Вы находитесь на странице: 1из 15


discussions, stats, and author profiles for this publication at:


Interpersonal Dependency and

Personality Pathology:
Variations in Rorschach Oral
Dependency Scores Across Axis
II Diagnoses

Article in Journal of Personality Assessment January 2001

Impact Factor: 1.84 DOI: 10.1207/S15327752JPA7503_08 Source: PubMed


15 51

4 authors, including:

James Christopher Fowler

Baylor College of Medicine


Available from: James Christopher Fowler

Retrieved on: 21 June 2016
Copyright 2000, Lawrence Erlbaum Associates, Inc.

Interpersonal Dependency and

Personality Pathology: Variations in
Rorschach Oral Dependency Scores
Across Axis II Diagnoses

Robert F. Bornstein
Department of Psychology
Gettysburg College

Mark J. Hilsenroth
Department of Psychology
University of Arkansas

Justin R. Padawer
Department of Psychology
University of Tennessee

J. Christopher Fowler
Erik H. Erikson Institute for Training and Research
Austen Riggs Center
Stockbridge, Massachusetts

Although several investigations have examined the relationship of Rorschach Oral

Dependency (ROD; Masling, Rabie, & Blondheim, 1967) scores to Axis I diagnosis,
there has been very little research assessing variations in ROD scores across Axis II
personality disorders (PDs). In this study, ROD scores were compared in 5 PD groups
(borderline PD inpatients, borderline PD outpatients, avoidantdependent PD outpa-
tients, narcissistic PD outpatients, and antisocial PD outpatients), and 2 non-PD com-
parison groups (psychotic disorder inpatients and college students). Borderline PD
inpatients had significantly higher ROD scores than borderline PD outpatients, anti-
social PD outpatients, and college students; no other between-group differences were
found. We discuss implications of these results for research on dependency and Axis
II psychopathology and offer suggestions for future studies.

Masling et al.s (1967) Rorschach Oral Dependency (ROD) scale has been
the most widely used projective measure of interpersonal dependency during
the last 30 years (Bornstein, 1996, 1999). ROD scores predict conformity,
compliance, help seeking, and interpersonal yielding in a variety of situations
and settings (Bornstein, Riggs, Hill, & Calabrese, 1996; Masling, Weiss, &
Rothschild, 1968; Shilkret & Masling, 1981). High scores on the ROD scale
are also associated with increased sensitivity to interpersonal cues (Masling,
Johnson, & Saturansky, 1974), an insecure attachment style (Duberstein &
Talbot, 1993), and difficulty terminating psychotherapy (Greenberg & Born-
stein, 1989).
These results support the convergent validity of the ROD scale as a measure of
interpersonal dependency (see also Viglione, 1999, for a review of research on this
issue). Other findings confirm that the ROD scale meets established psychometric
criteria for discriminant validity (Gordon & Tegtemeyer, 1983; Kertzman, 1980),
retest reliability (Bornstein, Rossner, & Hill, 1994), and interrater reliability (Juni
& Semel, 1982; Weiss & Masling, 1970). As expected, ROD scores show strong
positive correlations with scores on other projective measures of dependency
(Fowler, Hilsenroth, & Handler, 1996; Masling et al., 1967) and more modest cor-
relations with scores on self-report dependency measures (Bornstein, Rossner,
Hill, & Stepanian, 1994).
Given the widespread use of the ROD scale in clinical and research settings, there
has been surprisingly little exploration of the relationship between ROD score and
psychiatric diagnosis. In two early studies of this issue, Bertrand and Masling (1969)
and Weiss and Masling (1970) found that alcoholic psychiatric outpatients produced
higher ROD scores than nonalcoholic outpatients matched on salient demographic
and diagnostic variables. More recently, ROD scores have been found to distin-
guish eating-disordered from non-eating-disordered female psychiatric inpatients
(Bornstein & Greenberg, 1991), with eating-disordered patients showing higher
rates of both oral (i.e., food and mouth-related) and dependent imagery. High ROD
scores are also associated with elevated levels of depression in psychiatric inpa-
tients, outpatients, and nonclinical participants (ONeill & Bornstein, 1990, 1991);
in each of these samples, depressed participants had significantly higher ROD
scores than nondepressed participants matched on salient demographic and (in the
case of clinical participants) diagnostic variables.
Although the aforementioned studies assessed links between ROD scores and
several Axis I disorders, there has been only one investigation of the relationship
between ROD scores and Axis II personality disorder (PD) diagnoses. Blais,
Hilsenroth, Fowler, and Conboy (1999) assessed the correlations between ROD
scores and dimensional symptom ratings for each of the Diagnostic and Statisti-
cal Manual of Mental Disorders (4th ed. [DSMIV]; American Psychiatric As-
sociation, 1994) Cluster B PDs (i.e., antisocial, borderline, histrionic, and
narcissistic) in a mixed-sex outpatient sample. They found a significant negative

