertheless rule part of it inadmissible. In terms of the ongoing dispute in the United States over the standard
for the admissibility of expert evidence between the Frye test of general acceptance within *560 the relevant scientific
community and the stricter Daubert test of scientific reliability, the appropriate test for the admissibility of evidence from a
court-appointed evaluator should be specific to the family law context, but closer to the Frye test. It must be appreciated that
the role of the court-appointed child custody evaluator is unique in the legal system, because this is a professional appointed
by the court, and in a very different professional role from a party-retained expert. Accordingly, it is submitted that in a child-
related proceeding the evidence of a court-appointed evaluator acting within the terms of the original appointment should be
ruled inadmissible if the evaluator has failed to follow the generally accepted procedures or standards of his or her discipline
or of the evaluation field, or if the evaluator is expressing an opinion that is clearly beyond that professional's education or
training. This formulation of the admissibility standard accords significant respect to the evolving standards and practices of
the evaluation field, and recognizes the difficulty and lack of utility in delimiting the evaluator's evidence. It would, however,
clearly exclude certain types of testimony. It is submitted that this rule, while perhaps not fully articulated by the family courts,
is in fact consistent with the dominant judicial practices in North America. This proposed rule also allows testimony of an
evaluator to be ruled inadmissible on grounds other than lack of expertise, such as an opinion being formed through reliance on
illegally obtained materials. It is important to recognize that this is a proposed rule about admissibility and not about the weight
of evidence. Most of the deficiencies in an evaluator's report will go to its weight, not its admissibility.
THE SCOPE FOR EVIDENCE FROM PROFESSIONALS WHO ARE NOT COURT-APPOINTED
The argument made here relates to the role of the court-appointed evaluator and the admissibility of evidence from that
professional. It is submitted that other professionals involved in family litigation should generally be bound by the ordinary
rules of civil litigation in a jurisdiction. Thus, for example, an expert retained by one party to express an opinion about the
valuation of an asset in dispute in a family law cases should be required to meet the ordinary standards for the admission of
expert evidence. If a professional, like a physician, has treated one of the parents involved in litigation, that professional should
also be governed by the ordinary rules applying to party-retained experts, which will restrict the scope of that expert's testimony.
Further, in the absence of a court-appointed evaluator, the scope for testimony by a party-retained expert about best interests
issues should normally be limited. In particular, if there is no court-appointed evaluator and one party retains a mental health
professional, even one with extensive experience as a court-appointed evaluator, the opinions of that professional about the best
interests issues should generally be inadmissible.
14
There will be cases, especially in the child protection context, in which a
mental health professional retained by one party and without access to all of the parties may be able to give admissible, relevant
evidence, but such experts should be governed by the ordinary rules for expert evidence.
Where there has been a court-ordered evaluation, there should be some flexibility in the admission of expert evidence critiquing
the work of the court-appointed professional. Ideally, if one party has concerns about an evaluation, a second mental health
professional should be appointed by the court (or on consent of both parties) to conduct a second evaluation or offer a critique,
but this is not always possible. If there is no second court-appointed professional involved in a case, the parties should have some
latitude for retaining *561 professionals with experience in conducting evaluations and for introducing their evidence. While
ideally, all court-appointed evaluations will be conducted by well qualified professionals with sound judgment and awareness
of the limitations of their opinions and recommendations, this is not always the case. Given the relatively broad scope for
expressing opinions that I have advocated for the court-appointed evaluator and the interests at stake, it would be unfair to have
a narrower rule for the admissibility of opinions for a party-retained professional who is giving reply or critique testimony.
Ultimately, a court may decide to discount the weight of the evidence of the party-retained professional who challenges the
testimony of the court-appointed evaluator, as that professional will generally have less access to the parties and there may be
issues of bias in the selection of that professional. However, if appropriately qualified, fairness demands that this party-retained
professional should have as broad a scope as the original evaluator for expressing opinions.
13 cv 1944 CAB BLM
CONCLUSION: UNDERSTANDING THE LIMITATIONS OF EVALUATIONS
Tippins and Wittmann offer an important analysis of the limitations of social science research about the effects of parental
separation on children, and the difficulty in making predictions based on that research about the effects of different custodial
and visitation arrangements on children. They correctly argue for humility by mental health professionals involved in the
family court process, and vigilance by lawyers and judges about the use of the evidence of these professionals. I agree with
many of the recommendations that Tippins and Wittmann make, including the need for
15
clearer statements by evaluators in their reports of the limitations of existing research and about the limitations on an evaluator's
abilities to make reliable predictions about the future;
better education for judges and lawyers about social science research and its limitations;
better training and clearer qualification standards for those who perform court-ordered evaluations;
development of clearer standards by professional organizations whose members undertake evaluations, including the on-going
effort of the Association of Family and Conciliation Courts; and
further research to develop valid and reliable instruments for measuring the effects of separation and making predictions about
the effects of future child care arrangements on children.
