Child Abuse & Neglect 48 (2015) 119–130

Contents lists available at ScienceDirect

Child Abuse & Neglect

Research article

Behavior problems of children in foster care: Associations

with foster mothers’ representations, commitment, and the
quality of mother–child interaction夽
Karine Dubois-Comtois a,b,∗ , Annie Bernier c , George M. Tarabulsy d ,
Chantal Cyr e , Diane St-Laurent a , Anne-Sophie Lanctôt a , Janie St-Onge a ,
Ellen Moss e , Marie-Julie Béliveau c
a
Department of Psychology, Université du Québec à Trois-Rivières, Trois-Rivières, Canada
b
Department of Child Psychiatry and Research Center, Hôpital du Sacré-Coeur de Montréal, Montréal, Canada
c
Department of Psychology, Université de Montréal, Montréal, Canada
d
Department of Psychology, Université Laval, Québec, Canada
e
Department of Psychology, Université du Québec à Montréal, Montréal, Canada

a r t i c l e i n f o a b s t r a c t

Article history: This study investigated different environmental and contextual factors associated with mal-
Received 4 March 2015 treated children’s adjustment in foster care. Participants included 83 children (52 boys),
Received in revised form 16 June 2015 ages 1–7 years, and their foster caregivers. Quality of interaction with the foster caregiver
Accepted 30 June 2015
was assessed from direct observation of a free-play situation; foster caregiver attachment
Available online 15 July 2015
state of mind and commitment toward the child were assessed using two interviews;
disruptive behavior symptoms were reported by foster caregivers. Results showed that
Keywords:
quality of interaction between foster caregivers and children were associated with behav-
Maltreated children
ior problems, such that higher-quality interactions were related to fewer externalizing and
Foster care
Behavior problems internalizing problems. Foster caregivers’ state of mind and commitment were interrelated
Relationship quality but not directly associated with behavior problems of foster children. Type of placement
Representations moderated the association between foster caregiver commitment and foster child behavior
problems. Whereas greater foster caregiver commitment was associated with higher lev-
els of adjustment for children in foster families (kin and non-kin), this was not the case in
foster-to-adopt families. Finally, the associations between foster child behavior problems
and history of maltreatment and placement related-risk conditions fell below significance
after considering child age and quality of interaction with the foster caregiver. Findings
underscore the crucial contribution of the foster caregiver–child relationship to fostering
child adjustment and, thereby, have important implications for clinical services offered to
this population.
© 2015 Elsevier Ltd. All rights reserved.

Introduction

Evidence generally supports the notion that children placed in foster care are at heightened risk of experiencing behavior
problems. Indeed, prior research has suggested that the behavior problems of foster children are linked to past experiences

夽 This research was supported by grants received from the Fonds de recherche du Québec–Société et culture (FRQ-SC).
∗ Corresponding author.

http://dx.doi.org/10.1016/j.chiabu.2015.06.009
0145-2134/© 2015 Elsevier Ltd. All rights reserved.
120 K. Dubois-Comtois et al. / Child Abuse & Neglect 48 (2015) 119–130

of maltreatment, higher number of placements, and late age placement such as in the preschool and school-age period
(Newton, Litrownik, & Landsverk, 2000; Oosterman, Schuengel, Slot, Bullens, & Doreleijers, 2007; Taussig, 2002). However,
less is known about the characteristics of foster caregivers that are associated with behavior problems exhibited by their
foster children. As suggested by Chamberlain (1996), foster caregiver behaviors may serve to maintain, exacerbate, or reduce
behavior problems in foster children. Studies are needed to better understand what characteristics of foster caregivers act
as protective or risk factors for the development of behavior problems in foster children. Although numerous studies have
shown that relationship quality with the parent is an important predictor of child adjustment in various spheres (e.g., Dubois-
Comtois, Moss, Cyr, & Pascuzzo, 2013; Fearon, Bakermans-Kranenburg, van IJzendoorn, Lapsley, & Roisman, 2010; Groh,
Roisman, van IJzendoorn, Bakermans-Kranenburg, & Fearon, 2012), studies testing this association among foster families
are sparse. The way in which interactions between foster caregivers and children are related to behavior problems is unclear.
Studies have shown that types of foster family and foster parent motivation to care for placed children are associated with
differences in child adjustment trajectories as well as with differences in the success of fostering (Buehler, Cox, & Cuddeback,
2003; Rosenthal & Curiel, 2006). Less is known about foster caregiver mental representations and mother–child interactive
quality. Accordingly, the objective of the current study is twofold: (1) to evaluate the association between foster mother
representations, quality of the relationship with the foster child, and foster child behavior problems after controlling for
previous experiences of maltreatment; and (2) to examine type of placement as a potential risk factor in the association
between foster mother representations or quality of the relationship with the child, and foster child behavior problems.
According to the U.S. Department of Health and Human Services (2015), about 1% of children were abused or neglected
in the US in 2013. Neglect was the most frequent form of maltreatment (79.5%), followed by physical abuse (18%), sexual
abuse (9%) and psychological maltreatment (8.7%). The examination of US child protection narratives suggests that 36–91%
of children experience multiple types of maltreatment (Gilbert et al., 2009). In addition, between 21% and 42% of children
with substantiated maltreatment were re-reported to Child Protection Services (CPS) agency within 18 months and recur-
rence of maltreatment was between 6% and 24% (Casanueva et al., 2015). Recent data provided by the Quebec Association
for Child Protection (QACP, 2013) showed that 80,540 children were investigated for maltreatment by the Québec child
protection system in 2012–2013, of which 31.8% were between 0 and 5 years of age. For this group of children, substantiated
maltreatment-related investigations included physical abuse (12.9%), physical, material and health-related neglect (65.1%),
psychological and emotional abuse (18.4%), and sexual abuse (2.3%). Fifty-three percent of these children were placed in
out-of-home care (QACP, 2013).
Child maltreatment, which generally precedes child out-of-home placement, is considered to be an important risk factor
for child behavioral, emotional, and academic maladjustment (Cicchetti & Toth, 2005). Accordingly, it has been found that
children in foster care (aged 0–17 years) show rates of mental health and behavior problems 2.5 times higher than those
of children from at-risk populations with no history of abuse (Clausen, Landsverk, Ganger, Chadwick, & Litrownik, 1998;
dosReis, Magno Zito, & Safer Soeken, 2001; Zima et al., 2000). Others have reported that the levels of behavioral problems
presented by foster children (aged 0–12 years) are similar to those of abused children living with their biological parents
(Lawrence, Carlson, & Egeland, 2006; Mennen, Brensilver, & Trickett, 2010). In addition, children with disabilities, who have
experienced substantiated maltreatment, are about twice as likely to be in out-of-home placements, than are children (aged
0–18 years) with substantiated maltreatment without disabilities (Lightfoot, Hill, & LaLiberte, 2011). Taken as a whole, these
results clearly suggest that neglect and harsh experiences before placement are detrimental for children. Therefore, because
out-of-home placement is intended to promote child wellbeing through the development of a healthier caregiver–child
relationship, it is important not only to consider the background of children in foster care in explaining (mal) adjustment,
but also experiences of care with the foster parent.

Quality of Interactions between Foster Mother and Child

Bowlby (1969/1982) theorized that differences in the quality of early relationships with caregivers have a long-term
impact on child adjustment and mental health, given that parent-child interactions shape children’s internal working models
of self and relationships with others. Internal working models are defined as mental representations of self, caregivers, and
others that include expectations about caregivers’ responsiveness and child’s self-worth and strategies to seek comfort
(Bretherton, 1990). Mother–child interaction is a mutual process during which the mother guides and supports the child
in exploring affects and thoughts, thus assisting in the organization of child emotional experiences, and by extension,
the development of emotional and social abilities (Thompson, 2008). Studies have shown that children who experienced
sensitive care and harmonious interactions with their mothers are more likely to be socially adapted and use more effective
emotion regulation strategies than their peers who experienced less supportive mother–child interactions (e.g., Fearon
et al., 2010; Groh et al., 2012; NICHD, 2006). Evidence from foster care research appears to suggest a similar phenomenon.
A study by Oosterman and Schuengel (2008) showed that the interactive sensitivity of foster mothers, during a 15-minute
semi-structured interaction with children aged 2–7 years, was negatively associated with teacher-reported externalizing
problems in kindergarten or daycare. Another study, conducted with a sample of early adolescents in foster care (aged 13
years), showed that higher levels of positive family communication in the foster home were associated with lower levels of
adolescents’ externalizing and internalizing behavior problems (Vuchinich, Ozretich, Pratt, & Kneedler, 2001). In addition,
higher levels of negative family communication were correlated with greater child behavior problems. These two studies
 K. Dubois-Comtois et al. / Child Abuse & Neglect 48 (2015) 119–130 121

highlight the important role of parent-child interactions in foster families; however, in both cases, history of maltreatment
or other characteristics of foster caregivers were not considered.