correlation between ROD scores and borderline PD symptom ratings (r = .43)

but obtained only small (and statistically nonsignificant) correlations between
ROD scores and symptom ratings for the other three Cluster B PDs.
Further research on this topic is important for two reasons. First, variations in
dependency levels have long been assumed to distinguish different forms of per-
sonality pathology (Bornstein, 1997; Millon, 1996), yet data documenting these
hypothesized intercategory differences are lacking for most Axis II PD categories.
Second, high levels of patient dependency have been shown to influence the
course of inpatient and outpatient psychological treatment (Bornstein, 1993;
Masling, 1986); elucidating the links between ROD scores and different Axis II di-
agnoses may, therefore, aid in treatment planning.
The purpose of this study was to assess the RODPD link by examining differ-
ences in ROD scores across diagnostically defined criterion groups. Five PD
groups were included: borderline PD inpatients (BPDIs), borderline PD outpa-
tients (BPDOs), narcissistic PD outpatients (NPDOs), antisocial PD outpatients
(APDOs), and a combined group of dependent and avoidant PD outpatients (D/
APDOs). Two non-PD comparison groups were also included: psychiatric inpa-
tients diagnosed with various psychotic disorders (PSDIs), and nonclinical col-
lege student control (CSC) participants.
On the basis of recent analyses of the dependencyPD relation (Beck & Free-
man, 1990; Bornstein, 1997; Millon, 1996; Widiger & Bornstein, in press), we hy-
pothesized that D/APDO and BPDI participants should show the highest ROD
scores. NPDO, PSDI, and CSC participants should show lower ROD scores and
that the ROD scores in these groups would not differ from each other. We also hy-
pothesized that BPDO and APDO participants would show the lowest ROD
scores (Blais et al., 1999; Bornstein, 1997).
In addition to assessing differences in ROD scores across these seven criterion
groups, a secondary purpose of this investigation was to examine the frequencies
of different ROD subcategory scores. Although numerous studies have assessed
the frequency of ROD whole-scale scores in different clinical and nonclinical pop-
ulations, there are no published studies that have examined the frequencies of the
16 ROD subscale scores. Thus, these data represent preliminary norms for the
ROD subscales that may be useful for future studies of ROD scores in clinical and
research settings.



Inpatient groups: BPDI and PSDI. These groups consisted of (a) 27

adult inpatients with a DSMIV diagnosis of BPD and (b) 33 adult inpatients

with a DSMIV diagnosis of a psychotic disorder (paranoid schizophrenia n =

19, undifferentiated schizophrenia n = 7, schizoaffective disorder n = 6, and psy-
chotic disorder not otherwise specified n = 1). All patients were hospitalized in
an acute-care psychiatric unit because of an exacerbation of psychotic symp-
toms, and all patients received a comprehensive intake evaluation of symptoms
and history within 48 hr of admission. Because intake evaluations were con-
ducted shortly after admission, patients had received no inpatient psychotherapy
prior to being evaluated. All patients had received outpatient therapy at some
point prior to admission, and many patients had received psychotropic medica-
tions as well. However, reliable information regarding prior medication use was
not available for all patients.1
Each patient was interviewed and evaluated by a multidisciplinary treatment
team consisting of a board-certified psychiatrist, one or more psychiatric residents,
a licensed clinical psychologist, a clinical psychology intern, an occupational ther-
apist, a licensed clinical social worker, and a registered psychiatric nurse. All team
members observed, interacted with, and interviewed the patients before rendering
initial intake diagnoses. Patients also received a thorough medical evaluation
(health history, blood work, and physical examination) and were excluded if they
suffered from a general medical condition, substance-induced symptoms, or or-
ganic mental disorder.
Intake diagnoses were established in a consensus conference by this multi-
disciplinary treatment team 48 hr after admission. DSMIV diagnoses were
made using all available sources of information, including an integration of
interview data from the different disciplines with prior hospital records, con-
sultation with outpatient treaters, and interviews with knowledgeable infor-
mants to clarify patient history and premorbid level of functioning. This
method of diagnostic practice approximates the LEAD (longitudinal expert
evaluation using all data) standard of diagnosis (Pilkonis, Heape, Ruddy, &
Serrao, 1991; Skodol, Rosnick, Kellman, Oldham, & Hyler, 1988; Spitzer,
1983). The mean kappa for the DSMIV BPD criteria set was .85. Demo-
graphic characteristics of the PSDI and BPDI groups are summarized in
the top portion of Table 1.
All patients included in the study were administered the Rorschach shortly after
admission, always within the 1st week of hospitalization. However, in all cases,
the intake diagnosis was determined before the Rorschach was administered and
ROD scores were derived. Because of this, Rorschach data were unavailable to the
clinical team when case records were reviewed and diagnoses were assigned. This
strategy helped minimize criterion contamination and ensured that Rorschach data

1Although reliable information in this area was not always available, previous studies have found

that medication history and status are unrelated to dependency levels in clinical samples (Fava et al.,
1994; Peselow, Robins, Sanfilipo, Block, & Fieve, 1992).

