THE EXPERIENCE OF OUT-OF-HOME CARE
AND YOUTHS PERCEIVED QUALITY OF
Edyta Marie Biegunajtys
B.S., Colorado State University, 2006
A thesis submitted to the
University of Colorado Denver
in partial fulfillment
of the requirements for the degree of
Masters of Arts
This Thesis for Master of Arts
has been approved
Biegunajtys, Edyta Marie (M.A. Clinical Psychology)
The experience of out-of-home care and youths perceived quality of relationships
Childhood maltreatment can have a significant impact on healthy relationship formation. This
study examines the perceived quality of relationships of maltreated youth placed in out-of-
home care. Variables associated with out-of-home care, such as number, length, and type of
placements, number of school changes, and placement with or without siblings, were
considered in association with youths perceived quality of relationships to both caregivers
and peers. Baseline data from a preventive intervention for maltreated youth (ages 9 to 11)
placed in out-of-home care was analyzed using multiple correlation and regression
techniques. It was hypothesized that higher perceived quality of relationships with caregivers
and peers would be associated with a fewer number of placements, a longer duration of
current placement, fewer number of school changes, placement with siblings, and living with
kin as opposed to non-relative foster care. In general, the data did not support proposed
hypotheses. However, several interesting findings emerged. Child externalizing behavior
problems were significantly related to poorer perceived relationships with caregivers and
peers. A higher number of school placements also predicted poorer perceived caregiver
relationships for girls only. Gender was also a significant predictor of items reflecting greater
intimacy and self-disclosure with peers and caregivers, with girls reporting feeling more
comfortable talking about their problems and feelings with others. These factors are important
to consider when working with youth in out of home care on their interpersonal relationships.
This abstract accurately represents the content of the candidates thesis. I recommend its
I dedicate this thesis to my mom, dad and brother. Their value of education, love and
encouragement helped me understand and reach my own goals in life. I also dedicate this to
Trevor, my best friend and partner in crime. He saw me through all the late nights and
showed me nothing but support. Thank you!
I would like to thank Heather Taussig, Sara Culhane and the whole Fostering Healthy Futures
project for the opportunity to conduct this research. I would especially like to thank Christie
Petrenko for her constant support, patience time and generous contributions throughout the
process. Additionally, thank you to my committee members Peter Kaplan and Allison Bashe
for their contributions and sense of humor during my presentations!
TABLE OF CONTENTS
The Importance of Relationships...................................3
Relationships and Maltreatment....................................4
Relationships and Factors Associated with Out-of-Home Care........7
Number of Placements and Length f Stay.........................7
Type of Placement...............................................8
Number of School Changes........................................9
Need for Present Study............................................9
Specific Research Questions and Hypotheses.....................10
Potential Moderator Variables................................14
Data Analysis .................................................14
Descriptive Statistics of Study Variables......................16
A. PIML CAREGIVER SURVEY..............................................32
B. PIML PEER SURVEY...................................................34
LIST OF TABLES
1 Bivariate Correlations among Study Variables and IPPA Caregiver/Peer.........17
According to the U.S. Department of Social Services, 794,000 children were the
victims of maltreatment in 2007 (Child Maltreatment, 2007). Maltreatment in childhood, as
defined by Barnett, Manley, and Cicchetti (1993), can constitute physical, sexual, and
emotional abuse, being witness to domestic violence, involving the child in moral/legal
activities, as well as neglect in the form of lack of supervision, failure to provide, and
educational neglect. Maltreatment is a significant public health concern and can have a major
impact on the individual, family, community, and society as a whole. In fact, the cost to
society has been estimated at over 24 billion dollars in direct costs such as mental health,
medical, and social services needed. Indirect costs, however, are estimated at over 70 billion
dollars and can include losses of productivity, and costs associated with teen pregnancies,
juvenile delinquency, school failure, homelessness, and adult criminality (Prevent Child
Abuse America, 2001).
Maltreated children are generally at high risk for a multitude of adverse outcomes.
Academically, they are more likely to have poorer academic achievement, increased special
education services, as well as higher truancy and dropout rates (for a review see
Cicchetti, 1994). They also exhibit emotional and behavioral problems such as delinquency,
aggression, anxiety, depression, and interpersonal adjustment problems in the home, the
school, and the community (Wodarski, 1990). Lastly, issues with unemployment,
incarceration, teen pregnancy, and mental health difficulties give rise to problematic adult
outcomes. For example, Courtney and Piliavin (1998) found that in young adults (past the
age of 18 and 12-18 months after leaving foster care) 37% of foster youth had not finished
high school, 50% were unemployed, and 27% of males and 10% of females had been
Depending on a childs circumstances and the nature and severity of maltreatment,
youth are sometimes taken into the custody of social services to protect them from further
harm. Since the passage of the Child Welfare Reform Act in 1979 and the Federal Adoption
Assistance and Child Welfare Act of 1980, child welfare policy mandates that youth who
require out-of-home placements must be placed in the least restrictive level of care possible
(Pecora et al., 2000). Out-of-home care can include kinship care, foster care, and institutional
or group care (i.e., residential treatment centers, group homes). As of 2008, the Adoption and
Foster Care Analysis and Reporting System (AFCARS) estimates that over 510,000 children
in the U.S. currently reside in some form of out-of-home care. This number becomes even
more significant when considering the estimated 20,000 young people who "age out of the
U.S. foster care system every year.
In addition to suffering the consequences of child maltreatment, children placed in out-
of-home care are likely to experience additional trauma from being removed and often
isolated from their homes, schools, friends, and family (Taussig, 2002). A history of foster
care is a risk factor for a host of adverse outcomes such as poor physical health
(Gramkowski et. al, 2009), mental health issues (Clausen, Landsverk, Ganger, Chadwick, &
Litrownik, 1998; Stein et al., 1996), higher incidence of school failure (Blome, 1997; Crazier &
Barth, 2005; Salzinger, Feldman & Muriel, 1993), negative peer interactions (Parker & Asher,
1987; Price & Brew, 1998), and behavior problems (Kortenkamp & Ehrle, 2002; Snyder et al
2005). In fact, children in foster care have social and emotional problems at rates three to ten
times higher then those found in the general population (Nilsen, 2007). Due to a foster childs
diminished resources, quality of relationships with others becomes extremely important in
their development and coping ability.