Foster Mother Attachment State of Mind and Commitment

Parents mental representations, particularly those pertaining to relationships and experiences of traumatic events, are
likely to influence caregiving behaviors and thus to have an impact on child adjustment (Madigan et al., 2006; Van IJzendoorn,
1995). Attachment states of mind refer to the way in which adults process thoughts and feelings regarding their own
attachment experiences (Hesse, 2008). A secure-autonomous state of mind is indicated by a coherent and credible discourse
regarding one’s relationships with parents, along with valuing of attachment relationships. As parents, these adults are
most likely to be sensitive and responsive to their children’s needs (Pederson, Gleason, Moran, & Bento, 1998). Insecure-non
autonomous states of mind (dismissing or preoccupied) are characterized by incoherence and inconsistencies in the recall of
attachment-related experiences (for more details, see Main & Goldwyn, 1998). On the one hand, individuals with a dismissing
state of mind tend to minimize or even denigrate the importance of attachment relationships. They have few or no memories
of childhood relational experiences and while negative situations and their consequences are generally denied or minimized,
attachment figures are idealized. On the other hand, the discourse of individuals with a preoccupied attachment state of
mind is filled with uncertainty, anger, and self-blame. These individuals continue to experience great frustration concerning
former and present attachment experiences. Finally, adults can be assigned to a fourth category (unresolved/disoriented) if
their discourse regarding experiences of loss or abuse suggests some disorganization or disorientation of thought processes
(Ballen, Bernier, Moss, Tarabulsy, & St-Laurent, 2010). Evidence for such a disorganized response includes lapses in the
monitoring of discourse (e.g., failing to finish a sentence, becoming silent during the elaboration of an idea) or reasoning
(e.g., feeling responsible for being abused, feeling in contact with the deceased person).
Associations between parental state of mind and child adjustment have been assessed in a number of investigations.
DeKlyen (1996) found that insecure attachment states of mind are related to behavior problems in clinic-referred children.
Others found that adolescent mothers’ unresolved state of mind regarding attachment is related to disrupted maternal
behaviors and toddlers’ externalizing behavior problems (Madigan, Moran, Schuengel, Pederson, & Otten, 2007). In Zajac
and Kobak’s study (2009), maternal unresolved state of mind with respect to loss predicted teacher-reported behavior
problems in children aged 6 to 13 years.
In the foster care population, foster mother attachment state of mind has begun to be investigated. Bates and Dozier
(2002) found proportions of attachment states of mind comparable to that in the general population, although there was
a trend showing fewer secure-autonomous and preoccupied foster mothers, and more dismissing and unresolved foster
mothers. When children are placed with insecure foster mothers, they are more likely to show disorganized attachment
behaviors (Dozier, Stoval, Albus, & Bates, 2001) – a form of attachment that is a significant risk factor for child maladjustment
in the short- and long-term (Fearon et al., 2010; Groh et al., 2012). In line with these results, a recent study showed that
foster caregivers’ insecure states of mind were associated with increased atypical parenting while interacting with the foster
child (Ballen et al., 2010). These results suggest that foster caregiver insecure state of mind might constitute a risk factor
in the foster family that could contribute to child dysregulation. To our knowledge, however, no study has examined the
association between foster mother attachment state of mind and foster child behavior problems.
Another aspect of mental representations in adults that has an impact on foster child adjustment and is associated with
caregiving is the level of parental commitment toward the child. This concept is defined as the extent to which the caregiver
is motivated to have an enduring relationship with a particular child (Dozier & Lindhiem, 2006). Because foster children
have often experienced abandonment and lack of care, and because there is a certain degree of uncertainty as to whether the
child will stay in the family for an extended period of time, this concept is highly relevant for this population. A number of
studies by Dozier and her colleagues have highlighted the importance of foster caregiver commitment for child adjustment,
as well as related child and foster family characteristics. Foster mothers who are more committed have foster children with
fewer behavior problems (Lindhiem & Dozier, 2007), and those who are more accepting of their children have foster children
who develop more positive self-representations and effective solutions to separation scenarios than children whose foster
mothers are less accepting (Ackerman & Dozier, 2005). In addition, foster mothers with a secure-autonomous attachment
state of mind are more accepting of early-placed children than later-placed children (Bates & Dozier, 2002). Finally, higher
levels of foster mother commitment is found when children are placed at a younger age and among caregivers who have
fostered fewer children (Dozier & Lindhiem, 2006), suggesting that the type of foster family could impact on caregiver
commitment.
In summary, past studies have primarily examined the roles of foster placement history, including past experiences of
maltreatment, placement disruptions, and age at placement, as predictors of foster child adjustment. Only a few recent
studies have assessed the association between foster caregivers’ characteristics and foster child adjustment, but most have
overlooked the role played by past experiences of maltreatment and placement. This is an important limitation that could
lead to misinterpretation of results. One might expect that the relationship between a foster child, being a victim of past
maltreatment experiences, and a foster parent is likely to be problematic. Children form new relationships on the basis of
past relational experiences. However, foster children’s adjustment is also likely to be influenced by the characteristics of
new parental figures. This assumption is at the foundation of all foster care programs, as placement with a new parental
figure is essentially seen as a solution to promote child well-being. In order to better understand the role played by the
122 K. Dubois-Comtois et al. / Child Abuse & Neglect 48 (2015) 119–130

foster environment in foster child adjustment, it is imperative that studies examine the link between foster child behavior
problems and foster caregiver characteristics, while taking into account past experiences.
The central aim of this study is therefore to examine the association between foster mother attachment state of mind
and commitment, foster mother–child interactive quality, and behavior problems among foster children, while considering
child history of maltreatment and placement. First, we hypothesize that foster caregivers’ unresolved and insecure states of
mind (dismissing and preoccupied), low levels of commitment, and poor quality of interactions will be inter-related. Second,
it is expected that higher levels of child externalizing and internalizing problems, as reported by the foster caregiver, will
be associated with a greater number of past experiences of maltreatment and placement. We also predict that, over and
above past experiences of maltreatment and placement, higher levels of foster child behavior problems will be associated
with (a) foster caregivers’ unresolved and insecure states of mind, (b) lower levels of commitment, and (c) poorer quality
of interactions. Finally, because the foster-to-adopt families are in many ways different from regular foster families (Neil,
Beek, & Schofield, 2003), we will examine whether type of foster family moderates the association between foster child
adjustment and foster caregiver representations and interactive quality with the child.