Demographic Characteristics of the Clinical Groups

N Age (Years) Marital Status

Group Women Men Total M SD M SD U M D W

PSDI 17 16 33 38.76 12.91 11.69 2.10 15 6 10 2

BPDI 23 4 27 33.19 10.71 13.38 2.46 13 9 4 0
BPDO 20 2 22 27.46 6.39 14.55 1.95 14 2 6 0
D/APDOa 8 3 11 27.55 8.03 14.73 2.72 6 3 2 0
NPDO 4 7 11 27.64 5.10 15.73 1.42 6 2 3 0
APDO 0 14 14 26.50 6.04 12.36 2.71 8 1 5 0

Note. U = unmarried; M = married; D = divorced; W = widowed; PSDI = psychotic disorder

inpatient; BPDI = borderline personality disorder (PD) inpatient; BPDO = borderline PD outpatient;
D/APDO = dependentavoidant PD outpatient; NPDO = narcissistic PD outpatient; APDO =
antisocial PD outpatient.
aWithin the D/APDO group, n = 6 for avoidant PD patients and n = 5 for dependent PD patients.

did not influence the diagnostic process (the use of intake rather than discharge di-
agnoses also helped ensure independence of diagnoses and projective test data).

Outpatient groups: BPDO, D/APDO, NPDO, and APDO. Partici-

pants in the outpatient groups were drawn from an archival search of files at a uni-
versity-based psychological clinic, accomplished by an exhaustive screening of
approximately 800 cases seen over a 7-year period. The selection of clinic cases
proceeded in three phases.
In the first phase, 217 cases were initially identified as having a PD diagnosis
assigned by a clinical team.
In the second phase, these 217 cases were independently rated for the presence
or absence of DSMIV diagnoses by four advanced doctoral students in an Ameri-
can Psychological Association (APA) approved clinical psychology program. The
presence or absence of symptoms was determined through a retrospective review
of patient records that included an evaluation report, notes from the initial assess-
ment, session notes for the first 12 weeks of therapy (detailing patient reports of
history, symptoms, and topics discussed during the hour), and 3-month treatment
reviews. All information regarding patient identity, diagnosis, and test materials
(including all Rorschach data) was appropriately masked or made unavailable to
the raters reviewing the case records.
Of the 217 patients reviewed in this manner, 91 met DSMIV criteria for an
Axis II PD. This PD prevalence rate is in line with PD prevalence rates obtained in
other outpatient samples (Barber & Morse, 1994; Carter, Joyce, Mulder, Sullivan,

& Luty, 1999; Grilo et al., 1998). Interrater agreement in PD diagnosis was estab-
lished by independent ratings of a randomly selected pool of 31 PD outpatients. A
kappa coefficient of .90 regarding the presence or absence of a DSMIV PD diag-
nosis was obtained.
In the third phase, the records of these 91 patients were independently rated on
all DSMIV Cluster B PD symptom criteria using the same case material and meth-
odology described earlier. Interrater reliability was established by independent rat-
ings of a randomly selected pool of 25 patients. Kappa estimates of interrater
agreement regarding the presence or absence of individual DSMIV symptoms
were as follows: APD = .86, BPD = .80, and NPD = .90.
Demographic characteristics of the APDO, BPDO, NPDO, and D/APDO
groups are summarized in the bottom portion of Table 1. Axis I diagnoses in these pa-
tient groups were as follows: APDO (substance abuse disorder n = 10), mood disor-
der n = 5, adjustment disorder n = 1, impulse control disorder n = 1), BPDO (mood
disorder n = 9, substance abuse disorder n = 6, anxiety disorder n = 3, eating disorder
n = 3, adjustment disorder n = 2), NPDO (mood disorder n = 4, substance abuse dis-
order n = 3, anxiety disorder n = 1, adjustment disorder n = 1, impulse control disor-
der n = 1), and D/APDO (mood disorder n = 8, anxiety disorder n = 4).