However, for the reasons discussed in this article, I disagree with one of their central and most controversial conclusions, that
there should be a moratorium on evaluators providing recommendations about best interests decisions.
I respectfully agree with Justice Linda Dessau of the Family Court of Australia who, in a commentary on the Tippins and
Wittmann article, observed that in dealing with the excruciatingly difficult issues related to the future care of a child, the
combination of a good evaluation and a good judge is probably as good as it gets in the present system.
16
She concludes
that judges should have the benefit of receiving all the relevant evidence, including the opinions and recommendations of
the evaluators about the ultimate issue.
In practice, there is great variation in the experience and expertise of evaluators, but there is also great variation in the experience
and interests of judges who make best interests decisions. While ideally both the evaluator and judge will have the skills,
knowledge, and expertise to effectively promote the best interests of children involved in family litigation, in practice if one
should be weaker, it is hoped that the child's interests will be protected by having the other involved in the process. It is a reality
that, at least in North America, some judges lack the expertise and experience to deal effectively and appropriately with *562
family law cases; in these cases, the children involved will be fortunate if there has been a good evaluator involved, as the judge,
aware of his or her limitations, may at least rely on that professional's opinion. The reality is that resources for the resolution
of family law cases are limited; if too high a standard is set for the admission of evidence from a court-appointed evaluator in
a child-related case, this will not result in better evidence being introduced, it will simply result in the court making a decision
with less evidence, and inevitably more reliance on the values and sometimes ill-informed opinions of the judge.
Footnotes
a1
Correspondence: bala@post.queensu.ca
Author's Note: I wish to acknowledge funding support from the Social Sciences & Humanities Research Council of Canada for the
preparation of this article. I also wish to thank Prof. Andrew Schepard for his advice and support for writing this commentary and
Ms. Katherine Duvall-Antonacopoulos (Queen's University L.L.B. Candidate 2006) for her editorial assistance.
13 cv 1944 CAB BLM
Nicholas Bala is a Professor of Law at Queen's University in Kingston, Canada and was a Visiting Professor at Duke Law School.
1
Timothy M. Tippins & Jeffrey P. Wittmann, Empirical and Ethical Problems with Custody Recommendations: A Call for Clinical
Humility and Judicial Vigilance, 43 FAM. CT. REV. 193 (2005).
2
509 U.S. 579 (1993).
3
R. v. Mohan, [1994] 2 S.C.R. 9.
4
Tippins & Wittmann, supra note 1, at 215.
5
See, e.g., Jame Bow & Francella Quinnell, Critique of Child Custody Evaluations by the Legal Profession, 42 FAM. CT. REV. 115
(2004), who report on a study of family law practitioners in Michigan. They found that 84% of judges and 86% of attorneys want
evaluators to provide a recommendation about custody, and 91% of judges and 90% of attorneys want a recommendation about
visitation.
6
See, e.g., William O'Donohue & April R. Bradley, Conceptual and Empirical Issues in Child Custody Evaluations, 6(3) Clinical
Psychology: Science & Practice 310 (1999). See also Sarah H. Ramsey & Robert F. Kelly, Social Science Knowledge in Family Law
Cases: Judicial Gate-Keeping in the Daubert Era, 59 UNIV. MIAMI L. REV. 1 (2004) who offer a similar critique to Tippins and
Wittmann, though focusing on the (mis)use of social science research in family law cases.
7
See, e.g., S.M. v. G.M., New York Law Journal, Apr. 5, 2005, p. 20, col. 1 (Suffolk Cty Sup. Ct.); John A. v Bridget M., 16 A.D.3d
324, 791 N.Y.S.2d 421 (App.Div., 1st Dept, 2005).
8
See, e.g., Mayfield v. Mayfield, [2001] O.J. No. 2212 (QL) (Sup. Ct.).
9
There are special ethical rules that require a prosecutor to disclose to the accused and to court the opinions of experts whom the
prosecution consulted and who have expressed views that do not support the prosecution's case.
10
11
In the United States, see the tort cases of, for example, Daubert v Merrell Dow Pharaceuticals, 509 U.S. 579 (1993), and Kumho
Tire Company v Carmichael, 526 U.S. 137 (1999). In Canada, see, for example, Parsons v Canadian Red Cross (2000), 51 O.R.
(3d) 261(C.A.).