The Importance of Relationships
The ability to form and maintain relationships is central to the human experience. From
the day we are born, our social support system affects not only our personality development,
but also our health, mood, and even longevity. The ability to form relationships with others
serves as a buffer for many negative life outcomes such as depression (Panzarella, Alloy &
Whitehouse, 2006), cognitive decline in old age (Seeman, Lusignolo, Albert & Berkman,
2001; Wallace, 1986), and even impacts our ability to heal and overall physical health
In addition to serving a protective role, our social support system can provide physical,
emotional, and spiritual comfort, as well as stimulate personal growth. Our support networks
vary in purpose and complexity across the lifespan. During infancy, our attachment to a
primary caregiver is critical to our development of trust and belonging. As children develop,
the role of friendships gains importance, and adolescent and adult relationships grow in
complexity. Significant disruption of the normal development of key relationships can
contribute to poor adjustment and adverse outcomes.
Relationships & Maltreatment
Because social adjustment is an important predictor of childrens long-term outcomes
(Howe & Parke, 2000), it is crucial to understand the impact of childhood maltreatment on
relationship development. Maltreatment can often deter healthy relationship formation by
disturbing childrens social support systems, as will be detailed below. Unfortunately, children
in out-of-home care are at an even higher risk for poor social outcomes often due to higher
severity of maltreatment and disruption of existing relationships when placed in out-of-home
care. These children leave their families, neighborhoods, schools, and communities and are
typically placed into new situations quickly with little time for adjustment. The following
sections review the literature on caregiver and peer relationships in children with histories of
maltreatment and out-of-home care, which was the focus of the current study.
Whether children are in out-of-home care or not, relationships with caregivers provide
children with models of behaving in a social world and teach children how to negotiate
interactions with others (Cicchetti, Lynch, Shonk, & Manly, 1992). During normal
development, young children look to their primary caregivers for direction, particularly in
stressful situations in which the child needs support and assistance from the caregiver to
cope effectively. If a caregiver does not respond in an appropriate manner, this experience
can lead to an insecure attachment in which the child lacks trust in the caregivers availability,
support, and guidance (Egeland, Yates, Appleyard, & van Dulmen, 2002).
Experiencing abuse and neglect can significantly impact youths attachments to
caregivers, which can make forming close relationships to any caregiver challenging.
Removal from ones home and placement with strangers may only intensify such attachment
issues. Attachment to new caregivers may be particularly difficult for young children entering
foster care who often experience disruptions in care at a point when maintaining attachment
relationships is a key, biologically based task. Children in out-of-home care often push away
their new caregivers, even though such children are especially in need of nurturing care
(Dozier, &Lindhiem, 2009). Research finds that having at least one adult attachment figure
who is trustworthy and makes a child feel cared about can serve as a protective factor
against negative outcomes (Rutter, 1990).
Friendships are a factor known to contribute to better psychological adjustment
among at-risk youth (Legault, Anawati & Flynn, 2006). Making and maintaining friendships is
an especially important developmental milestone for maltreated youth, who are already at
high risk for poor adjustment. Children who lack basic social skills and fail to develop
successful peer relations during school entry are at greater risk for conduct problems, peer
rejection, and academic failure throughout childhood and adolescence (Brendgen, Vitaro,
Bukowski, Doyle, & Markiewicz, 2001; Dishon, 1990; Leve, et al, 2007; Snyder et al., 2005).
Peer rejection has been found to be one of the strongest predictors of school drop out as well
as later adult adjustment problems (Salzinger, Feldman & Hammer, 1993).
Studies with children who have been maltreated or been in foster care demonstrate a
consistent association between maltreatment/foster care and poor peer relations (Bolger,
Patterson, & Kupersmidt, 1998; Fantuzzo, Weiss, Atkins, Meyers, & Noone, 1998; Manly,
Kim, Rogosch, & Cicchetti, 2001; Parker & Herrera, 1996; Rogosch & Cicchetti, 1994).
Maltreated children display less intimacy and more conflict when interacting with their best
friends compared to non-maltreated children (Parker & Herrera, 1996). In addition, maltreated
children report less satisfactory relationships with peers and poorer self-concepts as
compared with children who have not been maltreated. (Cicchetti, Lynch, Shonk, & Manly,
1992; Mueller & Silverman, 1989; Vondra, Barnett, & Cicchetti, 1989).
The chronicity of maltreatment and gender of the child are associated with the quality
of peer relationships for children in out-of-home care. Children who have experienced a
longer duration of maltreatment show a greater decline in friendship quality over time (Bolger,
Peterson, Kupersmidt, 1998). Additionally, girls in foster care tend to have poorer
relationships with their peers (Leve et al., 2007). This may be in part to girls being less
accepting of externalizing behaviors in their peers, which is more common in both girls and
boys in out-of-home care.
While many children in out-of-home care struggle with peer relationships, having a
close friend can promote resilience. In one longitudinal study, chronically maltreated children
who reported a high quality friendship, showed an increase in self-esteem over time (Bolger,
Peterson, Kupersmidt, 1998). These results suggest youth can benefit from peer-based
intervention strategies to improve their social and personal adjustment. Due to youths
disrupted connections with caregivers its not surprising that peers would play an important
role in the lives of foster youth (Shook et. al, 2008).
Relationships and Factors associated with out-of-home care
The purpose of this study was to examine the influence of variables associated with
being placed in out-of-home care on perceived quality of relationships with caregivers and
peers. Existing literature on maltreated childrens relationships and variables of interest to this
study, such as number of placements, length of current placement, type of placement,
placement with siblings, and number of school changes, are detailed below.
Number of Placements and Length of Stay
In general, research shows that as the number of placements increases, the amount
of risk increases for adverse outcomes in a youth placed in out of home care (Anctil,
McCubbin & Pecora, 2007). Children who experience a greater number of placements have
less time to build quality relationships with caregivers, peers, and community members.
Smith (1995) found that foster children who experience multiple placements may be able to
reach only the early stages of friendship formation before they leave the foster home. This
type of friendship requires little emotional effort and is not likely to last onto later stages of
friendship, which require more effort. This suggests that foster children are at-risk for
developing a behavior pattern that is likely to be aversive to their peers simply because they
may not live in an environment that facilitates the development of stable long lasting
friendships. To my knowledge no studies at this time have linked foster childrens perceived
quality of relationships with the number and length of placements.