CSC group. Participants enrolled in undergraduate psychology classes at a

large southeastern university served as the nonclinical CSC group. These partici-
pants volunteered to take part in the study in exchange for extra course credit. Par-
ticipants in the CSC group were screened for a history of psychotherapy or
psychiatric hospitalization. A sample of 50 participants was administered the Ror-
schach by advanced graduate students in an APA-approved clinical psychology
PhD program. These graduate students had completed the entire testing curriculum
in this program prior to collecting test data. The final CSC group included 25 men
and 25 women with a mean of 14.84 years of education (SD = 1.10). The mean age
of CSC participants was 22.60 (SD = 5.30).2


The administration and scoring of all Rorschach protocols followed the procedures
articulated by Exner (1993). Rorschach protocols were scrutinized for validity, and
2A sample size of 50 was used for the CSC group because this sample size enhanced the statistical

power of the analyses without being so large as to introduce heteroscedasticity problems that might
confound the results. Although the CSC participants were somewhat younger than the clinical partici-
pants, studies have shown that from late adolescence through middle adulthood, age is unrelated to
ROD score (Bornstein, 1993; Kertzman, 1980; Masling, 1986). CSC participants did not differ from
the combined clinical participant group with respect to number of years of education.

any protocols with fewer than 14 responses and lambdas above 1.0 were omitted
from the study, as is standard in studies involving the Comprehensive System (see
Exner, 1993). For all participants, scoring of projective test data was done by Mark
J. Hilsenroth, who was unaware of group (diagnostic) assignment.
ROD scores were derived from the combined free-association and inquiry por-
tions of each participants Rorschach protocol, using the ROD scoring system of
Masling et al. (1967). In this system, a response is defined as oral dependent if it
falls into any of the following categories: (a) foods and drinks, (b) food sources, (c)
food objects, (d) food providers, (e) passive food receivers, (f) food organs, (g)
supplicants, (h) nurturers, (i) gifts and gift givers, (j) good luck symbols, (k) oral
activity, (l) passivity and helplessness, (m) pregnancy and reproductive anatomy,
and (n) negations of oral percepts (e.g., not pregnant, man with no mouth).
One point is assigned for each oral-dependent Rorschach response. Detailed re-
views of the construct validity of the ROD scale were provided by Bornstein
(1996) and Masling (1986).
To establish interrater reliability in ROD scoring, 20 Rorschach protocols were
chosen at random and rescored independently by J. Christopher Fowler, who was
unaware of the first coders scores and the participants group assignments. The
two sets of scored protocols were compared, and percentages of agreement were
calculated for the ROD. In addition, an intraclass correlation coefficient (ICC) was
computed between the two sets of scores. The resulting interrater agreement and
ICC statistics were 98% and .91, respectively.


A preliminary analysis of variance (ANOVA) revealed that number of Rorschach

responses per protocol (R) did not differ significantly across the seven groups, F(6,
161) = 1.00, p = .43. However, the number of oral-dependent percepts per protocol
was significantly correlated with overall R (r =.42, p < .0001). We, therefore, tabu-
lated a ROD percentage score by dividing the number of oral-dependent percepts
within a protocol by the R for that protocol to control for response productivity (see
Bornstein, 1996). This ROD percentage score was the basis of all further analyses.
Consistent with findings from previous studies (e.g., Bornstein, 1998b; Duberstein
& Talbot, 1993; Shilkret & Masling, 1980), women and men in our sample ob-
tained comparable ROD scores, F(1, 166) = 0.22, ns. Thus, analyses were con-
ducted with ROD scores collapsed across sex.
Table 2 contains data concerning the ability of the ROD scale to differentiate
the seven criterion groups. An ANOVA revealed a significant main effect of diag-
nosis on ROD scores, F(6, 161) = 5.74, p < .0001. Follow-up Tukey HSD tests re-
vealed that patients in the BPDI group produced protocols with significantly
higher ROD scores than those in the BPDO, APDO, and CSC groups (all ps <
.05). No other between-group differences were found.

Rorschach Oral Dependency (ROD) Scores in Different Criterion Groups

ROD Score

Group M SD Range

BPD-Ia .265 .118 .077.556

D/APDOb .204 .116 .050.421
NPDOb .202 .099 .050.417
PSDIc .185 .132 .000.583
CSCd .162 .080 .000.400
APDOe .117 .103 .000.312
BPDOf .109 .103 .000.368

Note. BPDI = borderline personality disorder (PD) inpatient; D/APDO = dependentavoidant

PD outpatient; NPDO = narcissistic PD outpatient; PSDI = psychotic disorder inpatient; CSC =
college student control; APDO = antisocial PD outpatient; BPDO = borderline PD outpatient. ROD
scores were calculated by dividing the total number of oral dependent Rorschach responses in each
protocol by R.
an = 27. bn = 11. cn = 33. dn = 50. en = 14. fn = 22.

