Marys Medicine

Fkn.org.il

This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier's archiving and manuscript policies are
encouraged to visit:

Author's personal copy
Child Abuse & Neglect 33 (2009) 429–440 Contents lists available at ScienceDirect Child Abuse & Neglect School functioning of children in residential care: The contributions ofmultilevel correlates Shalhevet Attar-Schwartz Research Group: Mental Health and Well-Being in Childhood and Adolescence, School of Social Work and Social Welfare, The Hebrew University of Jerusalem,Mount Scopus, Jerusalem 91905, Israel Article history: Objective: This study, using an ecological approach, examines the relationships between
Received 21 June 2007 problems in school functioning (including academic and behavior problems) of children in Received in revised form 11 December 2008 residential care with a number of variables describing the child and the care setting.
Accepted 17 December 2008 Methods: The study reports on 4,061 children and youth (ages 6–20) in 54 Israeli residen-
Available online 8 July 2009 tial care facilities supervised by the Ministry of Welfare. It is based on data derived from anongoing system of monitoring care based on annual reports by social workers on children in Keywords:Children in care care settings. Additionally, data on the characteristics of the settings were collected through a structured questionnaire completed by the supervisors at the Ministry of Welfare. Hierar- School performance chical Linear Modeling (HLM) was utilized to examine how characteristics of the individual Parental visitation children and the care settings were related to problems in school functioning among the Results: Most of the children (about 62%) had at least one problem in school functioning.
The most vulnerable children were boys, children who were taken from parental homes
by court decree, children with problems in quality of contact with their biological parents,
and children who stayed in the care setting for shorter periods. The settings' characteristics
most associated with poor performance at school are group structure (vs. mixed and family
home structures), higher levels of peer violence, fewer after-school activities, and settings
in which children tend to stay for shorter periods of times.
Conclusions: The findings demonstrate the need for an ecological perspective in addressing
children's problems in school functioning within the care system. The results help to identify
the types of placements that should benefit from additional resources in order to promote
adaptive performance in school among the children.
Practice implications: Social workers in residential care should give high priority to chil-
dren's positive academic involvement. The study demonstrates the need for identifying the
intersection of the individual, familial and institutional contexts in which problems in school
functioning are more prevalent. Therefore, it is important to allocate sufficient resources to
care settings which serve these children. The study suggests some priorities and directions
for policy and practice with children in residential care.
2009 Elsevier Ltd. All rights reserved.
Children who have been removed from home due to lack of adequate parental care rarely have access to consistent educational support, which is taken for granted by most children living with their parents (Shin, 2003). For these children,low school performance might be only one disadvantage among many (Ajdukovic & Franz, 2005), but it has some of the mostserious consequences for their life chances in adulthood (e.g., Schiff & Benbenishty, 2006).
0145-2134/$ – see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.chiabu.2008.12.010 Author's personal copy
S. Attar-Schwartz / Child Abuse & Neglect 33 (2009) 429–440 The aim of this study was to examine problems in school functioning among 4,061 children (in this paper, the term "children" includes adolescents) living in 54 residential care settings (RCSs) for children at risk under the responsibility ofthe Welfare Ministry in Israel.
Studies consistently report higher rates of academic and behavior problems at school among children in substitute care.
In addition, these children are at higher risk for entering adulthood with fewer educational qualifications than those who arenot in care (Emerson & Lovitt, 2003; Jackson, 1994; Jackson & Martin, 1998; Mitic & Rimer, 2002; Newburn, Shiner, & Young,2005; Schiff & Benbenishty, 2006; Stein, 2006; Vinnerljung, Oman, & Gunnarson, 2005; Vorria, Wolkind, Rutter, Pickles, &Hobsbaum, 1998a; Zetlin, Weinberg, & Kimm, 2005). Moreover, research shows that the academic outcomes of children incare are poor even in comparison to other children with similar socio-economic backgrounds (e.g., Vinnerljung et al., 2005).
Children in care also have higher rates of truancy, disciplinary referrals, performance below grade level, repetition of at leastone class, leaving school early, and dropping out (Brodie, 2000; Flynn & Biro, 1998; Sawyer & Dubowitz, 1994). Additionally,large percentages of children in care are classified as having special education needs (Shin, 2003; Zetlin et al., 2005). In Israel,about 40% of children in RCSs are classified as having special education needs, compared with less than 5% of the generalpopulation (National Council for the Child, 2006).
Contributing to the many risk factors for problems at school are, first, pre-care experiences, including: (a) high rates of parental abuse, which may have a lasting influence on the child's intellectual abilities (Vinnerljung et al., 2005); (b) over-representation of poor children in the care system; socio-economic difficulties have been associated in many studies withschool failure (Pagani, Boulerice, Vitaro, & Tremblay, 1999); (c) special emotional challenges experienced by children in care,which can negatively influence their ability to persevere with their studies (Aldgate, Colton, Ghate, & Heath, 1992; Mitic &Rimer, 2002); (d) the older age at which many children enter the care system, which it is argued, increases the intensity ofthe educational deficiencies and problems they bring with them.
It is also important to address children's in-care experiences as contributing to their difficulties at school. First, the system is characterized by high levels of residential instability, which result in multiple school transfers experienced by childrenin care, which disrupt the educational process (Emerson & Lovitt, 2003; Pecora et al., 2006; Zetlin et al., 2005). In addition,social workers tend to focus on emotional-behavioral difficulties of children in care, assigning low priority to educationalachievement (Jackson, 1994; Zetlin et al., 2005). Moreover, because institutional caregivers often lack higher educationthemselves, they have in many cases low expectations and little understanding of what is involved in academic work (Jackson& Martin, 1998). The RCSs often do not offer intellectual or academic support, such as assistance with homework and theprovision of books. Additional problems arise from interactions between the care and education systems and from thestigmatization of children in care at school (Brodie, 2000).
Fulcher (2001) and others (Vorria et al., 1998a; Whittaker, 1978) suggest using ecological approaches to examine the different effects of child, family, and RCS characteristics on children's status in RCSs. As obvious as this may seem, moststudies of children in RCSs overlook the multilevel nature of the phenomenon, and focus either on the children or on the caresettings as units of analysis. This study examines child, family, and RCS factors to explain variance in children's problems inschool functioning. The following sections present a review of the literature on the study correlates.
Level-1 correlates: Child and biological family characteristics Some research focuses on individual characteristics associated with problems in school functioning among children in care. For example, previous studies consistently show that boys have more problems in school performance and adjustmentthan girls, both in care and after leaving care (Hukkanen, Sourander, Bergroth, & Piha, 1999; Pecora et al., 2006; Schiff &Benbenishty, 2006; Schiff, Nebe, & Gilman, 2006; Vorria et al., 1998a; Wahba, 2003; Yule & Raynest, 1972). Studies also showthat older children (Glisson, Hemmelgarn, & Post, 2002; Heflinger, Simpkins, & Combs-Orme, 2000; Hukkanen et al., 1999)and immigrant children (Bates et al., 2005) in care have more psychosocial and school functioning problems.