12
In the United States, see, for example, Jennifer Groscup et al., The Effects of Daubert on the Admissibility of Expert Testimony in
State and Federal Criminal Cases, 8 PSYCHOL PU. POL'Y & L. 339 (2002), and Pamela Jensen, Frye Versus Daubert: Practically
the Same? 87 MINN. L. REV. 1579 (2003). In Canada, see, for example, R. v. Mohan, [1994] 2 S.C.R. 9.
13
There is a growing body of scholarship which suggests that in practice there may not be any real difference between the Frye and
Daubert standards in terms of their effects on the restriction of expert testimony: see, for example, Edward Cheng & Albert Yoon,
Does Frye or Daubert? A Study of Scientific Admissibility Standards, 91 VIRGINIA. L. REV. 471 (2005).
14
See, for example, foraund, Johnstone v. Brighton, Campbell J [2004] O.J. 3267 (Sup. Ct.-Fam. Ct.) where, in a relocation case, the
court refused to permit a psychologist with extensive experience in conducting court-ordered evaluations, but in this case retained
by the father, to testify as an expert witness regarding attachment and the effects of relocation on a 5-year-old child.
15
16
Linda Dessau, A Short Commentary on Timothy M. Tippins and Jeffrey Wittmann, Empirical and Ethical Problems with Custody
Recommendations: A Call for Clinical Humility and Judicial Vigilance, 43 FAM. CT. REV. 266, 268 (2005).
43 FAMCR 554
13 cv 1944 CAB BLM
AN EVALUATION OF A PROGRAM FOR MANAGING..., 40 Fam. Ct. Rev. 320
Family Court Review
July, 2002
Special Issue
Papers Celebrating the 25th Anniversary of the Family Court of Australia
Magellan's Discoveries
AN EVALUATION OF A PROGRAM FOR MANAGING FAMILY COURT
PARENTING DISPUTES INVOLVING CHILD ABUSE ALLEGATIONS
Thea Brown
a1
Copyright 2002 by Sage Publications, Inc.; Thea Brown
In 1998, the Family Court of Australia introduced an interagency, multidisciplinary pilot program for managing parenting
disputes that involved allegations of child abuse. The program, known as Project Magellan, is an example of the trend in
judicial administration toward specialized court programs.
1
It involved one hundred families and several state and federal
agencies and human service organizations. This article presents the findings of an evaluation of the new program and considers
the implications of those findings and the potential for the success of similar programs in other jurisdictions. It points to
the contribution the pilot program made to the development of knowledge about child abuse in the emotive and complex
circumstances of parental separation and suggests that the interrelationship between child abuse and relationship breakdown
should be acknowledged as a legitimate matter of serious concern.
RESEARCH ON CHILD ABUSE IN THE CONTEXT OF PARENTAL SEPARATION
Child abuse has been featured as a subject of research only in the past few decades and, when considered in the context of
parental separation, it has been even slower to attract research attention. The few small studies undertaken almost twenty years
ago were rightly criticized for their contradictory conclusions, which were a probable consequence of their small sample sizes.
2
Eventually, the United States National Center for Child Abuse sponsored a large and rigorously designed study, the results of
which were released in 1988.
3
This landmark research tracked several hundred cases of parenting disputes in several states
of the United States where allegations of child sexual abuse were involved and through family, children's, and criminal courts
as well as child protection services. One important finding to emerge from the study was that the rate of false allegations in
such cases was low and amounted to only 14 percent of all allegations made. This flies in the face of the commonly held view
that allegations made in such circumstances are manufactured as weapons fashioned for use in the divorce war and, therefore,
are highly likely to be fictitious.
4
The study also documented the serious nature of the abuse being experienced by children
and an interrelationship between such abuse and partnership violence. It also showed that family law and the child protection
services were relating poorly to each other, due in part to interorganizational obstructions *321 leading to a loss of focus on
the children. Cases drifted and outcomes were poor for the children, typically resulting in a lack of protective action.
Ten years later, another large studyreported in this journal
5
and referred to hereafter as the first studywas undertaken in
Australia by the Family Violence and Family Court Research Program in conjunction with the Family Court of Australia.
6
This study tracked 150 residence and contact disputes involving all types of child abuse allegations that were dealt with in the
Victorian and Australian Capital Territory registries of the Family Court. The research showed that such disputes began as a
small but increasing proportion of child-related applications in the court and became half its workload as they proceeded toward
13 cv 1944 CAB BLM
trial.
7
Few of these disputes were resolved without energetic court intervention, and approximately 30 percent proceeded as
far as a judicial determination.
8
The first study also found that these difficult cases took a long time to complete and involved, on average, five hearings over a
period of almost eighteen months. This posed many difficulties and potential dangers for the children, whose average age was
only four years. More than one-quarter (28 percent) of the children were considered by mental health staff to be suffering from
extreme distress, which increasedand their functioning deterioratedas the cases moved through the various court processes.