Type of Placement
Studies of maltreated youth suggest that children may have a difficult time forming
secure attachments with their caregivers (Cicchetti, Barnett, Rabideau, & Toth, 1991). This is
especially true of non-familial caregivers that the child has never met before. Placement in
kinship care as opposed to non-relative foster care is associated with better peer and
caregiver relationships (Inglehart, 1995). Review of the literature shows that when compared,
kinship care generally lasts longer, results in reunification less often, and is more stable
(Scannapieco, 1999). These characteristics are related to better overall outcomes for kids in
Siblings are often the only stable and enduring relationship for a substantial number
of foster children who routinely experience changes in foster home, school, and
neighborhood while in care (Linaares, 1998). Foster care placement and the presence of
siblings may influence the nature of a childs interactions with non-related peers (Smith,
1995). For instance, research finds that siblings, when placed together, tend to do better
socially than those placed separately so long as they are a positive influence on one another
and do not demonstrate abusive patterns (Smith, 1995). Kramer and Gottman (1992) show
that placement with a sibling may help build better friendships. However, Leathers (2005)
finds that sibling placement can sometimes confound friendship formation because
maltreatment effects can be projected onto unhealthy attachment styles in friendship
Number of School Placements
There has been little research on the impact of the number of school changes in the
foster care population. Studies in the general population however demonstrate that there is
loss of social connections found when children move schools (Pribesh & Downey, 1999).
Children in these situations often suffer academically as well as socially. Frequent school
changes may be particularly important for understanding peer social support disruption, as
school is the source of most childhood friendships. Additionally, the extra stressor of being a
new student could also affect youths perceived support from peers.
Need for Present Study
Due to the significant risk factors for youth placed in out-of-home care, it is extremely
important to determine what factors impact relationships for children in foster care to promote
resiliency and better outcomes. This study seeks to answer the question of how the
experience of being in out-of-home care influences the perceived quality of youth
relationships. While the above studies have examined some associations between childrens
relationships and factors associated with out-of-home care, none to my knowledge have
assessed multiple relationships and factors in a single study. The current study will focus on
childrens perceived relationships with caregivers and friends. Results from this study
highlight the characteristics of children in out-of-home care that increase their chances for
better relationship outcomes. These results may help inform targets for further intervention.
Specific research questions and hypotheses
The current study investigated two different research objectives. First, the number of
placements, duration of current placement, type of placement, whether siblings were placed
together, and number of schools attended were observed to see if they were associated with
youths perceived quality of relationships with caregivers. For this question it was predicted
that higher perceived quality of relationships with caregivers would be associated most with
fewer placements, longer duration of current placement and living with kin as opposed to
non-relative care. Next, the number of placements, duration of current placement, type of
placement, whether siblings were placed together, and number of schools attended were
examined to see if they were associated with youths perceived quality of relationships with
peers. It was hypothesized that higher perceived quality of relationships with peers would be
associated most with fewer placements, fewer numbers of school changes, and placement
Participants in the current study included 9-11 year old youth and their caregivers
who were recruited for a randomized controlled trial (RCT) known as Fostering Healthy
Futures (FHF). FHF is a preventive intervention program for preadolescent youth placed in
out-of-home care (for more on this program see Taussig, Culhane, & Hettleman, 2007).
Eligibility for the RCT included the following criteria: 1) Youth had been court ordered into out-
of-home placement within the preceding year 2) the placement was due to maltreatment and
3) at the time of the baseline interview, youth remained in out-of-home care. Ninety-two
percent of those individuals who met eligibility requirements agreed to participate in the RCT
study. The current study examined data that was collected during baseline assessments (pre-
randomization) from 2002 to 2009.
Although 455 youth were interviewed at baseline, the sample for this study included
335 youth and their caregivers. Participant data were excluded for the following reasons: 30
childrens achievement and intelligence scores indicated significant cognitive impairment (IQ<
70) limiting childrens ability to comprehend and answer assessment instruments; four
children were not proficient enough in English to understand the study questions; 39 children
were siblings of others included in the sample (when siblings were interviewed, one was
chosen at random to be included in the current studys analyses); 26 participants were
excluded due to incomplete youth or caregiver data; and 21 were excluded because the
institutional care (i.e., group home, shelters, residential treatment centers) category was too
small; relative to the foster care and kinship care groups for meaningful comparisons.
The final sample of 335 youths was approximately 49.9% female (n=167), with a
mean age of 10.31 (SD= .89). Youth were placed in kinship care (53.7%), or foster care
(46.3%) at relatively similar rates. At the time of the baseline interview, youth had lived an
average of 5.45 months (SD=3.88) in their current placement and had an average of 2.37
placements (SD=1.38). Of the 335 youths, 323 (96.4%) had siblings, out of those, 231 (69%)
had at least one sibling living with them in their current placement. Ethnic/racial distribution of
the youths sample, based on caregivers-report, is approximately 29.6% Hispanic, 49.9%
Caucasian, 26.3% African-American, 14.9% Native American, and 1.2% Asian or Pacific
Islander (non-exclusive categories).
The study protocol was approved by the Colorado Institutional Review Board. Both
assent (from the youth) and consent (from the legal guardian) was obtained prior to
participation in the baseline interview. Participants were paid $40.00 for their contribution to
the study. Youth and their current caregivers were interviewed separately for about 3-4 hours
over two-sessions, and 2 hours, respectively, most frequently in their own homes.
People in My Life (PIML: Gifford-Smith, 2000) is a 30-item measure of perceived
quality of relationships to caregivers and peers. The PIML was derived from a longer,
adolescent version of the Inventory of Parent and Peer Attachment (IPPA; Armsden &
Greenberg, 1987) and was modified for use with elementary school-age youth by the FAST
Track Project (Gifford-Smith 2000). The original IPPA demonstrated adequate internal
consistency (as for subscales range from .72 to .91) and construct validity (Armsden &
Greenberg, 1987). PIML has also demonstrated high internal consistency (as from .80 to
.83) and its subscales reliably discriminate between samples of high risk and normative
The explanatory variables were measured by youth self-report as well as caregiver
report. Predictors include number of placements (given on the placement history), length of
stay at the current placement, placement type, whether or not they are placed with siblings,
and the number of schools they have attended.
Potential Moderator Variables
Potential moderating variables including gender and behavior, as measured by the
Child Behavior Checklist (CBCL; Achenbach, 2001), were considered in analyses. The CBCL
is a widely used caregiver-report of childhood behavior. Caregivers rate 113 items on a
three-point scale to yield data on internalizing and externalizing behavior problems. The
Total Problems scale, which incorporates both internalizing and externalizing behaviors, was
used in this study. The CBCL has been standardized separately for boys and girls in different
age groups and its reliability and validity have been well established (Achenbach, 2001).
All data scoring and entry was rechecked prior to analyses in SPSS. Data was
examined for the presence of missing data, outliers, normality, and homogeneity of variance
prior to primary analyses. Missing data was excluded listwise. While univariate and
multivariate outliers were present, they did not meet cutoff criteria for influence statistics for
overall regression models (i.e., DFFITS) or individual regression coefficients (i.e., DFBETA).