Studies on the relationship between children's length of stay in care and their emotional-behavioral and academic func- tioning yield inconsistent results. Some researchers find a positive relationship between the two factors (Attar-Schwartz,2008; Davidson-Arad, 2005; Gilman & Hendwerk, 2001; Zemach-Marom, Fleishman, & Hauslich, 2002). Other studies, how-ever, find no association (Heflinger et al., 2000; Hukkanen et al., 1999) or a negative association (Hussey & Guo, 2002;Landsman, Groza, Tyler, & Malone, 2001; Mosek, 1993) between length of stay and children's outcomes.
Few studies in Israel have investigated the court's involvement in the removal of children from their homes. Often no association was found to link court involvement with the children's outcomes in care (e.g., Schiff & Benbenishty, 2006). About75% of all out-of-home placements in Israel are made with parental approval and do not involve the courts (National Councilfor the Child, 2006). The professional imperative is to make such decisions with as much parental consent as possible. Thecourt is involved only in cases in which the child is assessed as being at imminent risk. It can then order the removal of thechild from home without parental consent.
The child's biological family characteristics were examined by several studies. A large percentage of children in care come from single-parent families. In Israel, only about 38% of the biological parents of children in RCSs are married to each otheror living in the same household (National Council for the Child, 2006). Numerous studies show that children from single-parent families in the community as well as in out-of-home care have more psychosocial and educational difficulties than dochildren from two-parent households (e.g., Attar-Schwartz, 2008; Berger, 2004; Hegar & Yungman, 1989; Vorria, Wolkind,Rutter, Pickles, & Hobsbaum, 1998b; Wolock & Horowitz, 1984).
Author's personal copy
S. Attar-Schwartz / Child Abuse & Neglect 33 (2009) 429–440 Research has also looked at parent-child visitations in relation to children's functioning. Findings on the frequency of parent-child contact are inconsistent. On the one hand, some studies show a positive association between the frequency ofparent-child encounters and the child's well-being and functioning while in care and after leaving care (Cantos, Gries, &Slis, 1997; Fanshel & Shinn, 1978; Landsman et al., 2001; McWey, 2000; Oyserman & Benbenishty, 1992). Others find norelationship between these variables (Hukkanen et al., 1999; Vorria et al., 1998a). However, findings on the quality of contactbetween children in care and their parents are more consistent. Several studies report positive relationships between thequality of parent-child contact and the child's educational and psychosocial status in care and after leaving care (Attar-Schwartz, 2008; Hukkanen et al., 1999; McWey, 2000; Weiner & Kupermintz, 2001).
Level-2 correlates: Setting factors Some research focuses on institutional factors that are associated with the emotional-behavioral and school functioning problems of children in care. The following section presents a review of the literature on the institutional correlates includedin this research.
RCS structure. There are several possible institutional structures to accommodate children at risk in Israel. Traditional group institutions take care of large numbers of children who reside in small groups. Each group of children has a social worker, aswell as institutional caregivers in changing shifts. Clusters of family homes in a shared facility ("familial settings"), in which amarried couple with biological children cares for a small number of children at risk (up to 10) who share the same family unit.
All family units in the facility use the same administrative, cleaning, and general support services. Other settings include smallfamily-like settings in the community, combinations of residential group and family-like units (hybrid) in the same structure,and other arrangements (Children & Youth Service, 2005). There is some evidence that the emotional-behavioral status ofchildren varies across different institutional structures (e.g., Ajdukovic & Franz, 2005; Heflinger et al., 2000). For example,Mosek, Taus, and Shomodi (1997) describe an improvement in the well-being of Israeli children who were transferred froma traditional group setting to a cluster of family group homes.
RCS size. Research shows that smaller RCSs are warmer, more individualized, and promote more consistent relationships with the primary caregiver than large RCSs. Large numbers of children in RCSs may generate higher levels of noise, haveless personal space and privacy, and more interpersonal friction (Barter, Renold, Berridge, & Cawson, 2004; Gibbs & Sinclair,1999). These aspects may undermine children's sense of security, and inhibit their school performance (Bailey, 2002). Itshould be noted, however, that there are contradictory findings with regard to the effects of institution size. For example, ininterviews with children's home staff conducted by Barter et al. (2004), several of the staff members interviewed felt thatlarger setting was an advantage because it allowed children more privacy and allowed a range of activities to take placewithout children competing for space.
Physical conditions in the RCS Suitability of the physical environment to children's needs. Research shows that the physical characteristics of a facility influence the behavior and mental health of its residents (Bailey, 2002). For example, Tear et al. (1995) found that socialdensity in youth crisis shelters was associated with increased levels of behavior problems among the residents. Barter et al.
(2004) found links between children's life satisfaction in RCSs and their ranking of the general décor and amenities in thefacility (see also Shalom & Apple, 1994).
Food. Studies in the community have found associations between food insufficiency and psychosocial and academic problems among children (Alaimo, Olson, & Frongillo, 2001; Slack & Yoo, 2005). For children in RCSs, food has symbolic andemotional meanings of acceptance, domesticity, warmth, and care, beyond its nutritional value (Bailey, 2002; Rose, 1992).
This study examines associations between the quantity and variety of food and children's problems in school functioning inresidential care facilities.
Institutional functioning and climate characteristics After-school activities. Studies show that children in RCSs that offer more leisure-time activities and academic tutoring show greater satisfaction with their life at the RCS and achieve greater educational success in adulthood (Gilligan, 2007;Jackson & Martin, 1998; Shalom & Apple, 1994). Research also indicates that extracurricular interests and hobbies facilitatecontact with a wider range of non-care people (Pecora et al., 2006; Shin, 2003) and help to promote children's feelings ofcompetence (e.g., Romans, Martin, Anderson, O'shea, & Mullen, 1995).
Peer violence. The few studies on peer violence in care settings reveal alarming evidence regarding the high prevalence and destructive consequences of the phenomenon (e.g., Barter et al., 2004; Lees, Morris, & Wheatley, 1994; MacLeod, 1999).
For example, a follow-up study of children's homes for young people in the UK found that residents who reported beingbullied while in care were less adjusted to various aspects of life both in and after care (Gibbs & Sinclair, 1999).
Author's personal copy
S. Attar-Schwartz / Child Abuse & Neglect 33 (2009) 429–440 Table 1
Summary statistics of the study variables.