The frequency of the 16 individual criteria that make up the ROD scale was ex-
amined for each of the clinical groups included in this study. These data are pre-
sented in Table 3. The most frequently occurring ROD subcategories were Food
Organs and Oral Activity. The least frequently occurring subcategories were Ne-
gations of Oral Percepts, Baby-Talk Responses, Gifts and Gift Givers, and Passiv-
ity and Helplessness. There were no discernable patterns of ROD subcategory
differences across the clinical groups. The last two columns in Table 3 summarize
the overall frequencies of each ROD subcategory, collapsed across diagnosis.


These results provide partial support for our a priori hypotheses. As expected,
BPDI participants produced higher ROD scores than the APDO, BPDO, and
CSC groups. However, contrary to expectations, the ROD scores of the D/APDO
group did not differ from those of other PD groups.3
The lower-than-expected ROD scores of the D/APDO group might be due in
part to the methods used to derive PD diagnoses in this investigation. PD diagno-
ses were derived from a review of patients chart records, and much of the infor-
mation contained in these records was based on patients reports of PD-related
3Although ROD scores were not significantly elevated in the D/APDO group, it is worth noting

that patients in this group had the second highest mean ROD score (.204) and the highest mean ROD
score of all outpatient groups. Only the BPDI participants had a higher overall ROD score (.265) than
the D/APDO group.
Frequencies of Rorschach Oral Dependency (ROD) Scale Subcategories in Different Clinical Groups




Foods and Drinks 2 13 1 5 1 7 0 2 3 15 0 0 1 8

Food Sources 2 11 1 5 1 5 1 7 2 9 1 7 1 7
Food Objects 0 0 0 1 1 7 0 0 1 4 0 0 0 2
Food Providers 0 1 0 1 0 2 0 0 0 0 0 0 0 1
Passive Food Receivers 2 10 1 5 0 4 1 5 1 6 1 7 1 6
Begging and Praying 0 1 0 1 0 4 0 2 0 0 0 0 0 1
Food Organs 6 31 10 38 4 36 7 41 9 49 7 59 7 39
Oral Intruments 0 0 0 1 0 0 0 0 0 0 0 0 0 0
Nurturers 1 6 1 4 0 4 0 2 1 4 0 0 1 1
Gifts and Gift-Givers 0 2 0 0 0 0 0 0 0 0 0 0 0 0
Good Luck Objects 0 0 0 1 0 4 1 5 0 0 0 0 0 1
Oral Activity 4 21 7 28 2 21 4 24 2 11 2 17 4 22
Passivity and Helplessness 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Pregnancy and Reproduction 1 5 2 7 1 5 2 12 0 0 1 7 1 6
Baby-Talk Responses 0 0 0 1 0 0 0 0 0 2 0 0 0 1
Negations of Oral Percepts 0 1 0 0 0 2 0 0 0 0 0 0 0 0

Note. PSDI = psychotic disorder inpatient; BPDI = borderline personality disorder (PD) inpatient; BPDO = borderline PD outpatient; D/APDO =
dependentavoidant PD outpatient; NPDO = narcissistic PD outpatient; APDO = antisocial PD outpatient; BR = overall base rate of a given ROD subcategory
(i.e., subcategory percept frequency divided by R); ROD % = frequency of a given subcategory relative to other ROD subcategories (i.e., subcategory percept
frequency divided by the total number of oral dependent responses). All figures are percentages, rounded to the nearest whole number. Overall means were
calculated collapsing across clinical groups.
an = 33. bn = 27. cn = 22. dn = 11. en = 14.