Multiple regression/correlation techniques were utilized to examine the association of out-of-
home care variables with perceived quality of relationships with caregivers and peers. First, a
correlation matrix containing all response, explanatory, and control variables was inspected
to assess for multicollinearity and to identify moderating variables that should be included in
regression analyses. Multiple regression analyses were then completed for PIML-caregiver
and PIML-peer separately, with the control variables entered in hierarchically in the second
block. All regression weights were reported as standardized beta coefficients. With a sample
size of 335, power analyses utilizing GPOWER software (Erdfelder, E., Faul, F., & Buchner,
A.,1996) suggest this study was sufficiently powered to detect small to medium effects.
Specifically, with an alpha set at .05, assumed power of .80, and seven predicting variables,
regression analyses should be able to detect an effect size of f2 =.04.
Descriptive Statistics of Study Variables
PIML caregiver scores had a mean of 2.57 (out of a maximum of 3; SD= .35) and
PIML peer scores had a mean of 2.47 (SD= .36). The mean length of youths current
placement was 5.45 months (SD= 3.88) and the mean total time in out of home care was
6.45 months (SD= 3.63), with youth having experienced an average of 2.37 (SD= 1.38)
placements. 96.4% of the overall sample had siblings, out of those, 69% had at least one
sibling living with them in their current placement. By the time of the baseline interview, youth
had experienced an average of 4.25 (SD=3.04) school placements. The mean Total CBCL
score for this sample was 61.02 (SD= 11.83), which is in the borderline clinical range
(Borderline T = 60-63, Clinical >63).
Bivariate correlations among study variables are provided in Table 1. Overall,
correlations tended to be small and no evidence of multicollinearity was present. The
correlations show that lower PIML caregiver scores (i.e., poorer perceived relationship) were
related to being placed in foster care (vs. kinship care) and having greater behavior
problems. Having a lower score on the PIML peer is related to being male and having had
shorter placement duration. Foster care placement is related to youth having had a shorter
Table 1: Bivariate Correlations among Study Variables and IPPA Caregiver/Peer
Variable 1 2 3 4 5 6 7 8
1. IPPA Caregiver -.14* .04 -.07 .06 -.03 -.26* -.07
IPPA Peer .04 -.09* -.04 -.03 -.05 -.05 -.11*
2. Kinship vs Foster - -.16* .10* -.16* .05 .21* .01
3. Length of Current Placement - -.04 .08* -.01 .02 -.03
4. Number of Placements - -.09* .08* .13 -.10*
5. Siblings in Current Placement -- -.11* -.08* .00*
6. Number of School Placements - .13* .09*
7. CBCL Total Problem Score -- .01
8. Youth Gender -
Note: CBCL = Child Behavior Checklist.
current placement, more behavior problems, more placements overall, and lower likelihood of
being placed with siblings. Having been placed with siblings is related to higher chances of
being placed in kinship care, longer duration of current placement and fewer total
placements, fewer school changes, and fewer behavior problems. More school placements
were related to higher number of placements, lower likelihood of being placed with siblings,
more behavior problems, and higher probability of being male. Higher behavior problems
were related to being placed in foster care (vs kinship care), more school and home
placements and lower chances of being placed with siblings.
Due to significant correlations with explanatory and response variables, the CBCL
and gender variables were included in the model. The overall model (with CBCL and gender)
was significant, F (7, 327) = 4.47, p<.001, and accounted for 7% of the variance of the PIML
caregiver scores (adjusted R2= .07). Only the CBCL Total Problems score reached criteria
for statistical significance when controlling for other variables, (3 = -.24,7(327) = -4.42, p <
.001. Specifically, a higher level of behavior problems is related to a poorer perceived
relationship with the caregiver. The inclusion of the CBCL Total Problems Score was
important in understanding the overall model. When CBCL Total Problems was not included,
placement type was a significant predictor, p = -.13, f(329) = -2.28, p=.02, although the
overall model did not account for a significant amount of variance in perceived caregiver
relationships, F(5, 329) = 1.79, p = .12, adjusted R2= .01. Placement type likely predicted
caregiver relationships when CBCL Total Scores were not included in the model because
children in non-relative foster care tend to have more significant behavior problems, thus
requiring a more restrictive placement. However, behavior problems, when included,
accounted for significant variance above and beyond variables associated with out of home
Explanatory variables did not account for a significant portion of the variance
(adjusted R2 = .003) in PIML peer ratings, F(5, 327) = 1.18, p = .32 even when CBCL and
gender were considered; F(7, 325) = 1.63, p = .13.
Given that the selected explanatory variables did not account for a significant portion
of the variance in PIML scores, exploratory data analyses were conducted to further
understand the results. Because CBCL Total Problem scores predicted the PIML caregiver
scores above and beyond the other variables of interest, the model was re-run using CBCL
Externalizing and CBCL Internalizing scores in place of the CBCL Total score to see if the
nature of these behavior problems could be better understood within this sample. The overall
model for PIML caregiver model was significant F(8, 326) = 3.66, p < .001, and accounted for
6.0% of the total variance. Externalizing behavior problems (p = -.22, f(326) = -3.07, p =.002)
appear to have driven the prior relationship with total behavior problems, as internalizing
problems did not account for significant variance ((3 = -.02, f(326) = -0.29, p=.77) when
controlling for other variables.
When CBCL Externalizing and CBCL Internalizing scores were substituted in the
model for PIML peer scores, the overall model approached the significance criterion, F(8,
324) = 1.79, p = .08. Gender ((3 = -.13, f(324) = -2.33, p = .02) was a significant predictor and
Externalizing Problems approached significance criterion (p = -.14, f(324) = -1.90, p = .06),
when controlling for other variables. Males and youth with externalizing behavior problems
tended to rate their peers lower on the PIML peer survey.
While gender was not a significant predictor in either PIML-caregiver or PIML-peer
analyses, research suggests youths perceptions of their relationships may vary by gender
(Furman & Buhrmester, 1985; Leve et al., 2007). To test this hypothesis more directly,
regression analyses were repeated for boys and girls separately. The overall model for girls
was significant and accounted for 6.1% percent of variance in perceived caregiver
relationships, F(6,160) = 2.80, p= .01. Both the CBCL Total Problems (P = -.19, f(160) = -
2.40, p = .02) and number of school placements (P= -.17, f(160) = -2.09, p = .04) were
significant predictors. Specifically, more behavior problems and a greater number of school
placements were related to poorer perceived relationships with caregivers. The overall model
was also significant for boys and accounted for 8.2% of variance in perceived caregiver
relationships, F(6,161) = 3.50, p = .003. However, only CBCL Total Problems (P = -.28,
f(161) = -3.63, p < .001) was significant controlling for other variables. Similar to girls, having
greater behavior problems was related to poorer perceived relationships with caregivers.