Dependent variable Problems in school functioning (0 to 2 academic and behavioral problems) Child-level correlates Special education needs (no special needs) (0 = special education needs, 1 = no special needs) (1 = female, 0 = male) Immigration status (Israeli-born) (1 = Israeli-born, 0 = immigrant) Length of stay at the current RCS Court intervention (statutorily) (1 = statutorily, 0 = voluntarily) Family composition (two-parent families) (1 = two-parent family, 0 = single-parent family) Problems in frequency of parent-child visitations (normal) (1 = normal frequency, 0 = problematic frequency) Problems in quality of parent-child visitations (normal) (1 = normal quality, 0 = problematic quality) Contextual (RCS-level) correlates Setting type (therapeutic) (0 = rehabilitative, 1 = therapeutic) Setting structure (familial settings)a (1 = familial setting, 0 = group and "other" settings) Setting structure ("other" settings)a (1 = "other", 0 = group and familial settings) Suitability of the physical environment (1 = low suitability to 5 = high) (1 = low quality to 5 = high) After-school activities (1 = low level of activities to 5 = high level) (1 = strongly agree to 5 = strongly disagree) Note. N = 4,061 children, 54 residential care settings.
a Dummy variable, reference group = groups settings.
In summary, this study focuses on problems in school functioning (i.e., academic and behavioral problems at school) among children in RCSs. It improves on previous studies by addressing this issue from a multilevel perspective: it examinessimultaneously the effects of child and family characteristics and RCS factors on children's problems in school functioning,and the unique contribution made by both child and RCS variables in explaining the variance in school problems betweenchildren and between RCSs.
The current study examines the problems in school functioning of children in rehabilitative and therapeutic RCSs. In Israel, about 9,000 children live in out-of-home care, with about 80% of them placed in RCSs and the rest in family fostercare (National Council for the Child, 2006). Most of the children in Israel's care system are placed in RCSs that belong to thewelfare system and are specifically designed for children at risk. They are placed in RCSs through interdisciplinary decisioncommittees which operate within the framework of the Welfare Ministry. Depending on their needs and family background,the committees may decide to refer children at risk to rehabilitative, therapeutic, or post-hospitalization settings (Dolev,Benbenishty, & Timar, 2001).
About 40% of children in welfare RCSs are placed in rehabilitative RCSs. These settings serve children with satisfactory developmental potential who are removed from their homes because of their parents' inability to cope with their emotionaland educational needs. About 30% of children in welfare RCSs live in therapeutic RCSs, which are designed to treat childrenwith extreme family problems. Other settings, not included in the current study are post-hospitalization settings, whichserve nearly 5% of children in welfare settings who suffer from severe psychiatric problems and educational settings thatserve children from underprivileged backgrounds whose parents choose this option because they believe it will provide themwith a better education. Rehabilitative and therapeutic facilities are expected to differ from each other in the needs of thechildren under their responsibility. However, research shows that in some cases, the functioning of children in rehabilitativeand therapeutic settings is in a similar level (Dolev & Barnea, 1996). For cultural and historical reasons (Jaffe, 1982), 90% ofthe children in welfare RCSs are Jewish and 10% are Arab. This study focuses on Jewish children.
The original database of Jewish children in rehabilitative and therapeutic settings contained information on 4,631 children.
However, 316 children (6.82%) were excluded because they were in RCSs with fewer than 25 residents, and an additional209 (4.51%) were from four RCSs on which information about the RCS characteristics was unavailable for administrativereasons (e.g., turnover of supervisors). There were no significant differences between the socio-demographic characteristicsof the excluded children and the study's sample. An additional 45 children in the database were excluded from the study,because they were under school age. As shown in Table 1, the study is based on social workers' reports on 4,061 children in21 therapeutic RCSs (39% of the settings) and 33 rehabilitative RCSs (61% of the settings). Most of the children were between6 and 18 years old (M = 13.32, SD = 3.11). However, some participants were older, having repeated at least one grade level: 16participants (0.4% of the total sample) were 20 years old and 90 (2.2%) were 19 years old.
Author's personal copy
S. Attar-Schwartz / Child Abuse & Neglect 33 (2009) 429–440 Two sources of data were used in this study: (a) Existing database on children in RCSs. This Israel Welfare Ministry database consists of background information on children in RCSs and an annual status assessment made by each child's social worker. The social workers complete a structuredquestionnaire that provides an assessment of problems across an array of dimensions (e.g., depression, problems in schoolfunctioning, aggressive behavior, etc.) based on several sub-scales. Test-retest reliability examinations show medium tohigh overall reliability for the instrument (Walk, Zemach-Marom, Amihai-Ben Yaakov, Hauslich, & Branz, 2004). Therationale, content, and process of data collection are derived from the RAF method (Regulation, Assessment, and Follow-up; Zemach-Marom et al., 2002) of surveillance and continuous improvement of quality of care, which is based mainlyon the "tracer" approach (Kessner & Kalk, 1973). This study uses data collected in 2003–04 among children from theJewish sector.
(b) Questionnaires completed by RCSs quality control supervisors. Supervisors were asked to complete a questionnaire on each of the RCSs under their supervision because an excessive amount of missing information concerning RCSs characteristicswas found in the RAF database. The questionnaire was patterned after the reports required by the RAF system to describethe characteristics of each RCS. The instrument included questions about the organizational background characteristicsof the RCS (structure and size) and assessments of several domains such as physical conditions, environment and policycharacteristics at the RCS. Information was requested only on RCSs with more than 25 children (including several familygroup homes managed and supported by the same personnel). Overall, information on 54 RCSs (90% of the total numberof eligible therapeutic and rehabilitative facilities) was received. All 11 regional supervisors participated, each completingbetween 2 and 8 questionnaires (median = 5 questionnaires).
The research was approved by the Ethics Committee of the Hebrew University of Jerusalem. The data obtained from the Welfare Ministry does not include identification details of the children. Hence, the confidentiality of participants wasensured.
Dependent variable: Problems in school functioning. In the child questionnaire, social workers were asked about whether the child was classified as having special education needs, and then asked to complete questions regarding the following:(a) Achievement at school: For children in regular education they were asked whether the child's average academic gradesare low, and whether the child shows an educational gap of at least 2 years compared with other students in his/her class.
For children classified as having special education needs, they were asked to indicate the extent to which the child fulfilshis/her potential. According to the RAF method, children are considered to have an achievement problem if they have lowscores on average or if they have an educational gap of at least 2 years (no special education needs); and if they do not fulfill,or only partially fulfill, their academic potential (special education needs). (b) Behavior problems at school: Social workerswere asked whether the child has a problem of non-attendance or truancy, is late for classes, does not do homework or othertasks, and whether the child has discipline problems. The RAF method considers children to have a behavior problem if theyare late or do not attend school more than 5 consecutive days or 10 non-consecutive days a month, if they frequently fail todo their homework and other tasks, or if they frequently have discipline problems at school.