Method

Participants

Participants in this study were 83 foster children (52 boys) and their foster mothers who were recruited through CPS
in the province of Québec (Canada) to participate in a study evaluating the efficacy of a home-visiting relationship-based
intervention program. In the province of Québec, foster care is provided within a private home approved by child welfare
services. Whether the child is placed with unknown adults (regular foster or foster-to-adopt families) or with relatives
(kinship care), parental figures are responsible for the child. In particular, a foster-to-adopt program consists of placing
children who are considered at a high risk of abandonment in foster families wishing to adopt them (Pagé, Piché, Ouellette,
& Poirier, 2008). CPS agencies were asked to refer foster families who: (1) were fostering a child between 12 and 84 months
of age; (2) had been fostering the target child for at least two months; and (3) were engaged in long-term foster care or in
eventual child adoption. Identification of families was accomplished through liaison with an employee of the CPS who had
legal access to family records.
Children were on average 61.5 months of age (SD = 21.17, range from 12 to 84 months of age) and were placed in their
foster home for 20.37 months on average (SD = 11.65, range from 2 to 49 months). Foster mothers ranged in age from 25 to
63 years (M = 40.70 years, SD = 8.04). Nineteen percent of foster families had an income (in Canadian dollars) under 40,000$,
39% earned between 40,000$ and 75,000$, and 42% earned 75,000$ or more. Eighty-seven percent of foster families were
composed of two parents and 72% of foster mothers had completed a postsecondary education level. Twenty-six percent
of the sample were regular foster families, 10% kinship foster families, and 64% foster-to-adopt families. Foster-to-adopt
caregivers were significantly younger (M = 38.30 years, SD = 5.91) than caregivers of kinship (M = 49 years, SD = 11.51; Tukey
HSD, p = .001) and regular non-kinship (M = 43.63 years, SD = 8.31; Tukey HSD, p = .04) foster families (F[2,68] = 9.14, p < .001).
Foster-to-adopt caregivers were also more likely to have completed a postsecondary education level (86%; z = 3.7) than
caregivers of kinship (37.5%; z = 2.3) and regular non-kinship (50%; z = 2.5) foster families (2 [2] = 14.32, p < .001). Family
income and marital status did not differ according to foster family type.

Measures and Procedure

CPS employees, who were informed of the research purposes, initially approached eligible families and obtained their
consent to transfer personal information to a research assistant. Those who accepted were contacted by research assistants
who explained the research purposes and procedure. Written consent to participate was obtained from foster and biological
mothers. Foster caregivers consented to their own participation, and biological parents consented to children’s participation.
The protocol was approved by the institutions’ ethics committees. Participating families completed initial measures during
two one-hour visits. The first visit, conducted by trained research assistants, took place in the foster home, where foster
mother–child interactive quality was assessed through a 10-min video-recorded free play, and foster mothers completed
different questionnaires. The second visit was conducted either at home or in a lab setting, and involved the administration
of two interviews to foster mothers, that is, the This is my Baby interview (Bates & Dozier, 1998) and the Adult Attachment
Projective Picture System (George & West, 2012).
Foster caregiver commitment. The “This is my Baby” (TIMB; Bates & Dozier, 1998) interview is a semi-structured
interview designed for foster mothers who have a foster child living with them for at least two months. Ten standardized
questions act as probes for foster mothers’ feelings and attitudes concerning their foster child’s personality, their level of
commitment to raising the child, and how their relationship with this child had influenced and would continue to influence
them in the future. Interviews were rated on three Likert scales – acceptance, commitment, and belief in influence – ranging
from 1 to 5 with midpoints allowed (e.g., 1.5), where 1 indicated low commitment and 5, high commitment. Foster mothers
categorized as poorly committed did not try to form an affective bond with their foster child, showed no signs of psychological
adoption, and did not seem to mind whether or not the child had to stay or leave their care. Conversely, foster mothers
 K. Dubois-Comtois et al. / Child Abuse & Neglect 48 (2015) 119–130 123

categorized as highly committed considered their foster child as their own, and reported having a strong emotional bond
with the child (Dozier & Lindhiem, 2006).
Validity and test-retest reliability of the interview has been established in multiple studies (Ackerman & Dozier, 2005;
Dozier & Lindhiem, 2006; Lindhiem & Dozier, 2007). Using a transcription of the interviews, two independent coders who
were in contact with O. Lindhiem coded the interviews. Interrater reliability (intraclass correlations) calculated on 30% of
sample cases ranged from .72 to .87. Because the three scales were highly intercorrelated (r between .70 and .85), they were
averaged and combined into a mean score of foster mother commitment.
Attachment state of mind. The Adult Attachment Projective Picture System (AAP; George & West, 2012) is a
semi-structured interview that allows the classification of adults into one of the following attachment categories: secure-
autonomous, dismissing, preoccupied or unresolved. Eight pictures depicting attachment-related situations (separation, fear
or solitude) are designed to activate attachment representations. Foster mothers were asked to create stories based on the
pictures, and to comment on the emotions or thoughts of the characters. To assign a classification group, the attachment sto-
ries were coded according to three dimensions: story content, discourse, and defensive processes. The coding patterns were
then used to assign an attachment classification group to each participant. For a detailed description of the four classifications
see George and West (2012).
Validity of the AAP was established through strong convergence with the Adult Attachment Interview (AAI) (George &
West, 2001, 2011) and in studies from independent investigators (e.g., Béliveau & Moss, 2009). The instrument has also
been used with adolescents in foster care (Webster & Joubert, 2011). Protocols were coded by two independent coders who
completed training and a reliability trial with the developers of the AAP. Agreement for the 4-way classifications, calculated
on 30% of sample cases, was 88% (k = 0.81).
Quality of interaction. Foster mothers and children were asked to play together for 10 min. No specific instructions were
given to the dyads, and they were allowed to use the material that was brought by the research assistants (craft, blocks,
etc.) or to use their own toys. Video-recordings of interactions were evaluated using the Parent–Child Interaction Scale
for the Preschool and School Periods (Moss, Rousseau, Parent, St-Laurent, & Saintonge, 1998). The instrument consists of an
overall rating (from high quality [i.e., harmonious, reciprocal] to poor quality [i.e., indifferent or conflictual]) and eight 7-point
subscales (Coordination, Communication, Partner Roles, Emotional Expression, Responsivity/Sensitivity, Tension, Mood, and
Enjoyment) that are used to capture global aspects of parent-child interactive quality, with higher scores considered more
optimal.
These scales have been shown to distinguish mother–child dyads characterized by different attachment classifications,
and have shown concurrent and longitudinal associations with behavior problem ratings and other measures of quality
of parent–child interactions (e.g., Dubois-Comtois & Moss, 2004; Dubois-Comtois et al., 2013). Free-play videotapes were
evenly divided among three observers (blind with respect to other family information). Interrater reliability (intraclass
correlations) calculated on 30% of sample cases ranged from .65 to .82 for subscale ratings and from .77 to .84 for the overall
rating. Because a principal component factor analysis of the nine mother–child interaction scales yielded a single factor
explaining 81% of the variance, only the overall scale score representing reciprocated, balanced, and open communication
(at the high end) was included in data analyses.
Child behavior problems. Foster mothers completed the Child Behavior Checklist (CBCL), using either the version for
children 1–5 years of age (Achenbach & Rescorla, 2000) or that for children aged 6–18 years (Achenbach & Edelbrock, 1983),
which both consist of 113 items scored on a 3-point scale (not true, somewhat or sometimes true, very true or often true) that
are age relevant. The CBCL generates scores on two major scales: the internalizing scale includes behaviors of withdrawal,
somatic complaints, anxiety, and depression, whereas the externalizing scale includes inattention, delinquent and aggressive
behaviors. The computed raw scores are converted to gender- and age-standardized scores. The instrument has excellent
psychometric qualities, with high internal consistency and strong one-year test–retest reliability (Achenbach & Edelbrock,
1983; Achenbach & Rescorla, 2000; Ballespí, Claustre, & Dolors, 2013; Heubeck, 2000). Excellent Cronbach’s alpha scores
for the externalizing and internalizing scales were found for data of the current study (0.94 and 0.88, respectively). In our
sample, behavior problems in the clinical and borderline range were, for externalizing problems, 16.9% and 3.5%, and for
internalizing problems, 8.4% and 4.9%, respectively.
History of maltreatment/placement risk index. A risk index composed of past experiences of maltreatment and place-
ment related-risk conditions was calculated using five factors identified in the literature as posing a threat to foster child
development and adjustment. This index was similar to that used by Dozier and Lindhiem (2006). Information was retrieved
from each child CPS record. The history of placement risk index was computed by giving one point to the occurrence of
any of the following indices: physical abuse, sexual abuse, neglect, five or more placements (see Newton et al., 2000),
placement after 3 years of age (see Oosterman et al., 2007). Higher scores on the history of maltreatment/placement risk
index indicate greater vulnerability. Child neglect was largely predominant in this sample: 71 children were neglected,
12 were victims of physical abuse, and 5 were victims of sexual abuse. Ten children in the sample were placed in foster
care after 3 years of age or experienced at least five placements. Overall, 14.5% of children presented with no risk factor,
57% one factor, 17% two factors, 8.5% three factors, and 3.5% four risk factors. Because of the small number of children
in the four risk factors group, they were combined with the three factors group. Therefore, 12% of the current sample
experienced three or more risk factors prior to current placement. No children in the sample had experienced all five risk
factors. Children showing no risk factors were abandoned by their biological parents at birth and placed in a foster-to-adopt
family.
124 K. Dubois-Comtois et al. / Child Abuse & Neglect 48 (2015) 119–130

Table 1
Descriptive statistics for all main variables under study.