Consistent with overall analyses, when Peer PIML is examined separately by gender,
variables associated with out of home care did not account for variance in perceived peer
relationships (girls, F(6,158) = 0.58, p= .75; boys, F(6, 161) = 1.15, p= .33). In other words,
gender did not seem to make a difference.
CBCL Internalizing and Externalizing problems were also modeled separately for
boys and girls. For girls, the overall model was significant and accounted for 6.7% of the
variance in PIML caregiver scores, F(7,159) = 2.70, p = .01. Both the externalizing behaviors
(P = -.25, f(159) = -2.47, p = .01) and the number of school placements (p = -.16, f(159)= -
2.03, p = .04) were significant predictors. Youth with more externalizing behaviors and higher
number of school placements rated the relationships with their caregivers lower. The overall
model was also significant for boys ,F(7,160) = 2.53, p = .02, and accounted for 6.0% of
variance in PIML caregiver scores, however, no specific variables accounted for a significant
amount of variance above and beyond other variables included in the model. Similar to the
overall sample, variables associated with out of home care did not account for significant
variance for either boys or girls perceived relationships with peers, (girls, F(7, 157) = 0.78, p=
.60; boys, F(7,160) = 1.04, p = .40).
When examining the data in more detail, it was observed that some items on the
PIML showed little variability, whereas others elicited more variable responses from youth.
Items that elicited more variable responses tended to reflect more behavioral aspects of
relationships (e.g., I talk to my caregiver/friends about my problems). The limited variability of
some items may have reduced the variability of total PIML scores (which represents the
mean of all item scores) and the resulting ability to detect effects. To test this hypothesis,
regression analyses were repeated on individual items with more variability.
For PIML caregiver, items 6, 8,10,11, and 13 demonstrated greater response
variability than other items. The overall model accounted for significant variance (adjusted R2
= .033) for item 6, I talk to my caregivers about my problems, F(7, 327) = 2.64, p = .01.
CBCL Total Problems (P = -.13, f(327) = -2.22, p = .03) and gender (p = -.11, f(327) = -2.05, p
= .04) were both significant predictors for this item demonstrating that males and youth with
higher CBCL scores were more likely to give their caregivers lower ratings for this item.
Item 8, I share thoughts and feelings with my caregivers, showed that the overall
model accounted for a significant amount of variance (adjusted R2 = .042), F(7, 327) = 3.10,
p<.003. Both the placement type (p = -.13, f(327) = -2.32, p = .02) and gender (p = -.18,
f(327) = -3.26, p = .001) variables were significant predictors for this item, above and beyond
other variables. Specifically, males and children in non-relative foster care were more likely to
rate their caregivers lower on this question.
The overall model also accounted for significant variance (adjusted R2 = .03) for item
10, My caregivers dont understand what I am going through, F(7, 321) = 2.43, p = .02. The
placement type variable was a significant predictor for this item, p = -.14, f(321) = -2.36, p =
.02, indicating that foster yo.uth rated caregivers lower on this item than youth in kinship care.
Models for items 11 (F(7, 327) = 3.33, p = .002), I get upset easily with my
caregivers, and 13 (F(7, 327) = 2.99, p = .005), Its hard for me to talk to my caregivers
were also significant. CBCL Total Problems was a significant predictor for both models (item
number 11, p = -.24, f(327) = -4.24, p< .001 and item number 13, p = -.21, f(327) = -3.76, p<
.001), with children with more behavior problems rating poorer perceived relationships with
Overall, PIML peer responses were more variable than PIML caregiver responses.
For PIML peer, items 5, 6, 7, 8, and 10 showed greater response variability. The overall
model for item 6, I talk to my friends about my problems was significant and accounted for
2.4% percent of the variance, F(7, 325) = 2.19, p = .04. Gender ((3 = -.15, f(325)= -2.68, p =
.01) and placement type (p= .13, f(325) = 2.24, p = .03) were significant predictors. Males and
youth in kinship care rated their peer relationships as less supportive.
The overall model for item 8, I share thoughts and feelings with friends, was
significant F(7, 324) = 2.75, p = .009. Gender (p = -.17, f(324) = -3.06, p =.002) and the
number of school placements (p = -.12, f(324) = -2.15, p = .03) were significant predictors of
perceived peer relationships above and beyond other variables included in the model.
Specifically, Males and children with higher numbers of school placements reported less
supportive relationships with peers. Explanatory and control variables did not account for
significant variance for items 5 (My friends can tell when Im upset, F(7, 324) = 1.46, p =
.18), 7 (My friends ask if something is bothering me, F(7, 325) = 1.51, p = .16.) and 10 (My
friends dont understand what I am going through, F(7, 322) = 0.90, p = .51).
The purpose of this study was to examine youths perceived quality of relationships to
both caregivers and peers in relation to several factors associated with out of home care.
Both proposed hypotheses were not supported by the data. Specifically, the selected out of
home care variables (i.e., number of placements, length of current placement, type of
placement, placement with siblings, and number of school changes) did not predict youths
perceived relationships with their caregivers or peers. However, several interesting findings
Initial analyses (not controlling for gender or behavior problems) indicated placement type
was predictive of the quality of perceived caregiver relationships. Specifically youth in non-
relative foster care reported poorer quality relationships with caregivers. A possible
explanation for this relationship is that foster youth are less likely to know their caregiver
before they are placed than children in kinship care and may therefore be less likely to feel
supported. This finding supports research (Inglehart, 1995 & Scannapieco, 1999) that
showed youth living in foster care (vs kinship care) had poorer caregiver relationships.
However, when total behavior problems were included in the model, this relationship was no
longer significant. The relationship between being placed in foster care and poorer caregiver
relationships was likely due to children in non-relative foster care having greater behavioral
problems than those in kinship care (which is often a reason for placement in non-relative
foster care). When overall behavior problems were controlled for, placement type was no
longer a unique predictor of perceived quality of caregiver relationships. Follow-up analyses
revealed that externalizing behavior problems (e.g, breaking rules, aggressive behavior) were
driving this relationship, as internalizing problems (e.g., anxiety, depression) did not predict
quality of caregiver relationships. It is likely that having externalizing behavior problems in the
home would cause discord with the caregiver and inversely, foster youth were likely to feel
less positive about their caregivers.
Externalizing behavior problems also play a role in the perceived quality of peer
relationships. In overall analyses, no selected variables were predictive of perceived quality
of peer relationships. However, when total behavior problems were separated into
internalizing and externalizing behaviors, gender and externalizing behaviors were related to
youths perceived quality of peer relationships. Specifically, males and youth with
externalizing behavior problems tended to rate their peers as less supportive. Youth who
exhibit externalizing behaviors (i.e., not following directions, acting aggressively towards
others) towards their peers may be less liked by peers and inversely rate their peers lower on
the survey. This relationship may have been masked in the overall analysis by the total
behavior problem variable, which includes both internalizing and externalizing behavior
problems. In summary, the presence of externalizing behavior problems contributed uniquely
to youths perceived quality of relationships with both caregivers and peers.