The dependent variable in this study is based on the sum of the scores of children's behavior and achievement problems at school (˛ = 0.56). The summative index for children's problems in school functioning is therefore based on the followingscale: 2 = the child has both academic and behavioral problems; 1 = the child has problems in one of those areas; 0 = the childhas no problems in either of the areas.
Control variable: Special education needs. Some of the children in RCSs are classified as having special education needs by professionals in the welfare system. They may attend special classes within the RCS or special classes in the community,either in special or in regular schools. Others, with no special education needs, attend regular schools in the community.
The classification of children as having special education needs was controlled for in this study, as it was expected it wouldbe related to the children's problems in school functioning and not randomly distributed across other correlates related toproblems in school functioning examined in the current study.
Child-level correlates. Background information: Social workers reported on the child's gender, age, immigration status (Israeli-born children vs. immigrant children) and length of stay (in years) at the RCS. Other background informationincluded family composition (two-parent families—married or living together; or single-parent families—widowed, divorced,or never married); and court involvement in RCS placement (children placed by court decree vs. those placed with no courtinvolvement).
Problems with frequency of parent-child visitation: This variable is based on two questions: (a) whether the child's parents attend events held by the RCS or arrive of their own volition and (b) to what extent the child visits his or her parents' home.
According to the RAF method, children are considered to have a problem if their parents do not attend the RCS events, or ifthe children seldom or never visit their parents.
Author's personal copy
S. Attar-Schwartz / Child Abuse & Neglect 33 (2009) 429–440 Problems with quality of parent-child visitation. This variable is based on three questions: (a) whether the child has emo- tional/behavioral problems after visiting the parents' home; (b) whether the child is abused or suspected of being abusedwhile visiting home; and (c) whether the child is subject to considerable neglect when visiting the parents. Children are con-sidered to have a problem if they experience emotional-behavioral difficulties following every home visit or every secondvisit, or if they are victims or suspected victims of abuse or neglect during home visits.
Setting-level characteristics. Organizational factors: These factors include RCS type (rehabilitative or therapeutic), RCS structure (two dummy variables comparing between "familial" settings, "other", and "group" settings—see definitions in theIntroduction), and RCS size (number of children).
Physical conditions: Supervisors were asked to rate their level of agreement with a series of statements describing the physical conditions at the RCS. Responses were based on a 5-point scale ranging from 1 = "highly applicable" to 5 = "doesnot apply": (a) Suitability of the physical environment to the children's needs: this index (˛ = 0.80) is based on the mean scoreof three statements regarding the extent to which the RCS suits the children's needs. It includes two general statementsabout the physical conditions at the RCS (e.g., "the physical setting is suitable for the children's residential needs") and onequestion about the existence of recreational facilities at the RCS; (b) Food: this variable is based on one statement regardingthe adequacy of the amount and variety of food served in the RCS.
Functioning and environment characteristics: Supervisors were asked to rate their level of agreement with a series of statements describing the functioning and climate features of the RCS on a 5-point scale ranging from 1 = "highly applicable"to 5 = "does not apply to this facility": (a) After-school activities: this index (˛ = 0.78) is based on the mean of responses tothree statements regarding the extent to which the RCS is carrying out leisure-time and academic activities, such as personalacademic tutoring; (b) Peer violence: this index (˛ = 0.84) is based on the mean of responses to three statements regardingthe extent of verbal, physical, and sexual violence among children at an RCS.
Analytic plan. First the study presents the summary descriptive information about the academic and behavior problems of children in the RCSs examined in the current study and about the correlates of these problems. The main analytical toolused in the present study was Hierarchical Linear Modeling (HLM) which is a technique that allows the investigation ofdata organized hierarchically on more than one level (Bryk & Raudenbush, 2002; Khoury-Kassabri, 2006). In this study,children were nested within RCSs: the level-1 model represents the relationships between the child-level characteristics andproblems in school functioning; the level-2 model captures the relationships of RCS-level variables and children's problems inschool functioning. The HLM method provides separate estimates of the variance in children's problems in school functioningexplained by the children's characteristics and by the features of the RCSs.
In this analysis, child-level characteristics were first entered into the HLM equation in a sequential manner from the personal characteristics of the child to the characteristics of his/her contact with the biological parents. Then the contributionof RCS-level variables was examined after controlling for child-level variables. RCS-level factors were added to the equationfollowing a hierarchy from the most general context of the RCS (organizational factors) to the contexts in which RCSs mayhave the most direct control (functioning and environment).
Child-level variables were centered around the group mean because one of the research concerns was whether children with certain characteristics are associated with higher or lower levels of problems in school functioning within their own RCS.
For instance, do children who stay for longer periods have different levels of problems in school functioning than others intheir RCS? This type of question is best answered by centering around the group mean. However, this approach overlooksthe variation of these centered variables (e.g., length of stay and gender) between the RCSs participating in the study. Toovercome this problem, as required by the HLM method, the group means of the centered variables (e.g., the RCS' averagelength of stay and the percentage of females in the RCS) were reintroduced as second-level (RCS-level) variables (for furtherexplanation see Bryk & Raudenbush, 2002).
In this study, slopes that were not much different across RCSs (i.e., having insignificant variance components) were fixed.
For this research, the slopes of special education needs and length of stay were allowed to vary at the RCS level.
Based on the reports of the social workers, 61.7% of the children in the examined RCSs evidenced at least one problem in their functioning at school (M = 0.92 problems, SD = .82), including low achievement (51.6%) and behavior problems (40.3%).
As shown in Table 1 the average of length of stay in the RCS was 2.94 years (SD = 2.36). About 34% of the children were classified as having special education needs, 26.8% of them were removed from home by court decree, only 36.2% of themcome from two-parent families, and about 78% of them are Israeli-born children. According to the social workers' reports,about 33% of the children suffer from problems in the frequency of contact with their parents and about 41% suffer fromproblems in the quality of visitations. Most of the RCSs examined were traditional group settings (about 60%), with anaverage population of about 96 children (SD = 66.79). On average, these RCSs were described rather favorably as havingsuitable physical environment, food, and a range of after-school activities (means of 4.03, 4.20, and 4.01, respectively, on ascale from 1 = "low quality" to 5 = "high quality").
This study examines RCSs context variables that may be associated with children's problems in school functioning. This question assumes that RCSs differ with regard to the problems in school functioning of their residents. To address this Author's personal copy
S. Attar-Schwartz / Child Abuse & Neglect 33 (2009) 429–440 question, a fully unconditional two-level model (i.e., a model with no independent variables) was utilized. This providesuseful preliminary information on how much variance in children's problems in school functioning lies within RCSs andbetween RCSs. The results indicate that a considerable amount of the variance in children's problems in school functioningwas accounted for by variance between RCSs in reports of problems in school functioning (Intraclass Correlation Coefficient;ICC): 11.56% (Chi square = 452.219, df = 53, p < 0.001). Hence, a two-level analysis was conducted.