Variable Mean Standard deviation Observed range

CBCL externalizing problems (T scores) 54.59 11.97 30–89

CBCL internalizing problems (T scores) 51.04 9.90 30–87
Foster mother commitment 2.86 .83 1–4.5
Quality of interactions 4.04 .98 1–6
History of placement risk index 1.30 .95 0–3

n %

Foster mother attachment state of mind

Secure 25 30.1
Dismissing 8 9.6
Preoccupied 14 16.9
Unresolved 36 43.4

Results

Missing data. To maximize the sample size, we included cases with missing values in the analyses by estimating missing
data. Twenty-six CBCLs were missing or not filled correctly, and 11 interactions and three TIMB’s were missing due to
equipment malfunction or parents refusing to be filmed. Analysis of the existing dataset with Little’s Missing Completely
at Random (MCAR) Test indicated that the data were missing completely at random, 2 (9) = 14.51, ns. Missing data were
therefore replaced by means of Expectation Maximization (Tabachnick & Fidell, 2001). Analyses conducted with and without
imputing missing data revealed similar results.
Preliminary analyses. Table 1 presents the descriptive statistics for the study variables. Scores on study variables were
normally distributed. Analyses were undertaken to identify possible covariates related to the dependent variables, that
is, externalizing and internalizing behavior problems. T-tests and analyses of variance (ANOVAs) revealed no significant
differences in behavior problems as a function of child gender or foster mother’s education (secondary vs. postsecondary
education), marital status or income, ts between 0.16 and 1.02, ns and F between 0.82 and 0.88, ns. Correlations revealed
significant positive associations between child age and externalizing and internalizing behavior problems, r = .41, p < .001
and r = .33, p < .001, respectively, but non-significant associations with foster mother’s age or time spent in the current foster
family, r between .14 and .20, ns. The history of maltreatment/placement risk index was significantly correlated with child
externalizing, r = .22, p = .049, but not with internalizing behavior problems, r = .17, ns.
ANOVAs showed significant differences on externalizing and internalizing (marginal) behavior problems as a function of
type of placement, F(2,82) = 3.20, p = .046 and F(2,82) = 2.53, p = .086, respectively. Orthogonal comparisons revealed that chil-
dren in foster-to-adopt families showed fewer externalizing (M = 52.20, SD = 11.03) and internalizing (M = 49.24, SD = 9.41)
behavior problems than those in foster care families (kin and non-kin), t(80) = 2.47 p = .016, and t(80) = 2.12 p = .037, respec-
tively. Within the foster care families, no differences were found between children in kinship (for externalizing: M = 60.50,
SD = 13.89; for internalizing: M = 54.63, SD = 11.97) and non-kinship families (for externalizing: M = 58.20, SD = 12.36; for
internalizing: M = 54.08, SD = 8.18) on externalizing and internalizing behavior problems, t(80) = .48, ns and t(80) = .14, respec-
tively. Because type of placement will be examined as a potential moderator, it will not be included as a covariate. Overall,
subsequent analyses on behavior problems will be covaried only with child age and history of maltreatment/placement,
both variables being strongly correlated, r = .60, p = .001.

Interrelations among Independent Variables

Correlation analyses revealed that quality of interaction between foster mother and child was positively related to foster
mother commitment, r = .28, p = .011. To examine the association between foster caregivers’ attachment state of mind and
variables of foster caregiver commitment and quality of interaction with the foster child, we performed two one-way analyses
of variance (ANOVAs). A set of a priori simple contrasts comparing each insecure group to the secure group were also
performed, consistent with our hypotheses: Dismissing versus secure; preoccupied versus secure; and unresolved versus
secure.
Results of the ANOVA on quality of foster mother–child interaction revealed no significant omnibus effect,
F(3,82) = 1.84, ns. Despite this non-significant overall finding, we pursued with the analyses of planned comparisons, because
these were a priori hypothesis-driven. Results showed that preoccupied foster mothers experienced lower quality of inter-
actions with their foster child (M = 3.64, SD = .93) than secure-autonomous mothers (M = 4.35, SD = .75), t(79) = 2.18, p = .033.
Interactive quality of dismissing and unresolved foster mothers (M = 4.22, SD = 1.59 and M = 3.94, SD = .95, respectively)
resembled that of secure foster mothers, t(79) = .31, ns and t(79) = 1.60, ns, respectively.
Results of the ANOVA on foster mother commitment revealed a marginal omnibus effect, F(3,82) = 2.17, p = .098. Planned
comparisons showed that preoccupied and unresolved foster mothers had significantly lower levels of commitment toward
children (M = 2.64, SD = .71 and M = 2.74, SD = .87, respectively) than secure-autonomous foster mothers (M = 3.20, SD = .72),
 K. Dubois-Comtois et al. / Child Abuse & Neglect 48 (2015) 119–130 125

Table 2
Multiple regressions testing the association between study variables and child behavior problems.

Variable Externalizing problems Internalizing problems

Model 1 Model 2 Model 1 Model 2

B SE B ˇ B SE B ˇ B SE B ˇ B SE B ˇ

Child age .25 .07 .44** .23 .07 .40** .17 .06 .37** .16 .06 .33*
History of placement risk index −.77 1.81 −.06 .30 1.77 .02 −.69 1.55 −.06 −.13 1.54 −.01
Attachment state of mind −1.55 1.42 −.11 .01 1.25 .01
Foster mother commitment −1.92 1.56 −.13 −1.14 1.36 −.10
Quality of interactions −3.57 1.25 −.29** −2.63 1.09 −.26*

Adjusted R2 .15 .24 .09 .15

F for change in R2 8.08** 4.18** 4.96** 2.86*
*
p < .05.
**
p < .01.

Table 3
Multiple regression models testing the moderating role of foster family type on the relation between foster environment characteristics and child exter-
nalizing problems.

Variable Model 1 Model 2 Model 3

B SE B ˇ B SE B ˇ B SE B ˇ

Child age .01 .01 .43** .01 .01 .39** .01 .01 .39**
History of placement risk index −.06 .10 −.08 −.02 .10 −.02 −.01 .10 −.01
Foster mother commitment −.02 .01 −.17 −.10 .04 −.91*
Type of foster family −.02 .02 −.08 −.01 .02 −.06
Commitment × foster family .04 .02 .76*

Adjusted R2 .13 .14 .17

F for change in R2 6.95** 1.71 4.04*
*
p < .05.
**
p < .01.

t(79) = 2.05, p = .043 and t(79) = 2.14, p = .035, respectively. Dismissing mothers’ level of commitment (M = 2.67, SD = .71) did
not differ from that of secure-autonomous mothers, t(79) = 1.61, ns.

Prediction of Foster Child Behavior Problems

In order to examine the contribution of foster mother state of mind, commitment toward the child, and interactive
quality to externalizing and internalizing behavior problems, two multiple regression analyses were performed, one for each
dependent variable. Child age and history of maltreatment/placement were entered at step 1, followed by foster mother
attachment state of mind, commitment, interactive quality, at step 2. Attachment state of mind was coded as a continuous
variable, resulting in a score ranging from 1 to 3, with secure-autonomous individuals obtaining the highest score, insecure
not unresolved individuals (dismissing and preoccupied) the middle score, and unresolved individuals the lowest score.
Results of the two regression analyses are presented in Table 2.
Analyses of externalizing problems and internalizing problems revealed that in both cases, child age and quality of
interactions with the foster mother were significant predictors (see Table 2). In contrast, foster mother state of mind and
commitment toward the child, and child history of placement and maltreatment did not significantly contributed unique
variance. The final models predicted 24% of the variance in externalizing problems and 15% of the variance in internalizing
problems.