Parents were found to have many problems themselves, with confirmed partnership violence being a feature of 40 percent of
the families. Finally, the interface between the child protection service and the court worked very poorly, largely because of
substantial interorganizational problems.
Simultaneously, another large Australian study of residence and contact disputes involving allegations of child sexual abuse was
undertaken in South Australia and came to similar conclusions.
9
It also found a low rate (9 percent) of false allegations, a figure
identical to the proportion of false allegations recorded for all notifications of child abuse to the child protection authorities and
suggesting strongly that parental separation by itself does not precipitate a spate of false allegations.
Similar research was conducted almost simultaneously in Canada in another large study of parenting disputes in which
allegations of child physical and sexual abuse had been made, but including only cases that proceeded to a judicial
determination.
10
The Canadian study found a higher proportion of false allegations, but there were also many parallels with the
U.S. and Australian studies, especially in relation to the difficulties of the interface between child protection services and courts.
EVENTS PRECEDING THE INTRODUCTION OF PROJECT MAGELLAN
The publication of several research reports in Australia during 1997 added to the Family Court's growing concerns about its
management of child abuse cases.
The first was the publication of the Australian Law Reform Commission's Report, Seen and Heard: Priority for Children in
the Legal Process.
11
This report drew attention to the problems children faced in all legal processes, including the Family
Court. It suggested that children were being treated as peripheral in the various systems when they should be central, and it
characterized the length of time cases involving their safety took in the courts and the implications of delays for the children as
serious matters that required attention. It also pointed to the disjunction between the child protection services and the Family
Court. In Australia, as in most countries, child protection services are placed in organizations separate from the courts, and the
negotiations between these two independent organizations (which *322 have different legislative bases, goals, procedures,
and disciplines), present a number of challenges.
The commission's findings were confirmed by the Family Violence and Family Court Research Program's first report, which
also identified a major change in the nature of the matters being presented for resolution in the Family Court. The court was
established in 1976 to manage and resolve private family disputes that flowed from marriage breakdown, most commonly
divorce, and property and child related applications. However, the studies showed that the court had become a major forum for
the resolution of family violence issues, with child, spousal, and other family-related violence becoming a major component
of the court's workloadits core business, in effect. Furthermore, the Family Court was being used as an integral part of the
child protection system by parents, other family members, and the child protection services themselves.
12
The court's review
of its pending cases acknowledged these problems, much as the previous research studies had described.
13
THE INTRODUCTION OF PROJECT MAGELLAN
13 cv 1944 CAB BLM
Following these events, Chief Justice Nicholson established a steering committee located in the Melbourne Registry of the
Family Court to consider new approaches to the management of children's matters involving child abuse allegations. Justice
Linda Dessau was appointed to chair the committee, and its membership included representatives from the Victorian Department
of Human Services, the police, legal aid authorities, the Law Council of Australia, the Commonwealth Attorney General's
Department, and the Family Violence and Family Court Research Program.
Over a six-month period, the committee developed a new program based on a series of new principles derived from the previous
research. These were identified as
a child- (rather than parent-) focused approach that included automatic legal representation for all children funded by the
state legal aid authority;
a judge-led, tightly managed, fixed-time program with preset steps to prevent delays;
early intervention with up-front resources made available at the outset;
involvement of a multidisciplinary team, the membership of which remained constant throughout the program;
use of child protection staff and in-house court counselors to investigate and assess the child abuse allegations;
the provision of clear information about the program to the families and their lawyers at the outset and feedback to the families
about the progress of their matters, including the circulation of expert reports;
close collaboration between the various services involved in the program, using multiple coordination points; and
ongoing monitoring of the program by the judge responsible.
The program began in June 1998 as a pilot with a sample of one hundred new applications for parenting orders filed in either the
Melbourne or Dandenong registries of the Family Court. The cases were selected by the list registrar and the senior counselor,
who jointly scrutinized all new applications. The families so selected were informed of the opportunity to join the program in
advance, and all accepted the invitation.
*323 The program contained four court events that were managed within the court by a team of two judges, a senior counselor
(supported by a small team of counselors to undertake family reports), and the list registrar.
The first court event was a preliminary mention attended by the parents and their legal representatives. On this occasion, the
judge explained Magellan's objectives and procedures, made some procedural and interim orders, and ordered the appointment
of a legal representative for the child. Funding for such representation was guaranteed by Victoria Legal Aid and was an
important feature of the program.
The Family Law Act allows for the appointment of a children's lawyer, but as funding for this is almost invariably provided
by legal aid, there are limitations on the extent to which the appointments are approved by the aid authorities. In addition, the
federal government requires the states to impose separate funding caps on the amounts of aid available for children and adults
in family law proceedings.