It should be noted that an additional preliminary test was conducted in order to examine whether there is variability between the 11 supervisors in their reports of problems in school functioning. The ICC of this examination was insignificant(ICC = 0.0009%; Chi square = 547.695, df = 10, p = 0.267). This finding means that it did not make a difference who the supervisorwas; therefore there was no need to control for the identity of the reporting supervisors.
Multi-level model Child-level factors. The study estimated the relationships between child-level correlates and the child's problems in school functioning, after controlling for the child's education needs (special education needs vs. no special needs). The results inTable 2 show that, according to the social workers' reports, boys had more problems in school functioning than did girls.
Children whose period of residence at the current RCS was longer and those who had been removed from their parentalhome with no court decree had lower levels of problems in school functioning. Additionally, it was found that the child'simmigration status, age, family composition, and problems in frequency of contact with parents were not associated withproblems in school functioning. However, children suspected of being mistreated by their parents during visits had moreproblems in school functioning.
To examine the contribution of child-level variables to the explained variance in social workers' reports on problems in school functioning within the RCSs, the child-level correlates that were found to have significant associations with problemsin school functioning (see Table 2) were entered into the equation. The results show, overall, that child-level variablesexplained 4.60% of the variance within RCSs in children's problems in school functioning.
Setting-level factors. To estimate how the characteristics of the RCSs relate to children's problem in school functioning, the relationships between RCSs' means of problems in school functioning and RCS-level variables were examined (Table 3). Aftercontrolling for the RCS type (rehabilitative or therapeutic settings), the results revealed that children from RCSs with highproportions of children with no special education needs, Israeli-born children, children with no problems in quality of parent-child visitations and with higher average of length of stay of the residing children, showed fewer problems. Furthermore,children in small family-like settings and hybrid structures ("other") showed fewer problems in school functioning than didchildren in group settings. No significant association was found between the number of children in the RCS and problems inschool functioning. Regarding the physical conditions in the RCS, the findings indicate that the more the physical conditionsof the RCS suited the children's needs, the fewer problems the children had at school. However, diversity and quantity of foodin RCSs was not associated with the children's problems in school functioning. In addition, problems in school functioningwere less frequent in RCSs that emphasized after-school recreational activities and educational advancement of children,and that had lower levels of peer violence.
The variance explained by each setting-level correlate was examined. The variables consisted of those that showed sig- nificant relationships with problems in school functioning (see Table 3). Setting-level variables were found to explain 38.76%of the variance between the RCSs in problems in school functioning. The average length of stay of children residing in theRCSs and the percentage of children with no special education needs each explained more than 10% (11.23% and 10.45%,respectively). The setting structure ("other" vs. group settings) explained 6.30% of the variance, and peer violence at theRCS explained 4.30% of the variance between RCSs in children's problems in school functioning. The percentage of childrenwith problems in quality of parent-child visitations contributed 4.03% to the variance and after-school activities contributed2.45% to the variance. The suitability of the physical conditions to children's needs did not contribute to the explainedvariance.
Table 2
The relationships between children's problems in school functioning and child-level factors.
Independent variables-SF slopes, G10 Special education needs (no special needs) Immigration status (Israeli-born) Length of stay at the current RCS (years) Court intervention (statutorily) Family composition (two-parent family) Problems in frequency of parent-child visitations (normal) Problems in quality of parent-child visitations (normal) * p < 0.05.
** p < 0.01.
*** p < 0.001.
Author's personal copy
S. Attar-Schwartz / Child Abuse & Neglect 33 (2009) 429–440 Table 3
The relationships between setting-level characteristics and settings' means of problems in school functioning.
RCS type (therapeutic) Children aggregative variables Percentage of children with no special education needs Percentage of females Average age of children Percentage of Israeli-born children Average length of stay of children Percentage of children with statutory removals from parental home Percentage of children from two-parent families Percentage of children with normal frequency of visits with parents Percentage of children with normal quality of visits with parents Organizational factors Setting structure (familial settings)a Setting structure ("other" settings)a Physical conditions Suitability of the physical environment Functioning and environment After-school activities a Dummy variable, reference group = groups settings.
* p < 0.05.
** p < 0.01.
*** p < 0.001.
Problems in school adjustment have been identified by researchers and policy makers as a major difficulty for children in care. This disadvantage has most serious implications for their future life prospects for stable employment, higher education,and a range of other areas (Jackson & Martin, 1998; Rosenblatt & Rosenblatt, 1999; Schiff & Benbenishty, 2006). In recentyears there has been a consensus regarding the need to place high priority on the school involvement and progress of thesechildren.
Consistent with previous research, the current study shows that a high proportion of children in RCSs are classified as having special education needs, and that most of them (about 62%) have at least one problem related to school functioning.
The study aimed to understand what factors are associated with problems in school functioning of children in rehabilitative and therapeutic RCSs in Israel. The multilevel analysis used in the current study enabled to detect the groups of children thatare more vulnerable to problems in school functioning and the types of setting contexts associated with more problems.
Overall, the findings indicate that problems in school functioning are associated with several child (i.e., gender, length ofstay at the RCS, court involvement, and problems in the quality of parent-child visitation) and setting (i.e., RCS structure,suitability of the physical environment to children's needs, after-school activities, and peer violence) factors. The mainfindings regarding child-level and the setting-level correlates are addressed below.
This study found that boys exhibited more problems in school functioning than did girls (see also Hukkanen et al., 1999; Pecora et al., 2006; Schiff & Benbenishty, 2006; Vorria et al., 1998a). Girls appear to adjust more positively to changingevents and environments (Schiff et al., 2006; Wahba, 2003), and are more willing than boys to seek and accept both formaland informal help (e.g., Schober & Annis, 1996). The study also found that the longer children stayed in one RCS, the fewerproblems in school functioning they had. This finding is consistent with other studies (Davidson-Arad, 2005; Gilman &Hendwerk, 2001), and could signify that the children's problems are successfully addressed by the RCS after a period ofadaptation. However, causal conclusions cannot be drawn from this study: the findings should be tested and replicated withlongitudinal designs and, when possible, with experimental studies. It is also important to examine the length of stay withina wider array of care history variables (e.g., stability of placement), which might combine to explain the effects of children'scare experiences on school performance. In addition, it is interesting to see that the overall average length of stay of childrenin the RCS was found to be associated negatively with problems in school functioning, and had a considerable contributionto the variance explained between RCSs in children's problems in school functioning. This contribution on both levels isconsistent with the dual approach suggested by Bryk and Raudenbush (2002), who caution against the aggregation bias thatmay occur when a variable takes on different meanings at different organizational levels. In this context, it might indicate Author's personal copy
S. Attar-Schwartz / Child Abuse & Neglect 33 (2009) 429–440 that high turnover of children in a RCS might make it difficult for children to have consistent relationships with friends andto develop significant bonds.