Foster Family Type as a Moderator of the Association between Foster Caregiver Characteristics and Child Behavior Problems

Hierarchical multiple regression analyses were performed to examine if foster family type moderated the association
between foster caregiver characteristics and child behavior problems. Six models were tested, one for each combination of
dependent and independent variables, with predictors entered in the following order: step 1 included child age and history
of maltreatment/placement; in step 2, one of the following independent variable (regression 1: foster caregiver’s attachment
state of mind; regression 2: level of commitment; and regression 3: quality of interactions with the foster child) and type
of foster families (foster-to-adopt vs foster care) were entered; finally, step 3 included the interaction term Independent
variable X Foster family type (using variables that were centered).
Among the six regression models, two yielded significant results, and are presented in Tables 3 and 4. Analyses of external-
izing (see Table 3) and internalizing problems (see Table 4) revealed that the interaction between foster mother commitment
and foster family type was a significant predictor, above all other factors, thereby showing the moderating role of foster
126 K. Dubois-Comtois et al. / Child Abuse & Neglect 48 (2015) 119–130

Table 4
Multiple regression models testing the moderating role of foster family type on the relation between foster environment characteristics and child inter-
nalizing problems.

Variable Model 1 Model 2 Model 3

B SE B ˇ B SE B ˇ B SE B ˇ

Child age .01 .01 .38** .01 .01 .34** .01 .01 .33**
History of placement risk index −.05 .09 −.07 −.01 .10 −.02 −.01 .09 −.01
Foster mother commitment −.02 .01 −.15 −.10 .04 −1.04**
Type of foster family −.02 .02 −.09 −.01 .02 −.07
Commitment × foster family .04 .02 .91*

Adjusted R2 .09 .10 .15

F for change in R2 5.14** 1.39 5.60*
*
p < .05.
**
p < .01.

family type. The final model predicted 17% of the variance in externalizing problems and 15% of the variance in internaliz-
ing problems. Post hoc analyses revealed that for foster-to-adopt families, there was no association between foster mother
commitment and child externalizing (r = −.05, ns) and internalizing problems (r = −.05, ns), after controlling for child age and
history of placement. For foster families (kin and non-kin), commitment was negatively associated with child externalizing
(r = −.46, p = .015) and internalizing problems (r = −.52, p = .005) even after controlling for child age and history of maltreat-
ment/placement. Commitment does not differ according to foster family type, t(81) = .71, ns (for foster-to-adopt families:
M = 2.91, SD = .88; for foster families: M = 2.77, SD = .75).
Foster family type did not moderate the association between foster child behavior problem and foster mother state of mind
(externalizing: F[1,77] = .06, ns; internalizing: F[1,77] = .12, ns) or quality of interaction with the child (externalizing:
F[1,77] = 1.88, ns; internalizing: F[1,77] = 1.67, ns).

Discussion

The objective of this study was to examine the association between behavior problems of foster children and character-
istics of the foster caregiver, including foster caregivers’ state of mind, commitment and interactive quality with the child.
This study also took into consideration the impact of history of maltreatment and risk conditions associated with out-of-
home placement. Results showed that foster caregivers’ representations, namely attachment state of mind and commitment,
were interrelated but not associated with behavior problems of foster children overall. Foster caregiver commitment and
child behavior problems were related, however only for children in foster homes (kin and non-kin), not in foster-to-adopt
families. Quality of interaction between foster caregivers and children was associated with behavior problems, such that
higher-quality interactions were related to fewer externalizing and internalizing problems. These results have implications
for services offered to foster caregivers and children, and they are discussed in the following sections.

Quality of Interactions between Foster Caregivers and Children

The results of this study showed that quality of foster mother–child interaction is an important variable associated
with foster child adjustment in early childhood. These findings support those of past studies conducted with normative (e.g.,
Dubois-Comtois et al., 2013), at-risk (e.g., Easterbrooks, Bureau, & Lyons-Ruth, 2012), and foster care populations (Oosterman
& Schuengel, 2008). According to the foster caregivers of this study, children placed in foster families, where caregivers
were supportive and interactions were harmonious and pleasant, were more socially adjusted. Conversely, foster caregivers
reported higher levels of child externalizing and internalizing problems when their interactions were more unbalanced and
chaotic. Hence, when caregivers fail to provide adequate emotional support to foster children, the latter learn not to rely on
caregivers to regulate internal emotional states or to find adapted solutions to distress-eliciting problems. This situation is
likely to contribute to the development, maintenance, or exacerbation of internalizing and externalizing behavior problems.
In addition, this study suggests that when placed in a home with supportive and sensitive caregivers, children, despite harsh
living conditions prior to placement, may overcome difficulties and not show as high levels of externalizing and internalizing
problems. To confirm this hypothesis, however, longitudinal prospective studies are needed. It is also possible that foster
children’s behavior problems lead to difficulty in the foster mother–child relationship. Still, regardless of the direction of the
association, our results highlight the crucial role of the foster mother–child relationship in relation to foster child adjustment,
and support the use of parent-child intervention strategies in modifying the trajectory of foster children with behavioral or
emotional difficulties (e.g., Dozier, Zeanah, & Bernard, 2013; Fisher & Stoolmiller, 2008).

Foster Caregivers’ Mental Representations

The results of this study also showed that foster caregivers with a preoccupied state of mind had more problematic inter-
actions with their foster children than did caregivers with a secure state of mind, while unresolved and dismissing caregivers
 K. Dubois-Comtois et al. / Child Abuse & Neglect 48 (2015) 119–130 127