14
For the Magellan pilot, the caps were waived in relation to the costs of child representatives,
which allowed all children to be legally represented from the time of the first court event. Provided the parents passed the usual
means and merits tests, the cap on their costs was similarly waived.
At the first mention, the judge also ordered the State Department to conduct a child protection investigation and provide a
report of that investigation to the court within five weeks. This report was subsequently made available by the court to the
13 cv 1944 CAB BLM
parents' lawyers and the child's legal representative a week before the second hearing. The lawyers (but not the parents) were
also able to read the departmental file. After the hearing, the senior counselor and child's legal representative liaised with the
particular child protection worker allocated to investigate the allegations. This allowed a more holistic assessment of the child's
circumstances to be provided.
The second court event occurred seven weeks later, presupposing that the matter had not been resolved as a result of these
interventions.
15
At this time, the judge received the child protection services report and ordered that a family report be
prepared by the court counselors. Such reports provide information about relevant issues such as the parents' functioning; their
relationships with their children, each other, and additional family members; their views of the allegations; their attitudes to
their children; and their arrangements for them, including how (if at all) they saw the other parent being involved in the child's
life. In preparing the report, the counselors would interview the parents and the children and might also seek information from
other significant others such as stepparents, grandparents, and teachers. Other support services might also be approached for
information where this was considered necessary.
The third court event occurred ten weeks later, once again presupposing that no prior agreement had been reached. This took the
form of an informal prehearing conference conducted jointly by the list registrar and the senior counselor. Parents attended with
their legal representatives, and frequently child protection staff were also present. Informal discussions held with the registrar
helped identify areas of agreement and disagreement, and the registrar encouraged the parties to agree and negotiate future
arrangements for the children. If no agreement was reached, a trial date was fixed for ten to twelve weeks later. The trial was
the fourth and final court event.
*324 OUTCOMES OF THE NEW MODEL OF INTERVENTION
The pilot finished in December 2000 and was evaluated by the Family Violence and Family Court Research Program, which
had conducted the first study, funded by an Australian Research Council under a partnership scheme with the Family Court.
It sought to assess the program's achievements against seven indicators for measurement that were designed to show program
outcomes. These indicators were average number of court events per case, average length of time for state child protection
service to report back to the court, average length of time for the resolution of each case, incidence of cases proceeding to
trial, rate at which final orders broke down, changes in the size of the high level of distress group of children, and legal aid
costs per case.
The evaluation team used the same research design as was used in the first study, namely, an analysis of the court records in each
case, observations of the cases as they proceeded through the court, and interviews with staff from the diverse organizations
working on the program. In addition, as a new component, surveys were sent to the parents and to the legal representatives for
the parents and the children. A second new component involved a calculation of the costs to Victoria Legal Aid on all Magellan
cases and on a comparison group of similar cases that were not included in the program. Outcomes were assessed for each case,
on average, twelve months after completion.
It is not possible to report all the findings of the study here.
16
The remainder of this article concentrates on the program's
achievements as measured by the program indicators. While they represent only a summary, hopefully they give a clear picture
of the program's outcomes.
EVALUATION FINDINGS
Rigorous case management, combined with ongoing liaison between the different authorities, considerably reduced many of
the difficulties identified by the first study, particularly those that arose between the federal Family Court and the state child
protection authority. For example, the time taken by the child protection service to submit a report reduced from an average of
forty-two days to an average of thirty-two days. The reports themselves also changed considerably from being merely a series
13 cv 1944 CAB BLM
of boxes to be ticked or left blank to a three-page, detailed report of what investigations had been carried out, when, by whom,
and with whom. They included clear conclusions and showed an increasing recognition of the need to focus on more than the
children's short-term protection.
At the same time, the substantiation rate by child protection staff rose from 23 percent in the first study to 48 percent in Magellan.
This improvement may have been the result of the new program (in terms of better procedures and the better knowledge base
circulated to workers using the results of the first study). However, it may also have been attributable to the fact that Magellan
cases were selected on the basis of allegations of more serious abuse in which case substantiation rates would be expected to
increase.
Disputes were also resolved (whether by private agreement, negotiated agreement, consent orders, or judge order) far more
quickly than had previously been the case, with the average time to resolution falling from 17.5 months to 8.7 months. This
was particularly important given the findings of the first study that showed an association between delay and children's poor
functioning. The number of court interventions reduced also from an average of five events to three. This was important for
reasons of cost but also because the first study *325 found that each court event provided an opportunity for a change in the
children's residential arrangements. Previously, more than one-third (37 percent) of children had changed residence at any one
hearing. The stability of the children's living arrangements was, thus, improved by the pilot program's reducing the number of
hearings and the length of time between hearings.