The findings show that children who were taken from home by court decree have more problems in school functioning.
Therefore, this aspect might be informative of the circumstances that led to the child's placement. Studies show that thesecircumstances are linked to the child's status in care (Vorria et al., 1998b). The findings also indicate that children with poorquality of visitation with their parents had more problems in school functioning. This finding is in line with other studies(e.g., Hukkanen et al., 1999; McWey, 2000), as well as with social work policies that maintain that parent-child contact incare is important for the child's well-being and functioning and should be encouraged (e.g., Haight, Kagle, & Black, 2003).
This study is one of the few to examine the school functioning problems of children in care from a multilevel ecological perspective. It shows that children's problems in school functioning vary across different RCSs and that in RCSs with certaininstitutional characteristics children might have fewer problems in school functioning. Specifically, it was found that childrenin small family-like settings and hybrid structures had fewer problems in school functioning than did those in group RCSs(see also Mosek et al., 1997). This finding may indicate that intimate environments are necessary to promote children'sdevelopment and academic productivity and adjustment. However, no significant association was found between the numberof children at the RCS and their problems in school functioning. In other words, these findings may indicate that it is notthe number of children itself that affect the children's status but the organization of the sub-groups within the RCS. Tocompare the outcomes in various types of residential settings it is necessary in future research to gain a more comprehensiveunderstanding of the organizational and contextual aspects of each type of RCS. It may be necessary, for example, to reach abalance between keeping settings small and intimate yet large enough to ensure continued academic support and resources.
Furthermore, the superior school functioning of children in such RCSs might also be explained by selective allocation ofchildren with fewer problems to these facilities.
This study is one of the few to investigate associations between the physical conditions in RCSs and children's problems in school functioning. In RCSs with adequate physical conditions (e.g., those having recreational facilities), children hadfewer problems in school functioning. This finding is in line with research indicating that the physical characteristics of afacility play an important role in the behavior and mental health of its residents (Bailey, 2002). Stimulating and organizedenvironments that are adapted to children's needs can promote positive child behavior, including school performance. Futurestudies should continue investigating this issue with stronger designs including other physical aspects, such as social density.
It was also found that in RCSs with more after-school activities, children had fewer problems in school functioning.
Research shows that these activities may be associated with an institutional climate that focuses less on control and moreon encouraging friendly, safe, and positive relationships between institutional caregivers and children as well as among thechildren themselves. Such a climate places high priority on the children's educational success, and promotes their specialacademic needs and difficulties (Gilligan, 2000; Shin, 2003).
In settings with lower levels of peer violence, children exhibited fewer problems in school functioning (see also Gibbs & Sinclair, 1999; Tarantino, 2002). This aspect relates to the general atmosphere in the RCS rather than to specific childreninvolved in violence. Peer violence has not received the attention it deserves in the current literature although it is animportant aspect in children's lives in RCSs (Gibbs & Sinclair, 1999). It seems that an environment where peer violenceis prevalent can create a general feeling of insecurity (Benbenishty & Astor, 2005) and undermine internal resources thatshould be used to promote school functioning (Attar, 2006). This phenomenon should be addressed by practitioners andpolicy makers in RCSs.
Limitations and recommendations for future research As a cross-sectional design was used in this study, the causal direction of relationships among the variables cannot be determined. In order to establish a direct causal effect of RCSs and child factors on problems in school functioning,researchers should adopt longitudinal studies, following children in care settings for a prolonged period of time, and whenpossible employ experimental designs. The study also does not include information about factors that may have influencedthe allocation of children to their RCSs. It is possible that such factors confound with the study variables and that childrenwith better functioning end up in higher quality RCSs. Future studies should also include factors related to the allocation ofchildren to the different RCSs.
Although the study examines a wide range of potential risk factors, a substantial amount of variance between chil- dren's problems in school functioning remains unexplained. Future studies should examine the contributions of additionalimportant characteristics, such as number of placement changes, school transfers, pre-care status, the cause of referral, theallocation process leading to the placement, etc. Further contextual factors, such as the characteristics of the caregiver andcaregiver turnover rate, should also be explored.
This study relied on social workers as the sole informants on children's problems in school functioning. This poses ques- tions regarding the validity of the reports, as social workers are not educational figures in the children's lives. Future studiesshould include data from other sources, such as the children themselves and their teachers, academic records, and on-siteobservations. Furthermore, future studies should utilize well validated measures to examine RCS functioning and the child's Author's personal copy
S. Attar-Schwartz / Child Abuse & Neglect 33 (2009) 429–440 school adjustment. In addition, while the current study combines the behavioral and academic aspects of school functioningas one measure, future studies should examine these aspects separately by measures with sound psychometric properties.
Finally, social workers reported on the problems in school functioning of several children in their care and their characteris-tics, and supervisors reported on several RCSs for which they were responsible, which may cause a problem of dependencebetween observations at the child and RCS levels. In addition, the fact that in this cross-sectional design social workersreported simultaneously both on problems in children's school functioning and their characteristics may have ‘inflated' thecorrelations between these two sets of variables. Future studies should employ longitudinal designs and multiple informantsto enhance the understanding of children's school adjustment in residential care.
The results of the current study are based on data of children and youth in rehabilitative and treatment facilities for children at risk in Israel. These findings might be useful also for other settings not included in the current study, such aspost-hospitalization and educational settings in Israel. It is imperative to replicate this study in other contexts and countriesand to identify to what extent the findings of this study generalize to other contexts.
Implications for practice and policy Children in care are a high risk group for poor school functioning. Child welfare agencies, when functioning in loco parentis, should spare no efforts to help children overcome educational deficiencies and enhance school involvement and behavior.
This is especially relevant in the case of children who remain in care for long periods, for whom professionals have plentyof opportunity to invest in enhancing their adjustment (Vinnerljung et al., 2005). Pecora et al. (2006) suggest that targetededucational support, integrated social work and education case management, as well as continuous monitoring of educationoutcomes and school behavior should be important practice components in residential care facilities.
The research findings demonstrate the need for an ecological perspective in addressing children's problems in school functioning within the care system. It is important not to focus solely on the characteristics of children at risk for poor schoolfunctioning. Instead, it is essential to identify the intersection of the familial and institutional contexts in which schoolfunctioning problems are more prevalent. The study demonstrates the importance of allocating sufficient resources to RCSswhich serve more vulnerable children. Although the study design does not allow for causal inferences, it may suggest someclear priorities and directions for policy and practice. The study highlights the need to support parent-child contact as a wayof improving the children's functioning in school. It emphasizes the importance of the implementation of peer preventionand intervention programs and the need to provide enrichment through after-school activities and tutoring. Finally, resourcesshould be allocated to adjust the setting's physical environment to children's needs so that it can serve as a stimulating andwarm milieu.