behaviors, cognitions, and affective responses. In other words, these chart records
contained a great deal of information regarding self-attributed traits and character-
istics, that is, traits and characteristics that are explicitly acknowledged (albeit re-
luctantly at times) by the patient (McClelland, Koestner, & Weinberger, 1989).
In contrast, the ROD scale assesses implicit (i.e., underlying, often uncon-
scious) dependency strivings, which may or may not be reported by a participant.
Previous studies have shown that in both clinical and nonclinical samples, correla-
tions between measures of implicit and self-attributed dependency needs are mod-
est at best, and typically in the .20 to .30 range (Bornstein, 1998b; Bornstein,
Rossner, & Hill, 1994; Bornstein, Rossner, Hill, & Stepanian, 1994). In this con-
text, the absence of a significant relation between avoidantdependent PD symp-
toms and ROD scores is less surprising. It may simply represent another example
of the discontinuity between individuals implicit and self-attributed dependency
needs (see Bornstein, 1998a, for a detailed discussion of this issue).
The substantial (and statistically significant) difference between the ROD
scores of the BPDI and BPDO groups might reflect one (or more) of several un-
derlying processes. It may be that the BPDO groups concerns regarding inti-
macy, and the potential loss of identity associated with merging with a valued
other, caused these patients to invoke various defenses and coping strategies aimed
at denying their underlying dependency needs (see Blais et al., 1999; Duberstein &
Talbot, 1993; Fowler et al., 1996). To the extent that ROD scores do in fact tap im-
plicit dependency strivings, these results would suggest that such defensive and
coping strategies can influence participants indirect expression of these needs on
projective tests such as the ROD scale.
Alternatively, the significant inpatientoutpatient BPD differences obtained in
this study might reflect changes in participants mood states during inpatient treat-
ment. Studies have shown that the onset of negative mood states is associated with a
significant increase in ROD scores (Bornstein, Bowers, & Bonner, 1996). To the ex-
tent that borderline patients show increases in depression following psychiatric hos-
pitalization, increases in their ROD scores would be expected. Such an explanation
is supported by the fact that the vast majority of BPD inpatients in our sample were
hospitalized following a suicide attempt or gesture related to interpersonal conflict
or loss. Concerns regarding rejection and abandonment might well have caused
these BPD inpatients to become preoccupied, at least temporarily, with dependency-
related issues, thereby inflating their ROD scores.
A third possibility concerns the BPD diagnostic criteria themselves. As numer-
ous clinicians and researchers have noted, BPD is a particularly heterogenous di-
agnostic category, and considerable controversy remains regarding the most
appropriate and useful criteria for classifying borderline patients (see Carr, 1987;
Costello, 1996; Lerner, Albert, & Walsh, 1987; Millon, 1996). It may be that in-
patient and outpatient diagnosticians use somewhat different strategies and be-
havioral anchor points in diagnosing BPD, leading to subtle, albeit important,

differences in personality functioning and defensive style in members of these two

groups (see Edell, Joy, & Yehuda, 1990; Leichsenring, 1999).
Two conclusions can be drawn from the ROD subscale data summarized in Ta-
ble 3. First, these data suggest that certain ROD subscales (in particular, Food Or-
gans and Oral Activity) are particularly common among members of clinical
samples. Second, these data indicate that, by and large, ROD subscale frequencies
are comparable across different Axis II PD groups.
A much larger and more diverse patient sample clearly is needed to construct
definitive clinical norms for the 16 ROD subscales and to address more rigorously
the question of when, if ever, ROD subscale scores differ systematically across di-
agnostically defined criterion groups. These data represent a first attempt to ad-
dress these issues. Combined with additional information from future studies,
these results can contribute to researchers efforts to derive useful norms for this
widely used dependency test.
Three limitations of this study are worth mentioning. First, we assessed only
a subset of the DSMIV PD categories. Moreover, the low base rate of D/
APDOs in our outpatient sample did not permit us to construct separate de-
pendent PD and avoidant PD groups. Replication of this study on a larger partic-
ipant sample is needed to assess the variation of ROD scores across all relevant
PD categories.4
In this context, it would be useful to include inpatient and outpatient groups for
each PD category in future investigations, especially in light of the substantial dif-
ferences in ROD scores in BPD inpatients and outpatients. To the extent that ROD
scores are affected by state (e.g., mood) variables, or by factors such as severity of
symptomatology or disturbance, within-category comparisons of inpatients and
outpatients will be especially important.
Finally, future researchers may want to examine variations in both implicit and
self-attributed dependency needs across different PD groups. Several psycho-
metrically sound dependency questionnaires are available for such an investi-
gation (e.g., Interpersonal Dependency Inventory, Hirschfeld et al., 1977;
Depressive Experiences Questionnaire, Blatt, DAfflitti & Quinlan, 1976; Three-
Vector Dependency Inventory, Pincus & Gurtman, 1995). Only by examining the
variation of both implicit and self-attributed dependency needs across the entire
spectrum of DSMIV PD categories can researchers obtain a complete picture of
underlying and expressed dependency needs in different forms of personality

4It is worth noting that ADPOs had the second lowest mean ROD score of all the groups assessed

in this study (.117), only slightly higher than that of BPD outpatients (.109). This result is consistent
with recent theoretical analyses of dependency strivings in antisocial PD, which predict low levels of
both implicit and self-attributed dependency needs in antisocial individuals (e.g., Bornstein, 1997;
Millon, 1996).