showed interactive quality similar to that of secure caregivers. Similarly, Oyen, Landy, and Hilburn-Cobb (2000) found, in
an at-risk sample that, of all attachment groups, only preoccupied mothers differed significantly from secure-autonomous
mothers, and were the most insensitive during a 15-minute interaction with their child. Other studies also found links
between maternal preoccupation and increased angry, intrusive, and unsupportive behaviors toward the child during free-
play and structured tasks (Adam, Gunnar, & Tanaka, 2004; Bosquet & Egeland, 2001; Whipple, Bernier, & Mageau, 2011).
Because the play interaction used for this study was not intended to be stressful, some of the mothers with insecure attach-
ment representations may have been more available to their child than they would normally be under stress. Dismissing
mothers may function well in non-stressful playtime conditions (Oyen et al., 2000), although they may be insensitive when
their child needs assistance (Pederson et al., 1998). Similarly, unresolved individuals show lapses in their discourse and
disorientation in the face of trauma, but they can express themselves adequately when not in contact with traumatic events
(Hesse & Main, 2006). The task used in this study may therefore not have been stressful enough to elicit the insensitive and
frightening behaviors observed in caregivers with an unresolved state of mind in contexts such as extended home visits and
the Strange Situation procedure (see Hesse & Main, 2006). It is also possible that the relatively high SES of our sample or
other protective factors not taken into account in the present study (e.g., social support, mental health) may have prevented
unresolved foster mothers from showing poor quality interactions with the child.
Unexpectedly, we found a large proportion of foster caregivers with an unresolved state of mind (43%), compared to
18% in the general population (Bakermans-Kranenburg & van IJzendoorn, 2009). This meta-analysis also found that 32%
of parents in at-risk samples (e.g., low socioeconomic status, adolescent parenthood) and 43% of parents in the clinical
population have an unresolved state of mind with respect to loss or trauma. These proportions are similar to that observed
in the current study. According to Steele, Kaniuk, Hodges, Haworth, and Huss (1999), experiences of childhood adversity
often motivates individuals to foster or adopt children. Thus, unresolved experiences of loss or abuse may be more prevalent
among foster caregivers when compared with the normal population. It is possible also that this high unresolved prevalence
may be better accounted for by the use of the AAP instead of the AAI as the adult attachment measure. However, the
distributions of attachment classifications based on the AAP for different normative and clinical subsamples were found to
be similar to those documented for the AAI in Bakermans-Kranenburg et al.’s meta-analysis (Juen, Arnold, Meissner, Nolte, &
Buchheim, 2013). Moreover, in one of our previous paper with pilot data (half of the current sample), 46% of foster mothers
were classified unresolved on the AAI (Ballen et al., 2010). This evidence suggests that the attachment measure used in
the current study does not account for the overrepresentation of unresolved attachment. Hence, this overrepresentation
is clearly of concern and points to the importance of offering support to foster caregivers, whose past trauma may easily
be reactivated by foster children’s difficult life circumstances and behavior problems. It is, in fact, possible that the foster
mothers of this sample chose to participate in the study because of their difficulties and the opportunity to receive an
attachment-based intervention. These findings are nevertheless disquieting given that unresolved state of mind has been
associated with increased psychopathology and anomalous parenting behaviors (Bakermans-Kranenburg & van IJzendoorn,
2009; Madigan et al., 2006).
In addition, the present data indicate that foster parent commitment varies as a function of attachment representations.
Unresolved and preoccupied caregivers were less committed to their foster children than were secure mothers. These results
are the first to compare commitment levels of foster caregivers with different insecure attachment states of mind. A study
testing the predictive role of attachment state of mind on caregivers’ representations of their relationship with their infants
in biologically intact dyads showed that caregivers with insecure states of mind failed to reflect upon the internal experience
of their child and upon their own internal experience as a parent (Slade, Grienenberger, Bernbach, Levy, & Locker, 2005).
According to Slade et al. (2005), preoccupied individuals are overwhelmed by attachment-related representations to the
detriment of reflective functioning, and unresolved individuals are profoundly dysregulated when in contact with these
mental states. Thus, foster caregivers with these states of mind may have difficulty having a balanced view of their child and
respect for their individuality. Moreover, questioning caregivers about potential additional placement relocation of the foster
child might trigger unresolved memories of loss and trauma in certain individuals. As changing placement represents a real
possibility, which foster caregivers have to prepare for, it is important to consider how caregivers’ attachment representations
might influence this process, and adapt clinical support models accordingly.
In addition, our results indicated that foster caregivers who were less committed had interactions with their foster child
that were more problematic. These findings are in line with those of Dozier and colleagues showing that foster caregivers’
greater commitment toward their foster child was associated with more placement stability and better child adjustment
(Dozier & Lindhiem, 2006; Lindhiem & Dozier, 2007). Together with the results of the current study, this suggests that the
struggle for some foster caregivers to form an emotional connection with their foster child translates into the difficulty
of being sensitive, supportive and affectionate during playtime, which may represent a pathway through which caregiver
commitment impacts child adjustment.
Our results also revealed that commitment toward foster children is an important factor associated with behavioral
adjustment in foster families, however not in foster-to-adopt families. Lindhiem and Dozier (2007) also found an association
between commitment and child adjustment in a sample of regular foster families. Contrary to their counterparts in foster
families (kin and non-kin), children in foster-to-adopt families may have a greater sense of security as to placement stability
and of being a legitimate member of the foster family. Before placement, foster-to-adopt caregivers have explicitly expressed
their wish to adopt the child; making their commitment stronger and legislatively supported by the CPS organization. Another
factor that might explain the moderating effect found in this study is the fact that foster-to-adopt mothers were younger and
128 K. Dubois-Comtois et al. / Child Abuse & Neglect 48 (2015) 119–130

more educated than those in regular foster families. Low socio-economic status has been identified as a risk factor for the
development of child externalizing problems (see Campbell, Shaw, & Gilliom, 2000), although it was not related to behavior
problems in the present study. More studies are needed to disentangle the different factors that may account for the role
played by foster caregiver commitment on child adjustment in different types of foster families.

History of Maltreatment and Risk Conditions Associated with Out-of-home Placement

Finally, the results of this study revealed that history of maltreatment and placement related-risk conditions were asso-
ciated with foster child externalizing problems. These associations are consistent with results found in studies of maltreated
children living with their biological parents (see Cicchetti & Toth, 2005) or in foster care (e.g., Newton et al., 2000). However,
our study also showed that, after considering child age and quality of interaction with the foster caregiver, these associa-
tions were no longer significant. Similar results were found in another study of children (aged 7–11 years) in foster care
where quality of the foster environment, but not history of maltreatment or placement conditions, was associated with child
adaptation (Healey & Fisher, 2011). One may argue that after being placed for a more substantial period of time in foster
care, child adjustment may reflect the quality of the current environment rather than that of past experiences. Nevertheless,
despite the protection provided by a durable, safe and caring environment, older foster children are still at risk of showing
behavior problems as a result of the many challenges they need to face at each developmental transition and which can only
be exacerbated by early maltreatment history (Taussig, 2002). This issue is complex and future studies are needed to better
understand protective and risk factors associated with long-term adaptation of maltreated youth in foster care. Regarding
internalizing problems, such behaviors were not related to harsh living conditions prior to placement, which stands in con-
trast to results of other studies (see Cicchetti & Toth, 2005) and those found for externalizing problems. Tarren-Sweeney,
Hazell, and Carr (2004) found that foster parents and teachers may not be reliable informants of foster children’s internalizing
behavior problems. It has also been suggested that child evaluations by distressed mothers may reflect to a greater extent
their own personal difficulties than those of the child (Dubois-Comtois et al., 2013). Thus, internalizing problems in this
sample might have been underestimated, explaining the absence of significant links to history of maltreatment/placement.

Limitations

This study has several limitations that are important to note. First, there is a wide age range of children in our sample
that covers three developmental phases. Although the behavior problem measure used in the study has been validated for
such a wide range, one needs to consider problem behaviors from a developmental perspective. For instance, although it is
expected that toddlers present with higher levels of behavior problems, it is not developmentally expected for early school
age children. To better understand the differential impact of foster care on child behavior problem, it will be important, in
studies with more participants, to compare groups of children according to age. As of today, studies of foster children are
frequently composed of children aged 0–12 or 0–18 (e.g., Clausen et al., 1998; dosReis et al., 2001; Lawrence et al., 2006).
As a second limitation, the current sample is composed of more boys (63%) than girls. One should therefore be cautious in
interpreting results of this study. Still, because child gender was not associated with behavior problems in our sample, this
overrepresentation of boys should not affect the generalizability of our results. Somewhat smaller but similar proportions
were also found in a dataset combining data from child protection and the Canadian Census for the province of Québec, where
55.1% of placed children aged 0–9 years were boys. Third, our sample is composed of children who were predominantly
neglected. A different profile of results could be obtained from studies on physically or sexually abused children. It will be
important to replicate the study findings on a larger sample and to follow children over a period of time in order to assess
long-term foster children adaptation.

Summary and Clinical Implications

The objective of this study was to evaluate the association between foster mother representations, quality of the rela-
tionship with the foster child, and foster child behavior problems. We also explored type of placement as a potential risk
factor in the association between foster environment characteristics and foster child behavior problems. The significance of
our findings are strengthened by the inclusion of child history of maltreatment and placement related-risk conditions, thus
providing a more specific estimate of the characteristics of foster families associated with foster children’s (mal) adjustment.
The current study’s findings suggest that, in order to help reduce foster child behavior problems, intervention should
specifically target children experiencing poor quality of interactions with their foster caregiver and those living with care-
givers showing low levels of commitment. Parent-child interventions stand as methods of choice to modify the trajectories
of foster children with behavioral difficulties (e.g., Dozier et al., 2006; Dubois-Comtois, Cyr, Vandal, & Moss, 2012). These
interventions bring foster parents to recognize, interpret, and properly respond to foster child distress and cues, all of which
can be misleading owing to past experiences of inappropriate parental care prior to placement. In their interventions, prac-
titioners should focus on promoting foster caregivers’ sensitive behaviors and engagement toward the child and increase
positive and harmonious parent–child interactions in order to help children develop adequate emotion regulation strategies.
Results of our study not only inform on intervention targets that could promote foster child adaptation but they also
suggest important issues to assess with this population. For instance, our results point to the importance of screening
 K. Dubois-Comtois et al. / Child Abuse & Neglect 48 (2015) 119–130 129

for unresolved state of mind of future foster and adoptive caregivers in order to provide them with adequate support in
their crucial task of fostering an at-risk child. When assessing families, evaluators should pay special attention to flaws
in the organization of the adult’s discourse when addressing issues of loss and trauma, for example, temporal and spatial
incoherencies, signs of dissociation, prolonged pauses, and sources of profound anxiety, including moments when adults
blame themselves for the trauma. Practitioners should also help to promote foster parents’ engagement toward the child
not only through parent–child interventions but also by identifying and reducing factors that would refrain foster parents
from committing to the child. In that sense, helping foster parents to better cope with possible placement disruptions would
be noteworthy.