This research utilized the databases of the Israeli Welfare Ministry, which are based on the RAF method developed by the Brookdale Research Institute. I am grateful for the permission granted to use these databases. The cooperation of theWelfare Ministry supervisors is gratefully acknowledged. I would like to thank Prof. Rami Benbenishty for his kind supportand illuminating comments. Thanks are also due to Yasmin Alkalay and Elizabeth Miles for their assistance.
Ajdukovic, M., & Franz, B. S. (2005). Behavioral and emotional problems of children by type of out-of-home care in Croatia. International Journal of Social Welfare, 14, 163–175.
Alaimo, K., Olson, C. M., & Frongillo, E. A. (2001). Food insufficiency and American school-aged children's cognitive, academic, and psychosocial development.
Pediatrics, 108(1), 44–53.
Aldgate, J., Colton, M., Ghate, D., & Heath, A. (1992). Educational attainment and stability in long-term foste care. Children and Society, 6(2), 91–103.
Attar, S. (2006). Child, family, and institutional characteristics as predictors of the psychosocial and educational functioning of children at-risk in residential care settings. Unpublished dissertation: The Hebrew University of Jerusalem.
Attar-Schwartz, S. (2008). Emotional, behavioral and social problems among Israeli children in residential care: A multi-level analysis. Children and Youth Services Review, 30(2), 229–248.
Bailey, K. A. (2002). The role of physical environment for children in residential care. Residential Treatment for Children & Youth, 20(1), 15–27.
Barter, C., Renold, E., Berridge, D., & Cawson, P. (2004). Peer violence in children's residential care. Basingstoke: Palgrave.
Bates, L., Baird, D., Johnson, D. J., Lee, R. E., Luster, T., & Rehagen, C. (2005). Sudanese refugee youth in foster care: The "lost boys" in America. Child Welfare, Benbenishty, R., & Astor, R. A. (2005). School violence in context: Culture, neighborhood, family, school and gender. Oxford: Oxford University Press.
Berger, L. M. (2004). Income, family structure, and child maltreatment risk. Children and Youth Services Review, 26, 725–748.
Brodie, I. (2000). Children's homes and school exclusion: Redefining the problem. Support for Learning, 15(1), 25–29.
Bryk, A. S., & Raudenbush, S. W. (2002). Hierarchical linear models: Applications and data analysis methods. Newbury, CA: Sage.
Cantos, A. L., Gries, L. T., & Slis, V. (1997). Behavioral correlates of parental visiting during family foster care. Child Welfare, 76(2), 309–329.
Children Youth Service. (2005). Children and Youth Service: Children in residential care settings 2003–2004. Welfare Ministry and Brookdale Institute [Hebrew]: Davidson-Arad, B. (2005). Fifteen-month follow-up of children at risk: Comparison of the quality of life of children removed from home and children remaining at home. Children and Youth Services Review, 27, 1–20.
Dolev, T., & Barnea, N. (1996). Evaluation of the children's needs and intervention characteristics in nine institutions for children up to the age of 14. Jerusalem: Dolev, T., Benbenishty, R., & Timar, A. (2001). Decision committees in Israel: Their organization, work processes and outcomes. Jerusalem: JDC-Brookdale Institute Author's personal copy
S. Attar-Schwartz / Child Abuse & Neglect 33 (2009) 429–440 Emerson, J., & Lovitt, T. (2003). The educational plight of foster children in schools and what can be done about it. Remedial and Special Education, 24(4), Fanshel, D., & Shinn, E. (1978). Children in foster care: A longitudinal investigation. New York: Columbia University Press.
Flynn, R., & Biro, C. (1998). Comparing developmental outcomes for children in care with those for other children in Canada. Children and Society, 12, 228–233.
Fulcher, L. (2001). Differential assessment of residential group care for children and young people. British Journal of Social Work, 31(3), 417–435.
Gibbs, I., & Sinclair, I. (1999). Treatment and treatment outcomes in children's homes. Child and Family Social Work, 4, 1–8.
Gilligan, R. (2000). Adversity, resilience and young people: The protective value of positive school and spare time experiences. Children and Society, 14, Gilligan, R. (2007). Spare time activities for young people in care: What can they contribute to educational progress? Adoption & Fostering, 31(1), 92–99.
Gilman, R., & Hendwerk, M. L. (2001). Changes in life satisfaction as a function of stay in residential setting. Residential Treatment for Children & Youth, 18(4), Glisson, C., Hemmelgarn, A. L., & Post, J. A. (2002). The shortform assessment for children: An assessment and outcome measure for child welfare and juvenile justice. Research on Social Work Practice, 12(1), 82–106.
Haight, W. L., Kagle, J. D., & Black, J. E. (2003). Understanding and supporting parent-child relationships during foster care visits: Attachment theory and research. Social Work, 48(2), 195–207.
Heflinger, C. A., Simpkins, C. G., & Combs-Orme, T. (2000). Using the CBCL to determine the clinical status of children in state custody. Children and Youth Services Review, 22(1), 55–73.
Hegar, R. L., & Yungman, J. J. (1989). Toward a casual typology of child neglect. Children and Youth Services Review, 11, 203–220.
Hukkanen, R., Sourander, A., Bergroth, L., & Piha, J. (1999). Psychosocial factors and adequacy of services for children in children's homes. European Child & Adolescent Psychiatry, 8(4), 268–275.
Hussey, D. L., & Guo, S. (2002). Profile characteristics and behavioral change trajectories of young residential children. Journal of Child and Family Studies, Jackson, S., & Martin, P. Y. (1998). Surviving the care system: Education and resilience. Journal of Adolescence, 21, 569–583.
Jackson, S. (1994). Educating children in residential and foster care. Oxford Review of Education, 20(3), 267–279.
Jaffe, E. D. (1982). Child welfare in Israel. New York: Pregar Publisher.
Kessner, D., & Kalk, C. (1973). A strategy for evaluating health services. Washington DC: National Academy of Sciences, Institute of Medicine.
Khoury-Kassabri, M. (2006). Student victimization by educational staff in Israel. Child Abuse & Neglect, 30, 691–707.
Landsman, M. J., Groza, V., Tyler, M., & Malone, K. (2001). Outcomes of family-centered residential treatment. Child Welfare, 80(1), 351–379.
Lees, B., Morris, S., & Wheatley, H. (1994). Time to listen. London: Child-Line.