Far fewer cases (13 percent) proceeded to a judicial determination than had done so previously (30 percent), and this had
important ramifications for the families and professionals involved in the program. Obviously, the smaller proportion of cases
going to trial (the most expensive of all court events) reduced financial costs to the court, to child protection services (whose
staff would usually attended such hearings), to parents, and to the legal aid authority. Litigation in family law matters is also
associated with increased parental hostility and makes effective parenting thereafter problematical.
In addition, only 5 percent of the final orders made broke down, contrasted with 37 percent in the first study. The resilience of
orders was measured by interrogating the court's computerized case management system, on average twelve months after the
matter appeared to have resolved, to check whether an appeal had been sought or further child-related applications filed.
Victoria Legal Aid spent on average Aus.$13,770 on all parties per case (for parents and children) averaged over all cases in
Magellan. This figure was well below the cap allowed for legal aid expenditure on family law cases and also well below the
average cost of the non-Magellan comparison group of cases (Aus. $19,867).
17
Unfortunately, it was not possible to establish
costs for the cases in the first study, although this would have been the preferred option.
An examination of the court files in the first study showed an alarmingly high incidence (28 percent) of children experiencing
high levels of disturbance as reported in affidavit and other material by psychologists, doctors, and social workers. Magellan
files were examined again to assess children's distress levels, and the incidence of those considered to be highly distressed
appeared to amount to 4 percent. The improved figure may be related to the faster court response times, as in the first study,
distress levels appeared to increase as time in the court process increased. However, this cannot be conclusively linked to the
introduction of the new program, and other factors may have come into play.
Parents were very supportive of the program and reported high levels of satisfaction for themselves and their children. Their
most common recommendation for change was for an increase in post-court services (particularly court-sponsored services) to
ensure that the orders made were enforced. Legal practitioner satisfaction levels were also high, with those representing children
tending to be more satisfied than those representing parents. Lawyers' recommendations for change were very similar to those
of the parents and were directed at the introduction of more post-court services for families. Both parents and practitioners
wanted all components of the program to be maintained.
13 cv 1944 CAB BLM
CAN THE PROJECT MAGELLAN PROGRAM OPERATE ELSEWHERE?
The Magellan program was uncomplicated in design, and only two relatively minor problems were identified in the evaluation.
The first was that 29 percent of cases were delayed because proceedings in other jurisdictions had not been completed. The
problem was particularly serious in the 15 percent of cases that involved concurrent criminal proceedings relating *326 to
the child abuse allegations. Such delays are obviously outside the control of the Family Court and thus not remediable by
changes to the Magellan design. The second problem occurred even less frequently and was caused by parties' asking for new
parenting arrangements to be tried for a fixed period that extended the usual time limits of the program. The court agreed with
the request in 4 percent of cases, and all these arrangements subsequently broke down, but it cannot be said that they did not
warrant being tested in the first place.
The most difficult aspect of the pilot program was obtaining and maintaining the collaboration of the various state and
federal authorities and human service organizations. The involvement of staff members with a strong commitment to change
greatly assisted the process, and maintaining the composition of the committee organization was significant to its success.
Coordination needs were continuous, and all members of the steering committee moved literally and figuratively between the
member organizations. Finally, the vigorous leadership of Justice Dessau, combined with her keen understanding of Magellan's
objectives, were essential to its success.
There are therefore strong reasons to suggest that Magellan could operate effectively in other Australian states. The Family
Court of Western Australia
18
is now piloting a similar program, which has been called Columbus. The major difference
between the two is that cases included in Columbus contain allegations of both serious physical and sexual child abuse and
partnership violence.
THE REALITY OF CHILD ABUSE IN THIS CONTEXT
The study's findings threw more light on child abuse in this controversial context. They showed (as had the first study) that
child abuse is a major cause of parental separation.
19
The evaluation of Magellan documented the extent (48 percent) to which
parents left their partners because of incidents of child abuse, serious partnership violence, or both.
20
Sadly, the study also echoed the findings of previous studies that leaving an abusive partner does not bring the abuse of children
to an end. As those working with separated families are only too aware, abuse frequently continues at the time of hand-over
and during contact periods.
21
This was found to be the reality for 80 percent of families in the Magellan program.
Recent research also suggests that parental separation may precipitate child abuse. Wilson analyzed all research on children's
well-being postseparation and postdivorce
22
and concluded that the loss of the protection gained from two parents living
together rendered female children vulnerable to sexual abuse, no matter what parenting arrangements were put in place after
the separation. She saw family courts as central to child protection.