While gender did not account for significant variance in perceived quality of either
caregiver or peer relationships for the overall sample, research (Furman & Buhrmester, 1985;
Leve, et. al, 2007) suggests youths perceptions of their relationships may vary by gender.
Results from the current study partially support this hypothesis. Specifically, the number of
school placements accounted for significant variance in perceived relationships with
caregivers for girls only, in addition to externalizing behavior problems, which was significant
for both boys and girls. This suggests that changing schools may be more difficult for girls
than for boys, and may uniquely impact the quality of their relationships with their caregivers,
even after controlling for externalizing behavior problems and other variables associated with
out of home care. In contrast, no selected variables were significant predictors of perceived
relationships with peers for either boys or girls when examined separately. This is somewhat
surprising given the previous finding of gender and externalizing behavior problems as
predictors of perceived peer relationship quality for the combined sample. It is possible that
the analyses lacked sufficient power to detect effects for boys and girls separately due to
smaller sample sizes within groups.
Lastly, the lack of variability of the individual item responses may have accounted for the
hypotheses not being supported. Total PIML scores are means, and less variable items may
obscure variable items that may be more reflective of relationship quality. To test this
hypothesis, individual item response variability was examined. Items that elicited more
variable responses tended to require more from the youth in terms of self-disclosure and
require them to be more active participants in the relationship. They also were more likely to
be more intimate in nature and require more personal investment from the youth then more
general items (i. e., 1 talk to my caregivers/friends about my problems vs. My
caregivers/friends pay attention to me). Social desirability may be one possible explanation
for the lack of variability of some item responses. For example, items such as My
caregivers/friends care about me may elicit more homogeneous response patterns than
other items such as I share thoughts and feelings with my caregivers/friends, which are
more open ended and less value-laden.
While some items that elicited more variable responses yielded similar results to overall
mean PIML caregiver scores (i.e., I get upset easily with my caregivers, Its hard for me to
talk to my caregivers), others demonstrated alternate patterns. Specifically, placement type
was a significant predictor for two items (I share thoughts and feelings with my caregivers,
and My caregivers dont understand what I am going through,) even when controlling for
behavior problems, which was not significant. For both items, children in foster care felt less
supported by their caregivers than youth in kinship care. This again may be related to
knowing your caregiver before living with them and having some level of comfort in confiding
in them. Gender was also a significant predictor for the item I share thoughts and feelings
with my caregivers, with girls feeling more comfortable sharing their thoughts and feelings
with caregivers than boys. It has been generally found that girls tend to form closer
relationships with others and are more open to share, especially with caregivers and intimate
peers (Furman & Buhrmester, 1985). Additionally, this research suggests that boys may not
be as comfortable with closer relationships (such as with caregivers) as they do better in
bigger groups and less intimate situations.
Overall PIML peer items elicited more variability than PIML caregiver items. Two of the
most variable items, which reflect greater intimacy with peers, (I talk to my friends about my
problems and I share thoughts and feelings with friends,) were related to gender,
suggesting limited item variability may have accounted for lack of findings with the quality of
perceived peer relationships. Consistent with previous research highlighting greater intimacy
in girls relationships (Furman & Buhrmester, 1985), girls reported feeling more comfortable
talking with friends and sharing their thoughts and feelings than boys. In addition, youth in
kinship care (vs. foster placements) reported having more difficulty talking with peers about
their problems. This finding is interesting as it goes against previous research (Inglehart,
1995) that shows living in kinship care has a positive impact on both peer and caregiver
relationships. Its possible that living in kinship care fosters more loyalty to kin members and
youth are less likely to talk about problems related to family matters with peers. Additionally,
having a better relationship with caregivers (which is more likely in kinship care) may
decrease the need to talk to peers about problems. Lastly, children with higher numbers of
school placements reported less supportive relationships with peers per lower ratings on item
8 (I share thoughts and feelings with friends ). This finding is consistent with research
(Pribesh & Downey, 1999) that finds higher school placements can lead to strained
relationships with others.
While some interesting findings have resulted from this investigation, there are limits
to the information provided. For instance, the PIML peer data was only the youths description
of their peer relationships; this did not measure reciprocated friendships. As Smith (1995)
discussed, children in foster care tend to build shallow relationships, and this may be the
case with this sample, even those children reporting high quality relationships with others.
Youth may have also given socially desirable responses when interviewed, wanting
the interviewer to see them as getting along well with their caregivers and peers. The
limitation may be remedied by obtaining ratings from caregivers and peers on their perceived
relationships with the youth. Teacher ratings of youths friendships could also be beneficial.
Having multiple reporters of a given relationship would provide a better picture of the youths
actual reciprocated friendships vs. those that are imagined or assumed by the youth.
Another limitation to the current study was that maltreatment type and severity were
not examined. While all youth in this study were maltreated, it is unknown at this time if
maltreatment type and/or severity would predict differential relationship quality for youth. This
could be addressed in future studies.
Despite study limitations, this study has both programmatic and theoretical
implications. Treatment planning with foster care populations should consider externalizing
behavioral problems as an important treatment target, if present, as it can take its toll on
relationship formation with both caregivers and peers. Some practical implications of these
results include engaging youth and caregivers in family therapy to address behavior problems
to improve their relationships. In addition, being sensitive to gender when providing specific
help with relationship formation is warranted. Perhaps males need more intensive services to
assist in relationship/ intimacy formation. It may also help females to target ways they could
feel more connected to their school or address their academic needs when relationship
problems are noticed.
Future studies should address the limitations of this study, specifically by using
multiple reporters to assess reciprocity of relationships. Another helpful future study may
include observing foster children at school and at home to better understand the nature of
their relationships with peers and caregivers. Hypothetical vignettes could also be given to
youth to assess how they evaluate relationship quality in general or how they might deal with
common problems in their relationships with others.
Future studies could also explore the nature of those externalizing behaviors that are
the most critical for lower ratings on perceived quality relationships with others. It would be
helpful to separate externalizing behaviors into more specific categories or behaviors such as
aggression or not following directions. This type of analysis could provide a much clearer
nature of the behavior problems in the context of relationships within the foster care sample.