References

Achenbach, T. M., & Edelbrock, C. S. (1983). Manual for the child behavior checklist and revised child behavior profile. Burlington: Department of Psychiatry,
University of Vermont.
Achenbach, T. M., & Rescorla, L. A. (2000). Manual for the ASEBA preschool forms & profiles. Burlington, VT: University of Vermont.
Ackerman, J. P., & Dozier, M. (2005). The influence of foster parent investment on children’s representations of self and attachment figures. Journal of Applied
Developmental Psychology, 26(5), 507–520.
Adam, E. K., Gunnar, M. R., & Tanaka, A. (2004). Adult attachment, parent emotion, and observed parenting behavior: Mediator and moderator models. Child
Development, 75(1), 110–122.
Bakermans-Kranenburg, M. J., & van IJzendoorn, M. H. (2009). The first 10,000 Adult Attachment Interviews: Distributions of adult attachment represen-
tations in clinical and non-clinical groups. Attachment & Human Development, 11(3), 223–263.
Ballen, N., Bernier, A., Moss, E., Tarabulsy, G. M., & St-Laurent, D. (2010). Insecure attachment states of mind and atypical caregiving behavior among foster
mothers. Journal of Applied Developmental Psychology, 31(2), 118–125.
Ballespí, S., Claustre, J., & Dolors, R. (2013). Reliability and validity of a brief clinician-report scale for screening behavioral inhibition. Journal of Psychopathol-
ogy and Behavioral Assessment, 35, 321–334. http://dx.doi.org/10.1007/s10862-013-9344-7
Bates, B., & Dozier, M. (1998). This is my baby coding manual (Unpublished manuscript). University of Delaware, Newark.
Bates, B. C., & Dozier, M. (2002). The importance of maternal state of mind regarding attachment and infant age at placement to foster mothers’ represen-
tations of their foster infants. Infant Mental Health Journal, 23(4), 417–431.
Béliveau, M., & Moss, E. (2009). The role of stressful life-events on the intergenerational transmission of attachment. European Review of Applied Psychology,
59(1), 47–58.
Bosquet, M., & Egeland, B. (2001). Associations among maternal depressive symptomatology, state of mind and parent and child behaviors: Implications
for attachment-based interventions. Attachment & Human Development, 3(2), 173–199.
Bowlby, J. (1969/1982). Attachment and loss, Vol. 1: Attachment. New York: Basic Books.
Bretherton, I. (1990). Communication patterns, internal working models, and the intergenerational transmission of attachment relationship. Infant Mental
Health Journal, 11, 237–252.
Buehler, C., Cox, M., & Cuddeback, G. (2003). Foster parents’ perceptions of factors that promote or inhibit successful fostering. Qualitative Social Work:
Research and Practice, 2(1), 61–83. http://dx.doi.org/10.1177/1473325003002001281
Campbell, S. B., Shaw, D. S., & Gilliom, M. (2000). Early externalizing behavior problems: Toddlers and preschoolers at risk for later maladjustment.
Development and Psychopathology, 12(03), 467–488.
Casanueva, C., Tueller, S., Dolan, M., Testa, M., Smith, K., & Day, O. (2015). Examining predictors of re-reports and recurrence of child maltreatment using
two national data sources. Children and Youth Services Review, 48, 1–13.
Chamberlain, P. (1996). Intensified foster care: Multiple level treatment for adolescents with conduct disorder in out-of-home placement. In E. Hibbs, & P.
Jensen (Eds.), Psychosocial treatment for child and adolescent disorders: Empirically-based strategies for clinical practice (pp. 475–496). Washington, DC:
APA.
Cicchetti, D., & Toth, S. L. (2005). Child maltreatment. Annual Review of Clinical Psychology, 1, 409–438.
Clausen, J. M., Landsverk, J., Ganger, W., Chadwick, D., & Litrownik, A. (1998). Mental health problems of children in foster care. Journal of Child and Family
Studies, 7, 283–296.
DeKlyen, M. (1996). Disruptive behavior disorder and intergenerational attachment patterns: A comparison of clinic-referred and normally functioning
preschoolers and their mothers. Journal of Consulting and Clinical Psychology, 64(2), 357–365.
dosReis, S., Magno Zito, J., & Safer Soeken, K. L. (2001). Mental health services for youths in foster care and disabled youths. American Journal of Public Health,
91, 1094–1099.
Dozier, M., & Lindhiem, O. (2006). This is my child: Differences among foster parents in commitment to their young children. Child Maltreatment, 11(4),
338–345.
Dozier, M., Peloso, E., Lindhiem, O., Gordon, M. K., Manni, M., Sepulveda, S., & Ackerman, J. (2006). Developing evidence-based interventions for foster
children: An example of a randomized clinical trial with infants and toddlers. Journal of Social Issues, 62, 767–785.
Dozier, M., Stoval, K. C., Albus, K. E., & Bates, B. (2001). Attachment for infants in foster care: The role of caregiver state of mind. Child Development, 72(5),
1467–1477.
Dozier, M., Zeanah, C. H., & Bernard, K. (2013). Infants and toddlers in foster care. Child Development Perspectives, 7(3), 166–171.
Dubois-Comtois, K., Cyr, C., Vandal, C., & Moss, E. (2012). Le placement en famille d’accueil: Vulnérabilité socio-affective de l’enfant et modèle d’intervention
relationnelle. In G. Dans, M. Tarabulsy, J.-P. Provost, A. Lemelin, C. Plamondon, & Dufresne (Eds.), Développement social et émotionnel chez l’enfant et
l’adolescent, Tome 2: Applications pratiques et cliniques (pp. 29–45). Québec: Les Presses de l’Université du Québec.
Dubois-Comtois, K., & Moss, E. (2004). Relation entre l’attachement et les interactions mère-enfant en milieu naturel et expérimental à l’âge scolaire.
Canadian Journal of Behavioural Science/Revue canadienne des sciences du comportement, 36(4), 267–279. http://dx.doi.org/10.1037/h0087236
Dubois-Comtois, K., Moss, E., Cyr, C., & Pascuzzo, K. (2013). Behavior problems in middle childhood: The predictive role of maternal distress, child attachment,
and mother–child interactions. Journal of Abnormal Child Psychology, http://dx.doi.org/10.1007/s10802-013-9764-6
Easterbrooks, M. A., Bureau, J.-F., & Lyons-Ruth, K. (2012). Developmental correlates and predictors of emotional availability in mother–child interaction:
A longitudinal study from infancy to middle childhood. Development and Psychopathology, 24, 65–78.
Fearon, R., Bakermans-Kranenburg, M. J., van IJzendoorn, M. H., Lapsley, A., & Roisman, G. I. (2010). The significance of insecure attachment and disorgani-
zation in the development of children s externalizing behavior: A meta-analytic study. Child Development, 81, 435–456.
Fisher, P. A., & Stoolmiller, M. (2008). Intervention effects on foster parent stress: Associations with child cortisol levels. Development & Psychopathology,
20, 1003–1021.
George, C., & West, M. (2001). The development and preliminary validation of a new measure of adult attachment: The Adult Attachment Projective.
Attachment and Human Development, 3, 30–61.
George, C., & West, M. (2011). The adult attachment projective picture system: Integrating attachment into clinical assessment. Journal of Personality
Assessment, 93(5), 407–416.
George, C., & West, M. L. (2012). The Adult Attachment Projective Picture System: Attachment theory and assessment in adults. New York: Guilford Press.
130 K. Dubois-Comtois et al. / Child Abuse & Neglect 48 (2015) 119–130