The evaluation study showed that those responsible for the abuse were family members. Fathers were most commonly
responsible (53 percent), then mothers and fathers jointly (11 percent), a family group of people such as a father and a grandfather
(11 percent), mothers (8 percent), siblings or step siblings (7 percent), stepmothers (5 percent), and stepfathers (5 percent).
In terms of which alleged perpetrators were likely to have allegations made against them substantiated, the evaluation showed
that those less frequently accused of sexual abuse (stepmothers, stepfathers, grandfathers, siblings, and stepsiblings) were
the most likely to be substantiated as perpetrators. Surprisingly, mutual allegations of any abuse made by both parents
simultaneously were highly likely to be substantiated. The allegations least likely to be substantiated were those of physical
abuse inflicted by mothers. It should be emphasized that no allegation against any one category of person was invariably found
13 cv 1944 CAB BLM
to be untrue. Apparently *327 unlikely or unusual allegations (such as a mother's sexually abusing a daughter or stepson) were
substantiated in some cases. Similarly, seemingly paranoid allegations of sexual abuse of all children in a family by a father,
uncle, and grandfather were substantiated in three cases.
The child protection service substantiated abuse in its first investigation in 48 percent of the Magellan cases, and the court
counselors noted that children were at risk of abuse in a further 16 percent of cases. In 4 percent of all cases where abuse had
not been substantiated by the child protection service, the counseling staff renotified the service of abuse, and these cases were
substantiated subsequently by that service. In one judicially determined matter, judgement found there to be no abuse despite
the previous finding of the child protection service.
Physical, sexual, and emotional abuse were all found to occur with some frequency, but neglect was less likely, given the focus
of the program. Sexual abuse in the context of parental separation was more common than in the profile of abuse relating to
notifications to the state child protection services. This is probably due to the propensity for parents to leave their partners once
sexual abuse is discovered and the subsequent need to seek protective court orders. Children were most commonly found to
have experienced more than one category of abuse.
Little has been written about the extent (if any) to which child abuse that occurs in and around parental separation is different
from or similar to the etiology of such abuse in other contexts. The Magellan evaluation showed many similarities (in terms
of perpetrators and types of abuse) and some differences (such as the higher incidence of sexual abuse and lower incidence
of neglect). Child protection workers commented on the greater anger of these parents in relation to other parents, which may
be related to the anger associated with the separation, the reasons for it, and the loss of trust that it causes. A theme of family
secrecy emergedfor example, family members with secret histories of abuse or with secret relationships to the children and
family members whose very existence had been kept secret. Often the discovery of such secrets led one parent to leave the
other, thereby bringing the abuse allegations to the attention of the Family Court.
CONCLUSIONS
Project Magellan was found to have achieved its goals, and its continuance was recommended by a subsequent Commonwealth
government inquiry into the Australian family law service system.
23
The pilot program was simple in its design and execution,
and there are no reasons to suggest that it would not be transferable elsewhere. Finally, the evaluation of the pilot program
suggested that child abuse and parental separation should be regarded as interrelated and that child abuse is an event along the
pathway to separation and vice versa.
Footnotes
a1
Thea Brown is a professor of social work at Monash University in Victoria, Australia, and director of the Family Violence and Family
Court Research Program, which undertook the evaluation of Project Magellan. She has completed a number of research studies
on child abuse in the context of parental separation. She was recently a member of the Family Law Pathways Advisory Group, an
Australian Commonwealth government committee that reviewed family law services.
Author's Note: The author wishes to acknowledge the enormous amount of assistance, including financial, staffing, and administrative
support, given to the research project by the Family Court, and in particular the support given by Chief Justice Alastair Nicholson
and Justice Linda Dessau. The project was funded jointly by the Australian Research Council and the Family Court of Australia. The
author wishes to acknowledge the assistance provided by the steering committee, which included members from the Commonwealth
Attorney General's Department; Victoria Legal Aid; the Department of Human Services, Victoria; Victorian Police; and the Family
Law Section of the Law Council of Australia and the Family Court.
1
Mario Fabri, In Search of Quality: Judicial Governance, Innovation and Trends in Europe. In Justice, Courts and Community: The
Continuing Challenge, 25th Anniversary Conference, Family Court of Australia, Sydney (2001).
13 cv 1944 CAB BLM
2
Charles Schudson, Antagonistic Parents in Family Courts: False Allegations or False Assumptions of Child Abuse? 1 JOURNAL
OF CHILD SEXUAL ABUSE, 111-13 (1992); and Patricia Toth, All Child Abuse Allegations Demand Attention, 1 JOURNAL OF
CHILD SEXUAL ABUSE, 117-18 (1992).