While study results indicate that the explanatory variables were not robust predictors
of perceived quality of caregiver or peer relationships, some interesting insights were gained
from the exploratory portion of the study. Particularly, externalizing behavior problems are
important to understanding youths relationships with their caregivers and peers above and
beyond factors associated with out of home care. In addition, the factors that influence
relationship quality are likely to differ by gender. More frequent school changes are uniquely
associated with poorer relationships with caregivers for girls more so than boys. In addition,
girls report higher intimacy and self-disclosure in their relationships with caregivers and
peers. These factors are important to consider when working with youth in out of home care
on their interpersonal relationships.
PIML Caregiver Survey
These questions make statements about the people who care for you now. Please indicate
how true the statements are.
Not True Sometimes True Often True
1 2 3
IPPAc 1 My [caregivers] listen to me. 1 2 3
IPPAc 2 My [caregivers] accept me. 1 2 3
IPPAc 3 My [caregivers] care about me. 1 2 3
IPPAc 4 My [caregivers] help me with my problems. 1 2 3
IPPAc 5 My [caregivers] can tell when Im upset. 1 2 3
IPPAc 6 I talk to my [caregivers] about my problems. 1 2 3
IPPAc 7 My [caregivers] ask if something is bothering me. 1 2 3
IPPAc 8 I share thoughts and feelings with my [caregivers]. 1 2 3
IPPAc 9 My [caregivers] pay attention to me. 1 2 3
IPPAc 10r My [caregivers] dont understand what I am going
through. 1 2 3
IPPAc 11 r I get upset easily with my [caregivers]. 1 2 3
IPPAc 12r I feel angry with my [caregivers]. 1 2 3
IPPAc 13r Its hard for me to talk to my [caregivers]. 1 2 3
IPPAc 14r I feel scared at home. 1 2 3
IPPAc 15 My [caregivers] are proud of me. 1 2 3
PIML Peer Survey
These questions make statements about friends. Please indicate how true the statements
are for your friends. When we ask these questions, we dont mean adults who are friends.
We mean other kids who are friends.
Not True Sometimes True Often True
1 2 3
IPPAf 1 My friends listen to me. 1 2
IPPAf 2 My friends accept me. 1 2
IPPAf 3 My friends care about me. 1 2
IPPAf 4 My friends help me with problems. 1 2
IPPAf 5 My friends can tell when Im upset. 1 2
IPPAf 6 I talk to my friends about my problems. 1 2
IPPAf 7 My friends ask if something is bothering me. 1 2
IPPAf 8 I share thoughts and feelings with friends. 1 2
IPPAf 9 My friends pay attention to me. 1 2
IPPAf 10r My friends dont understand what I am going through. 1 2
IPPAf 11 r I get upset easily with my friends. 1 2
IPPAf 12r I feel angry with my friends. 1 2
IPPAf 13r Its hard for me to talk to my friends. 1 2 3
IPPAf 14r I feel scared when Im with my friends. 1 2 3
IPPAf 15 My friends are proud of me. 1 2 3
Achenbach, T. M. (2001). Integrative guide for the 2001 CBCL/4-18. YSR. and TRF profiles.
Burlington, VT: University of Vermont Department of Psychiatry.
Anctil.M.T., McCubbin, L.D., OBrien, K., & Pecora, P. (2007). An evaluation of recovery
factors for foster care alumni with physical or psychiatric impairments: Predictors of
psychological outcomes. Children and Youth Services Review. 29,1021-1034.
Blome, W. W. (1997). What happens to foster kids: Educational experiences of a random
sample of foster care youth and a matched group of non-foster youth. Child and
Adolescent Social Work Journal, 14, 41-53.
Bolger, K. E., Patterson, C. J., & Kupersmidt, J. B. (1998). Peer relationships and
self-esteem among children who have been maltreated. Child Development 69,
Brendgen, M., Vitaro., F., Bukowski, W. M., Doyle, A. B., & Markiewicz, D. (2001).
Developmental profiles of peer social preference over the course of elementary
school: Associations with trajectories of externalizing and internalizing behavior.
Developmental Psychology, 37, 308-320.
Christenfeld, N. & Gerin, W. (2000). Social support and cardiovascular reactivity.
Biomedecine & Pharmacotherapy,54, 251-257.
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, 13,
Cicchetti, D. (1994). Advances and challenges in the study of the sequelae of child
maltreatment. Development and Psychopathology, 6,1-3.
Cicchetti, D., Lynch, M., Shonk, S., & Manly, J.T. (1992). An organizational perspective on
peer relations in maltreated children. In R. D. Parke & G.W. Ladd (Eds.) Family-peer
relationships: Modes of linkage (pp. 345-383). Hillsdale, NJ: Erlbaum.
Cicchetti, D., Barnett, D., Rabideau, G. J., & Toth, S. (1991). Toward the development of a
transactional model of risk taking and self-regulation: Illustration through the study of
maltreated children. In L. Lipsitt & L.Mitnick (Eds.), Self-regulatory behavior and risk
taking: Causes and consequences (pp. 165-198). Norwood, NJ: Ablex.
Courtney, M. E., & Piliavin, I. (1998). Foster youth transitions to adulthood. Madison,Wl:
School of Social Work and Institute for Research on Poverty, University of Wisconsin
Crazier, J. C., & Barth, R. P. (2005). Cognitive and academic functioning in maltreated
children. Children and Schools, 27, 197- 206.
Dishon, T. J. (1990). The family ecology of boys peer relations in middle childhood.
Child Development, 61,874-892.
Egeland, B., Yates, T., Appleyard, K., & van Dulmen, M (2002). The long- term
consequences of maltreatment in the early years: A developmental pathway
model to antisocial behavior. Childrens Services: Social Policy, Research,
and Practice. 5, 249-260.
Erdfelder, E., Faul, F., & Buchner, A. (1996). GPOWER: A general power analysis program.
Behavior Research Methods, Instruments, & Computers, 28,1-11.
Ertel, K. A., Glymore, M. M., & Berkman, L. F. (2009). Social networks and health
A life course perspective integrating observational and experimental evidence.
Journal of Social and Personal Relationships, 26, 73-92.
Fantuzzo, J. W., Weiss, A. D., Atkins, M., Meyers, R., & Noone, M. (1998). A
contextually relevant assessment of the impact of child maltreatment on the
social competencies of low-income urban children. Journal of the American
Academy of Child and Adolescent Psychiatry, 37,1201-1208.
Furman, W. & Buhrmester, D. (1985). Childrens perceptions of the personal relationships in
their social networks. Developmental Psychology. 21,1016-1024.
Gramkowski, B., Kools, S., Paul, S., Boyer, C. B., Monasterio, E., Robbins, N. (2009).
Health risk behavior in foster care. Journal of Child and Adolescent Psychiatric
Nursing, 22, 77-85.