American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th
ed.). Washington, DC: Author.
Barber, J. P., & Morse, J. Q. (1994). Validation of the Wisconsin Personality Disorders Inventory with
the SCIDII and PDE. Journal of Personality Disorders, 8, 307319.
Beck, A. T., & Freeman, A. (1990). Cognitive therapy of the personality disorders. New York: Guilford.
Bertrand, S., & Masling, J. (1969). Oral imagery and alcoholism. Journal of Abnormal Psychology, 74,
Blais, M. A., Hilsenroth, M. J., Fowler, C., & Conboy, C. A. (1999). A Rorschach exploration of the
DSMIV borderline personality disorder. Journal of Clinical Psychology, 55, 110.
Blatt, S. J., DAfflitti, J. P., & Quinlan, D. M. (1976). Experiences of depression in normal young adults.
Journal of Abnormal Psychology, 85, 383389.
Bornstein, R. F. (1993). The dependent personality. New York: Guilford.
Bornstein, R. F. (1996). Construct validity of the Rorschach Oral Dependency Scale, 19671995. Psy-
chological Assessment, 8, 200205.
Bornstein, R. F. (1997). Dependency in the personality disorders: Intensity, insight, expression, and de-
fense. Journal of Clinical Psychology, 54, 175189.
Bornstein, R. F. (1998a). Implicit and self-attributed dependency needs: Differential relationships to
laboratory and field measures of help seeking. Journal of Personality and Social Psychology, 75,
Bornstein, R. F. (1998b). Implicit and self-attributed dependency needs in dependent and histrionic per-
sonality disorders. Journal of Personality Assessment, 71, 114.
Bornstein, R. F. (1999). Criterion validity of objective and projective dependency tests: A meta-analytic
assessment of behavioral prediction. Psychological Assessment, 11, 4857.
Bornstein, R. F., Bowers, K. S., & Bonner, S. (1996). Effects of induced mood states on objective and
projective dependency scores. Journal of Personality Assessment, 67, 324340.
Bornstein, R. F., & Greenberg, R. P. (1991). Dependency and eating disorders in female psychiatric in-
patients. Journal of Nervous and Mental Disease, 179, 148152.
Bornstein, R. F., Riggs, J. M., Hill, E. L., & Calabrese, C. (1996). Activity, passivity, self-denigration,
and self-promotion: Toward an interactionist model of interpersonal dependency. Journal of Per-
sonality, 64, 637673.
Bornstein, R. F., Rossner, S. C., & Hill, E. L. (1994). Retest reliability of scores on objective and projec-
tive measures of dependency: Relationship to life events and intertest interval. Journal of Personal-
ity Assessment, 62, 398415.
Bornstein, R. F., Rossner, S. C., Hill, E. L., & Stepanian, M. L. (1994). Face validity and fakability of ob-
jective and projective measures of dependency. Journal of Personality Assessment, 63, 363386.
Carr, A. C. (1987). Borderline defenses and Rorschach responses: A critique of Lerner, Albert, and
Walsh. Journal of Personality Assessment, 51, 349354.
Carter, J. D., Joyce, P. R., Mulder, R. T., Sullivan, P. F., & Luty, S. E. (1999). Gender differences in the
frequency of personality disorders in depressed outpatients. Journal of Personality Disorders, 13,
Costello, C. G. (Ed.). (1996). Personality characteristics of the personality disordered. New York:
Duberstein, P. R., & Talbot, N. L. (1993). Rorschach oral imagery, attachment style, and interpersonal
relatedness. Journal of Personality Assessment, 61, 294310.
Edell, W. S., Joy, S. P., & Yehuda, R. (1990). Discordance between self-report and observer-rated
psychopathology in borderline patients. Journal of Personality Disorders, 4, 381390.
Exner, J. E., Jr. (1993). The Rorschach: A comprehensive system: Vol. 1. Basic foundations (3rd ed.).
New York: Wiley.