Gilbert, R., Widom, C. S., Browne, K., Fergusson, D., Webb, E., & Janson, S. (2009). Burden and consequences of child maltreatment in high-income countries.
Lancet, 373(9657), 68–81.
Groh, A., Roisman, G., van IJzendoorn, M., Bakermans-Kranenburg, M., & Fearon, R. (2012). The significance of insecure and disorganized attachment for
children’s internalizing symptoms: A meta-analytic study. Child Development, 83, 591–610.
Healey, C. V., & Fisher, P. A. (2011). Young children in foster care and the development of favorable outcomes. Children and Youth Services Review, 33(10),
1822–1830.
Hesse, E. (2008). The adult attachment interview: Protocol, method of analysis, and empirical studies. In J. Cassidy, & P. R. Shaver (Eds.), Handbook of
attachment: Theory, research, and clinical applications (2nd ed., pp. 552–598). New York: Guilford Press.
Hesse, E., & Main, M. (2006). Frightened, threatening, and dissociative parental behavior in low-risk samples: Description, discussion, and interpretations.
Development and Psychopathology, 18(02), 309–343.
Heubeck, B. G. (2000). Cross-cultural generalizability of CBCL syndromes across three continents: From the USA and Holland to Australia. Journal of Abnormal
Child Psychology, 28(5), 439–450. http://dx.doi.org/10.1023/a:1005131605891
Juen, F., Arnold, L., Meissner, D., Nolte, T., & Buchheim, A. (2013). Attachment disorganization in different clinical groups: What underpins unresolved
attachment? Psihologija, 46(2), 127–141.
Lawrence, C. R., Carlson, E. A., & Egeland, B. (2006). The impact of foster care on development. Development and Psychopathology, 18, 57–76.
Lightfoot, E., Hill, K., & LaLiberte, T. (2011). Prevalence of children with disabilities in the child welfare system and out of home placement: An examination
of administrative records. Children and Youth Services Review, 33(11), 2069–2075.
Lindhiem, O., & Dozier, M. (2007). Caregiver commitment to foster children: The role of child behavior. Child Abuse & Neglect, 31(4), 361–374.
Madigan, S., Bakermans-Kranenburg, M. J., Van IJzendoorn, M. H., Moran, G., Pederson, D. R., & Benoit, D. (2006). Unresolved states of mind, anomalous
parental behavior, and disorganized attachment: A review and meta-analysis of a transmission gap. Attachment & Human Development, 8(2), 89–111.
Madigan, S., Moran, G., Schuengel, C., Pederson, D. R., & Otten, R. (2007). Unresolved maternal attachment representations, disrupted maternal behavior
and disorganized attachment in infancy: Links to toddler behavior problems. Journal of Child Psychology and Psychiatry, 48(10), 1042–1050.
Main, M., & Goldwyn, R. (1998). Adult attachment classification system. London: University College.
Mennen, F. E., Brensilver, M., & Trickett, P. K. (2010). Do maltreated children who remain at home function better than those who are placed? Children and
Youth Services Review, 32(12), 1675–1682.
Moss, E., Rousseau, D., Parent, S., St-Laurent, D., & Saintonge, J. (1998). Correlates of attachment at school age: Maternal reported stress, mother–child
interaction, and behavior problems. Child Development, 69, 1390–1405. http://dx.doi.org/10.2307/1132273
Neil, E., Beek, M., & Schofield, G. (2003). Thinking about and managing contact in permanent placements: The differences and similarities between adoptive
parents and foster carers. Clinical Child Psychology and Psychiatry, 8(3), 401–418.
Newton, R. R., Litrownik, A. J., & Landsverk, J. A. (2000). Children and youth in foster care: Disentangling the relationship between problem behaviours and
number of placements. Child Abuse & Neglect, 24, 1363–1374.
NICHD Early Child Care Research Network. (2006). Infant–mother attachment classification: risk and protection in relation to changing maternal caregiving
quality. Developmental Psychology, 42(1), 38–58.
Oosterman, M., & Schuengel, C. (2008). Attachment in foster children associated with caregivers’ sensitivity and behavioral problems. Infant Mental Health
Journal, 29(6), 609–623.
Oosterman, M., Schuengel, C., Slot, N. W., Bullens, R. A. R., & Doreleijers, T. A. H. (2007). Disruptions in foster care: A review and meta-analysis. Children and
Youth Services Review, 29, 53–76.
Oyen, A. S., Landy, S., & Hilburn-Cobb, C. (2000). Maternal attachment and sensitivity in an at-risk sample. Attachment & Human Development, 2(2), 203–217.
Pagé, G., Piché, A.-M., Ouellette, F.-R., & Poirier, M.-A. (2008). Devenir parent sans donner naissance: la construction d’un lien avec un enfant en contexte
d’adoption. In C. Parent, S. Drapeau, M. Brousseau, & E. Pouliot (Eds.), Visages multiples de la parentalité (pp. 89–121). Sainte-Foy, QC: Presses de
l’Université du Québec.
Pederson, D. R., Gleason, K. E., Moran, G., & Bento, S. (1998). Maternal attachment representations, maternal sensitivity, and the infant–mother attachment
relationship. Developmental Psychology, 34(5), 925.
Quebec Association for Child Protection. (2013). Bilan des directeurs de la protection de la jeunesse: Un élan pour voir grand. Québec, Canada: Gouvernement
du Québec.
Rosenthal, J. A., & Curiel, H. F. (2006). Modeling behavioral problems of children in the child welfare system: Caregiver, youth, and teacher perceptions.
Children and Youth Services Review, 28, 1391–1408.
Steele, M., Kaniuk, J., Hodges, J., Haworth, C., & Huss, S. (1999). The use of the Adult Attachment Interview: Implications for assessment in adoption and
foster care. In Assessment, preparation and support: Implications from research. London: BAAF.
Slade, A., Grienenberger, J., Bernbach, E., Levy, D., & Locker, A. (2005). Maternal reflective functioning, attachment, and the transmission gap: A preliminary
study. Attachment & Human Development, 7(3), 283–298.
Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics (4th ed.). Needham Heights, MA: Allyn & Bacon.
Tarren-Sweeney, M. J., Hazell, P. L., & Carr, V. J. (2004). Are foster parents reliable informants of children’s behaviour problems? Child: Care, Health and
Development, 30(2), 167–175.
Taussig, H. (2002). Risk behaviors in maltreated youth placed in foster care: A longitudinal study of protective and vulnerability factors. Child Abuse &
Neglect, 26(11), 1179–1199.
Thompson, R. A. (2008). Early attachment and later development: Familiar questions, new answers. In J. Cassidy, & P. R. Shaver (Eds.), Handbook of attachment:
Theory, research, and clinical applications (2nd ed., pp. 348–365). New York: Guilford Press.
U.S. Department of Health and Human Services, Administration for Children and Families, Administration on Children, Youth and Fam-
ilies, Children’s Bureau. (2015). Child maltreatment 2013. Available from http://www.acf.hhs.gov/programs/cb/research-data-technology/
statistics-research/child-maltreatment
Van IJzendoorn, M. (1995). Adult attachment representations, parental responsiveness, and infant attachment: A meta-analysis on the predictive validity
of the Adult Attachment Interview. Psychological Bulletin, 117(3), 387.
Vuchinich, S., Ozretich, R. A., Pratt, C. C., & Kneedler, B. (2001). Problem-solving communication in foster families and birthfamilies. Child Welfare, 81(4),
571–594.
Webster, L., & Joubert, D. (2011). Use of the Adult Attachment Projective Picture System in an assessment of an adolescent in foster care. Journal of Personality
Assessment, 93(5), 417–426.
Whipple, N., Bernier, A., & Mageau, G. A. (2011). A dimensional approach to maternal attachment state of mind: Relations to maternal sensitivity and
maternal autonomy support. Developmental Psychology, 47(2), 396.
Zajac, K., & Kobak, R. (2009). Caregiver unresolved loss and abuse and child behavior problems: Intergenerational effects in a high-risk sample. Development
and Psychopathology, 21(01), 173–187.
Zima, B. T., Bussing, R., Freeman, S., Yang, X., Belin, T. R., & Forness, S. R. (2000). Behavior problems, academic skill delays and school failure among school-aged
children in foster care: Their relationship to placement characteristics. Journal of Child and Family Studies, 9, 87–103.