Gerin, W., Milner, D., Chawla, S., & Pickering,T. G. (1995). Social support as a moderator of
cardiovascular reactivity in women: A test of the direct effects and buffering
hypotheses. Psychosomatic Medicine,57,16-22.
Haskett, M. E., Nears, K., Ward, S. C., & McPherson, A. V. (2006). Diversity in
adjustment of maltreated children: Factors associated with resilient functioning.
Clinical Psychology Review. 26, 796- 812.
Howe, T. R. & Parke, R. D. (2001). Friendship quality and sociometric status:
between-group differences and links to loneliness in severely abused and
non-abused children. Child Abuse and Neglect, 25, 585- 606.
Inglehart, A. P. (1995). Readiness for independence: Comparison of foster care, kinship
care, and non-foster care adolescents. Children and Youth Services Review, 9, 81-
Kramer, L Gottman, J. M. (1992). Becoming a sibling: With a little help from my
friends. Developmental Psychology, 28, 685-699.
Kinard, E. M. (1999). Percieved social skills and social competence in maltreated
children. American Journal of Orthopsychiatry, 69, 465- 481.
Kortenkamp, K. & Ehrle, J. (2002). The well-being of children involved with the child
welfare system: A national overview. Washington, DC: Urban Institute.
Leathers, S. J. (2005). Seperation from siblings: Associations with placement adaptation and
outcomes among adolescents in long-term foster care. Children and Youth Services,
Legault, L., Anawati, M., & Flynn, R. (2006). Factors favoring psychological resilience among
fostered young people. Children and Youth Services Review. 28, 1024-1038.
Leve, L.D., Fisher, P.A., & DeGarmo, D.S. (2007). Peer relations at school entry: Sex
differences in outcomes of foster care. Merrill-Palmer Quarterly. 53, 557 577.
McCoy, J. K., Brody, G. H., & Stoneman, Z. (1994). A longitudinal analysis of sibling
relationships as mediators of the link between family processes and youth best
friendships. Family Relations, 43, 400-408.
Mueller, E., & Silverman, N. (1989). Peer relations in maltreated children. In D. Cicchetti and
V. Carlson (Eds.), Child Maltreatment: Theory and research on the causes and
consequences of child abuse and neglect (pp. 529-578). New York : Cambridge
Nilsen, W. (2007). Fostering futures: A preventative intervention program for school-
aged children in foster care. Clinical Child Psychiatry. 12(1), 45- 63.
Panzarella, C., Alloy, L. B., & Whitehouse, W. G. (2006). Expanded hopelessness
Theory of depression: On the mechanism by which social support protects
against depression. Cognitive Theory Research, 30, 307-333.
Parker, J. G., & Asher, S. R. (1987). Peer relations and later personal adjustment: Are
low-accepted children at risk? Psychological Bulletin. 102, 357- 389.
Parker, J. G., & Herrera, C. (1996). Interpersonal processes in friendship: A comparison of
abused and non-abused childrens experiences. Developmental Psychology, 32,
Pecora, P., Whittaker, J.D, Maluccio, A.N., & Barth, R. (2000). The child welfare
challenge.Policy, practice and research. New York: Aldine De Gruyter.
Pribesh, S.,& Downey D., 1999. Why are residential and school moves associated with poor
school performance? Demography, 36, 521-534.
Price, J. M., & Brew, V. (1998). Peer relationships of foster children: Developmental and
mental health service implications. Journal of Applied Developmental Psycholgy, 19,
Prince, J. (1996). Friendships of maltreated children and adolescents: Contexts for
expressing and modifying relationship history. In W.M. Bukowski, A. F.
Newcomb, & W.W. Hartup (Eds.), The company they keep: Friendship in
childhood and adolescence (pp. 262- 285). New York: Cambridge University
Rutter, M. (1990). Psychosocial resilience and protective factors. In J. Rolf, A. Masten, D.
Cicchetti, K. Nuechterlein, & S. Weintraub (Eds.), Risk and protective factors in the
development of psychopathology (pp. 181-214). New York, NY: Cambridge
Salzinger, S., Feldman, R. S., Hammer, M., Rosario, M. (1993). The effects of physical
abuse on childrens social relationships. Child Development, 64, 169-187.
Scannapieco, M. (1999). M. Kinship care in the public child welfare system: A systematic
review of the research. In: R.L. Hegar and M. Scannapieco, (Eds), Kinship care:
Policy, practice & research, (pp.141-154). Oxford, New York.
Seeman, T. E., Lusignolo, T. M., Albert, M. & Berkman, L. (2001). Social relationships, social
support, and patterns of cognitive aging in healthy, high functioning older adults:
MacArthur studies of successful aging. Health Psychology, 20, 243-255.
Shook, J. J., Vaughn, M. G., Litschge, C., Kolivoski, K., & Schelbe, L. (2008). The
importance of friends among foster youth aging out of care: Cluster profiles
of deviant peer affiliations. Children and Youth Services Review, 31, 284-291.
Smith, M.C., (1995). A preliminary description of non-school- based friendship in young high
risk children. Child Abuse and Nelgect. 19, 1497-1511.
Snyder, J., Schrepferman, L., Oeser, J., Patterson, G., Stoolmiller, M., Johnson, K., et al.
(2005). Deviancy training and association with deviant peers in young children:
Occurances and contribution to early-onset conduct problems. Development and
Psychopathology. 17, 397-413.
Stein, E., Evans, B., Mazumdar, R., & Rae-Grant, N.(1996). Mental health in
foster care: A comparison with community and clinical samples.
Canadian Journal of Psychiatry. 41,385-391.
Taussig, H. N. (2002). Risk behaviors in maltreated youth placed in foster care: a
longitudinal study of protective and vulnerability factors. Child Abuse &
Taussig, H. N., Culhane, S.E., & Hettleman, D. (2007). Fostering healthy futures: An
innovative preventative intervention for preadolescent youth
in out-of-home care. Child Welfare League of America, 86, 113-131.
Thornberry, P. T., Ireland, T. O., & Smith, C. A. (2001). The importance of timing:
The varying impact of childhood and adolescent maltreatment on multiple
problem outcomes. Development and Psychopathology, 13, 957- 979.
Vondra, J., Barnett, D., & Cicchetti, D. (1989). Percieved and actual competence among
maltreated and comparison school children. Development and Psychopathology. 1,
Wallace, R. B. (1986). Established populations for epidemiologic studies of the elderly:
Resource data book (NIH Publication No. 86-2443). Bethesda, MD: National Institute
Wodarski, J. S., Kurtz, P. D., Gaudin, J. M. & Flowing, P. T. (1990). Maltreatment and the
school-age child: Major academic, socioemotional, and adaptive outcomes. Social
Work, 35, 506-513.