Fava, M., Bouffides, E., Pava, J., McCarthy, M. K., Steingard, R. J., & Rosenbaum, J. F. (1994). Person-
ality disorder comorbidity with major depression and response to fluoxetine treatment. Psychother-
apy and Psychosomatics, 62, 160167.
Fowler, C., Hilsenroth, M. J., & Handler, L. (1996). A multimethod approach to assessing dependency:
The Early Memory Dependency Probe. Journal of Personality Assessment, 67, 399413.
Gordon, M., & Tegtemeyer, P. F. (1983). Oral dependent content in childrens Rorschach protocols.
Perceptual & Motor Skills, 57, 11631168.
Greenberg, R. P., & Bornstein, R. F. (1989). Length of psychiatric hospitalization and oral dependency.
Journal of Personality Disorders, 3, 199204.
Grilo, C. M., McGlashan, T. H., Quinlan, D. M., Walker, M. A., Greenfield, D., & Edell, W. S. (1998).
Frequency of personality disorders in two age cohorts of psychiatric inpatients. American Journal of
Psychiatry, 155, 140142.
Hirschfeld, R. M. A., Klerman, G. L., Gough, H. G., Barrett, J., Korchin, S. J., & Chodoff, P. (1977). A
measure of interpersonal dependency. Journal of Personality Assessment, 41, 610618.
Juni, S., & Semel, S. R. (1982). Person perception as a function of orality and anality. Journal of Social
Psychology, 118, 99103.
Kertzman, D. (1980). Dependency, frustration tolerance, and impulse control in child abusers.
Saratoga, CA: Century Twenty-One.
Leichsenring, F. (1999). Development and first results of the Borderline Personality Inventory: A self-
report instrument for assessing borderline personality organization. Journal of Personality Assess-
ment, 73, 4563.
Lerner, H., Albert, C., & Walsh, M. (1987). The Rorschach assessment of borderline defenses: A con-
current validity study. Journal of Personality Assessment, 51, 334348.
Masling, J. M. (1986). Orality, pathology, and interpersonal behavior. In J. Masling (Ed.), Empirical
studies of psychoanalytic theories (Vol. 2, pp. 73106). Hillsdale, NJ: Lawrence Erlbaum Associ-
ates, Inc.
Masling, J. M., Johnson, C., & Saturansky, C. (1974). Oral imagery, accuracy of perceiving others, and
performance in Peace Corps training. Journal of Personality and Social Psychology, 30, 414419.
Masling, J. M., Rabie, L., & Blondheim, S. H. (1967). Obesity, level of aspiration, and Rorschach and
TAT measures of oral dependence. Journal of Consulting Psychology, 31, 233239.
Masling, J. M., Weiss, L., & Rothschild, B. (1968). Relationships of oral imagery to yielding behavior
and birth order. Journal of Consulting and Clinical Psychology, 32, 8991.
McClelland, D. C., Koestner, R., & Weinberger, J. (1989). How do self-attributed and implicit motives
differ? Psychological Review, 96, 690702.
Millon, T. (1996). Disorders of personality: DSMIV and beyond. New York: Wiley.
ONeill, R. M., & Bornstein, R. F. (1990). Oral dependence and gender: Factors in help-seeking re-
sponse set and self-reported psychopathology in psychiatric inpatients. Journal of Personality As-
sessment, 55, 2840.
ONeill, R. M., & Bornstein, R. F. (1991). Orality and depression in psychiatric inpatients. Journal of
Personality Disorders, 5, 17.
Peselow, E. D., Robins, C. J., Sanfilipo, M. J., Block, P., & Fieve, R. R. (1992). Sociotropy and auton-
omy: Relationship to antidepressant drug treatment response and endogenousnonendogenous di-
chotomy. Journal of Abnormal Psychology, 101, 479486.
Pilkonis, P., Heape, C., Ruddy, J., & Serrao, P. (1991). Validity in the diagnosis of personality disorders:
The use of the LEAD standard. Psychological Assessment, 3, 4654.
Pincus, A. L., & Gurtman, M. B. (1995). The three faces of interpersonal dependency: Structural
analysis of self-report dependency measures. Journal of Personality and Social Psychology, 69,
Shilkret, C. J., & Masling, J. M. (1981). Oral dependence and dependent behavior. Journal of Personal-
ity Assessment, 45, 125129.

Skodol, A., Rosnick, L., Kellman, D., Oldham, J., & Hyler, S. (1988). Validating structured DSMIIIR
personality disorder assessments with longitudinal data. American Journal of Psychiatry, 145,
Spitzer, R. L. (1983). Psychiatric diagnosis: Are clinicians really necessary? Comprehensive Psychia-
try, 24, 399411.
Viglione, D. J. (1999). A review of recent research addressing the utility of the Rorschach. Psychologi-
cal Assessment, 11, 251265.
Weiss, L. R., & Masling, J. (1970). Further validation of a Rorschach measure of oral imagery: A study
of six clinical groups. Journal of Abnormal Psychology, 76, 8387.
Widiger, T. A., & Bornstein, R. F. (in press). Histrionic, narcissistic, and dependent personality disor-
ders. In H. E. Adams & P. B. Sutker (Eds.), Comprehensive textbook of psychopathology (3rd ed.).
New York: Plenum.

Robert F. Bornstein
Department of Psychology
Gettysburg College
Gettysburg, PA 17325
E-mail: bbornste@gettysburg.edu

Received August 11, 1999

Revised March 7, 2000