3
Nancy Thoennes and Jessica Pearson, Summary of Findings from Child Sexual Abuse Project. In Sexual Abuse Allegations in Custody
and Visitation Cases, edited by B. Nicholson and J. Bulkey (Washington, DC: National Legal Resource Center for Child Advocacy
and Protection, 1988).
4
Mary Kaye and Julie Tolmie, Father's Rights Groups in Australia, 12 AUSTRALIAN JOURNAL OF FAMILY LAW, 19-67 (1998).
However, as mentioned later in this article, low false allegation rates have been a characteristic of the several studies conducted in
Australia in recent years.
5
Thea Brown, Margarita Frederico, Lewsley Hewitt, and Rosemary Sheehan, Problems and Solutions in the Management of Child
Abuse Allegations in Custody and Access Disputes in the Family Court, 36 FAMILY AND CONCILIATION COURTS REVIEW,
431-43 (1998).
6
Thea Brown, Margarita Frederico, Lewsley Hewitt, and Rosemary Sheehan, Violence in Families: Report Number One, The
Management of Child Abuse Allegations in Custody and Access Disputes in the Family Court of Australia (Clayton, Victoria,
Australia: Family Violence and Family Court Research Program, Monash University, 1998), at 112.
7
The Family Law Act provides a number of opportunities for parties to settle their disputes without judicial intervention. Mediation,
case conferences, and other interventions are available, and parties are encouraged (and in some instances required) to avail themselves
of alternative dispute resolution.
8
For many years, the court has noted that the percentage of all applications that result in a court determination presided over by a
judge has remained constant at 5 to 6 percent.
9
Marie Hume, Child Sexual Allegations and the Family Court (unpublished master's thesis, Humanities and Social Sciences,
University of South Australia, 1997).
10
Nicholas Bala and John Schuman, Allegations of Sexual Abuse When Parents Have Separated, 17 CANADIAN FAMILY
QUARTERLY, 192-243 (1999).
11
Australian Law Reform Commission and Human Rights and Equal Opportunity Commission, Seen and Heard: Priority for Children
in the Legal Process (Canberra: Commonwealth of Australia, Australian Government Publishing Service, ACT, 1997).
12
Although the term Family Court implies that child protection matters are determined at the federal level, the Australian constitution
does not give the court power to deal with child protection (public law) issues in which the state is a party; see Hon. Alastair Nicholson,
Setting the Scene: Australian Family Law and the Family CourtA Perspective from the Bench 40 FAMILY COURT REVIEW,
277-94 (2002 [this issue]).
13
See supra note 11.
14
Aus.$10,000 to $12,000 for adults and $15,000 to $17,000 in relation to children.
15
One-third of the Magellan cases were resolved before this stage of the process was reached.
16
Thea Brown, with Rosemary Sheehan, Margarita Frederico, and Lesley Hewitt, Resolving Violence to Children, an Evaluation of
Project Magellan (Clayton, Victoria, Australia: Family Violence and Family Court Research Program, Monash University), at 132.
17
Only twenty cases could be found that matched those in Magellan on a number of characteristics but were not included in the pilot.
Thus, there are limitations in matching the two groups for costing purposes.
18
The Family Court of Australia exercises jurisdiction across the country in all states and territories except Western Australia, where
a separate Family Court was established in 1976. The two courts work closely together and their legislation is very similar.
19
See supra note 6.
13 cv 1944 CAB BLM
20
Id.; and Family Law Pathways Advisory Group, Out of the Maze: Pathways to the Future for Families Experiencing Separation
(Report of the Family Law Pathways Advisory Group) (Canberra: Commonwealth of Australia, Australian Government Publishing
Service, ACT, 2001).
21
See supra note 18; Marian Hester and Lesley Radford, DOMESTIC VIOLENCE AND CHILD CONTACT ARRANGEMENTS IN
ENGLAND AND DENMARK (Bristol, UK: Policy Press, 1996); and Helen Rhoades, Regina Graycar, and Margaret Harrison, The
Family Law Reform Act, 1995: The First Three Years (Sydney: University of Sydney and Family Court of Australia, 2000).
22
Robyn Wilson, Children at Risk. Paper presented to the 10th International Conference in Family Law, Brisbane, Australia (2000).
23
Family Law Pathways Advisory Group, Out of the Maze.
40 FAMCR 320
13 cv 1944 CAB BLM

1 Family Court Review April 2005 Debate Master

Загружено:

Сведения о документе

Исходное описание:

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

1 Family Court Review April 2005 Debate Master

Загружено:

Авторское право:

Доступные форматы

EMPIRICAL AND ETHICAL PROBLEMS WITH CUSTODY..., 43 Fam. Ct. Rev.

Вам также может понравиться