MacLeod, M. (1999). The abuse of children in institutional settings: Children's perspectives. In J. M. N. Stanely, & B. Penhale (Eds.), Institutional abuse: Perspectives across the life course (pp. 44–65). London: Routledge.
McWey, L. (2000). I promise to act better if you let me see my family: Attachment theory and foster care visitations. Journal of Family Social Work, 5(1), Mitic, W., & Rimer, M. (2002). The education attainment of children in care in British Columbia. Child and Youth Care Forum, 31(6), 397–414.
Mosek, A. (1993). Well-being and parental contact of foster children in Israel: A different situation from the USA? International Social Work, 36, 261–275.
Mosek, A., Taus, R., & Shomodi, J. (1997). The implication of an organizational change in residential care setting for children: A transition from group setting to family group homes. Welfare and Society, 18(2), 303–325 [Hebrew].
National Council for the Child. (2006). Children in Israel: Statistical abstract. Jerusalem: Israel National Council for the Child (Hebrew).
Newburn, T., Shiner, M., & Young, T. (2005). Dealing with disaffection: Young people mentoring and social inclusion. Portland, OR: Willan.
Oyserman, D., & Benbenishty, R. (1992). Keeping in touch: Ecological factors related to foster care visitation. Child and Adolescent Social Work Journal, 9(6), Pagani, L., Boulerice, B., Vitaro, F., & Tremblay, R. E. (1999). Effects of poverty on academic failure and delinquency in boys: A change and process model approach. The Journal of Child Psychology and Psychiatry and Allied Disciplines, 40, 1209–1219.
Pecora, P. J., Williams, J., Kesslers, R. C., Hiripis, E., O'brien, K., Emerson, J., Herrick, M. A., & Torres, D. (2006). Assessing the educational achievements of adults who were formerly placed in family foster care. Child and Family Social Work, 11, 220–231.
Romans, S. E., Martin, J. L., Anderson, J. C., O'shea, M. L., & Mullen, P. E. (1995). Factors that mediate between child sexual abuse and adult psychological outcome. Psychological Medicine, 25, 127–142.
Rose, M. (1992). Treasure Island: The metamorphosis of deprivation and injury in the residential setting. Residential Treatment for Children & Youth, 10, 75–87.
Rosenblatt, J. A., & Rosenblatt, A. (1999). Youth functional status and academic achievement in collaborative mental health and education programs: Two California care systems. Journal of Emotional and Behavioral Disorders, 7(1), 21–30.
Sawyer, R. J., & Dubowitz, H. (1994). School performance of children in kinship care. Child Abuse & Neglect, 18(7), 587–597.
Schiff, M., Nebe, S., & Gilman, R. (2006). Life satisfaction among Israeli youth in residential treatment care. British Journal of Social Work, 36, 1325–1343.
Schiff, M., & Benbenishty, R. (2006). Functioning of Israeli group-homes alumni: Exploring the differences and in-care correlates. Children and Youth Services Review, 28, 133–157.
Schober, R., & Annis, H. M. (1996). Barriers to help-seeking for change in drinking: A gender-focused review of the literature. Addictive Behaviors, 21(1), Shalom, H., & Apple, Z. (1994). The evaluation of life quality in institutions: The perspectives of children and youth. The Israeli Association for Education Research and Development, 23, 22–25 [Hebrew].
Shin, S. H. (2003). Building evidence to promote educational competence of youth in foster care. Child Welfare, 82(5), 615–632.
Slack, K. S., & Yoo, J. (2005). Food hardship and child behavior problems among low-income children. The Social Service Review, 79(3), 511–536.
Stein, M. (2006). Research review: Young people leaving care. Child and Family Social Work, 11, 273–279.
Tarantino, J. (2002). The child's perspective on what works in residential treatment [Abstract]. Dissertation Abstracts International: Section B: The Sciences and Engineering, 63(5B), 2607.
Tear, J. F., Smith, R. W., Osgood, D. W., Peterson, R. W., Authier, K., & Daly, D. L. (1995). Ecological influences in youth crisis shelters: Effects of social density and length of stay on youth problem behavior. Journal of Child and Family Studies, 4, 89–101.
Vinnerljung, B., Oman, M., & Gunnarson, T. (2005). Educational attainments of former child welfare clients: A Swedish national cohort study. International Journal of Social Welfare, 14(4), 265–276.
Vorria, P., Wolkind, S., Rutter, M., Pickles, A., & Hobsbaum, A. (1998a). A comparative study of Greek children in long-term residential group care in two-parent families. I. Social, emotional, and behavioral differences. Journal of Child Psychology and Psychiatry and Allied Disciplines, 39(2), 225–236.
Vorria, P., Wolkind, S., Rutter, M., Pickles, A., & Hobsbaum, A. (1998b). A comparative study of Greek group care in two-parent families. II. Possible mediating mechanisms. Journal of Child Psychology and Psychiatry and Allied Disciplines, 39(2), 237–245.
Wahba, D. A. (2003). Adjustment of children from disrupted and adoptive homes in residential treatment facility [Abstract]. Dissertation Abstracts Interna- tional: Section B: The Sciences and Engineering, 68(8B), 3651.
Walk, D., Zemach-Marom, T., Amihai-Ben Yaakov, V., Hauslich, Z., & Branz, B. (2004). Reliability examination of RAF child instrument at welfare childcare institutions. Jerusalem: Brookdale Institute.
Weiner, A., & Kupermintz, H. (2001). Facing adulthood alone: The long-term impact of family break-up and infant institutions. British Journal of Social Work, Whittaker, J. K. (1978). Changing character of residential child-care: Ecological perspective. Social Service Review, 52(1), 21–36.
Wolock, I., & Horowitz, B. (1984). Child Maltreatment as social problem: The neglect of the neglect. American Journal of Orthopsychiatry, 54(4), 530–543.
Author's personal copy
S. Attar-Schwartz / Child Abuse & Neglect 33 (2009) 429–440 Yule, W., & Raynest, N. V. (1972). Behavioral characteristics of children in residential care in relation to indices of separation. Journal of Child Psychology and Psychiatry, 13, 249–258.
Zemach-Marom, T., Fleishman, R., & Hauslich, Z. (2002). Improving quality of care in residential settings in Israel through the RAF method. In P. M. Bergh, E.
J. Knorth, F. Verheij, & D. C. Lane (Eds.), Changing care: Enhancing professional quality and client involvement in child and youth care services (pp. 30–41).
Amsterdam: SWP Publishers.
Zetlin, A. G., Weinberg, L. A., & Kimm, C. (2005). Helping social workers address the educational needs of foster children. Child Abuse & Neglect, 29(7),

Source: http://www.fkn.org.il/webfiles/fck/files/Attar-Schwartz_2009_CAAN.pdf