The early home environment and developmental outcomes for young children in the child welfare system

Children and Youth Services Review 33 (2011) 1392–1403

Contents lists available at ScienceDirect

Children and Youth Services Review j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / c h i l d yo u t h

The early home environment and developmental outcomes for young children in the child welfare system Brenda Jones Harden a,⁎, Jessica Vick Whittaker b a b

University of Maryland, MD, USA University of Virginia, VA, USA

a r t i c l e

i n f o

Available online 14 April 2011 Keywords: Home environment Preschool Infancy Developmental outcomes Child maltreatment Foster care

a b s t r a c t The quality of the early home environment is predictive of young children's subsequent cognitive, academic, and behavioral functioning. Limited research has focused on the effects of the early caregiving environment on the functioning of young children involved with the child welfare system. This study investigated the influence of children's home environments (i.e., number of children in the home, number of moves the child experienced, level of cognitive stimulation, and level of emotional support) during the first 2 years of life on their preschool developmental outcomes (i.e., cognition, language, social skills, and behavior problems). As anticipated, a high-quality early home environment promoted the well-being of preschool children who had entered the child welfare system as infants. Children who lived with greater numbers of children incurred more compromised cognitive, language, behavioral, and social outcomes. No significant associations emerged between the total number of placements and developmental outcomes; children who remained in the same home during infancy (typically the birth family home) had more compromised developmental outcomes in every domain except behavioral problems. Both cognitive stimulation and emotional support in the home predicted higher cognitive and language scores, decreased behavioral problems, and increased social skills. Early out-of-home placement and lack of emotional support interacted to predict children's behavioral problems. These findings are considered in the context of extant research and policy relevant to young children in the child welfare system. © 2011 Elsevier Ltd. All rights reserved.

1. Introduction The significance of home as a physical venue, a psychological experience, and a family context has been addressed across social science disciplines (Burton & Clark, 2005; Mallet, 2004). Child development experts have underscored the salience of home life for parents' engagement in the essential functions for the care of their children (e.g., feeding), the manifestation of parent–child emotional bonds, and the contextual influences that shape children's and families' daily activities (Bradley, 2006). Although the quality of the home environment contributes to the development of all children, it may be especially important for children of specific ages and from particular backgrounds. For example, the quality of the early home environment plays a critical role in determining the developmental trajectories of very young children, and is predictive of their later cognitive, academic, and behavioral functioning (Bradley, 2006; Duncan, Ziol-Guest, & Kalil, 2010; Evans, 2004, 2006). Additionally, studies suggest that a stable caregiving environment and the provision of stimulating and

⁎ Corresponding author. E-mail address: [email protected] (B.J. Harden). 0190-7409/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.childyouth.2011.04.009

sensitive experiences in the home may be particularly important for children living in impoverished and other high-risk environments (e.g., Crosnoe, Leventhal, Wirth, Pierce, & Pianta, 2010; Evans et al., 2010; Linver, Brooks-Gunn, & Kohen, 2002; Mistry, Biesanz, Chien, Howes, & Benner, 2008). Despite the growing body of evidence on the benefits of a high-quality home environment, there has been a lack of research specifically devoted to the effects of the early caregiving environment on the functioning of young children involved with the child welfare system.

2. Infancy and the child welfare system Infancy comprises the most rapid and complex development of the human organism (Bornstein & Lamb, 1992; Bremner & Fogel, 2001). Paradoxically, it is a time of extreme opportunity for children and is also a time of extreme vulnerability, particularly for those reared in high-risk environments (Knitzer & Perry, 2009; Shonkoff & Phillips, 2000). Within this developmental context, children entering the child welfare system as infants have caregiving and service experiences that are distinct (see Administration for Children and Families, 2006; Jones Harden, 2007). More extensive study of the epidemiologic and developmental trends for this population must be undertaken to

B.J. Harden, J.V. Whittaker / Children and Youth Services Review 33 (2011) 1392–1403

inform child welfare policy and practice concerning very young children and their families. From an epidemiologic perspective, infants in the child welfare system are distinguishable in various ways from their older counterparts (see Administration for Children and Families, 2006; Jones Harden, 2007). They are more likely to be maltreated, to experience a recurrence of maltreatment, and to die from maltreatment, particularly if they are under 1 year of age (Administration for Children and Families, 2009a, 2009b; National Clearinghouse on Child Abuse and Neglect Information, 2004). In the National Survey of Child and Adolescent Well-Being (NSCAW) sample, infants had a higher likelihood of all forms of maltreatment except sexual abuse (Administration for Children and Families, 2006). Neglect is particularly pernicious for infants, with consequences perhaps more deleterious for them than for older children (Connell-Carrick & Scannapieco, 2006; Erickson & Egeland, 2002; Pears & Fisher, 2005a, 2005b). The foster care placement rate for infants is significantly higher than that for the older children as well (Administration for Children and Families, 2010). In the NSCAW study, approximately a quarter of the foster care population comprised children younger than 2 years old (Administration for Children and Families, 2003, 2005). These higher placement rates have been attributed to caseworkers' concerns about infants' vulnerability, specifically infants' inability to protect themselves in situations of maltreatment (e.g., disclose their maltreatment and seek help from another adult; Barth, 1997; Wulczyn, 2004). Infants in the child welfare system are uniquely vulnerable to a host of environmental risks, including poverty, parental psychopathology, and impaired parenting (Brown, Cohen, Johnson, & Salzinnger, 1998; Scannapieco & Connell-Carrick, 2005). Such ecological risks have been linked to findings of developmental vulnerability in this population across a variety of outcomes, including physical health, cognitive and language skills, and socioemotional functioning (for a review, see Jones Harden, 2007). The NSCAW study, for example, found that more than half of children younger than age 2 were at risk for developmental delay (Administration for Children and Families, 2005, 2007). Smaller, cross-sectional studies have reported similarly high rates of developmental delay, specifically with regard to early cognition (e.g., Leslie, Gordon, Ganger, & Gist, 2002; Pears & Fisher, 2005a; Stahmer, Leslie, Landsverk, Zhang, & Rolls, 2007). However, few studies have traced the effects of early child welfare involvement on the cognitive abilities of these infants at older ages. A notable exception is the longitudinal research conducted on families with neglected children, in which early experiences of social deprivation were associated with later cognitive and academic difficulties (Erickson, Egeland, & Pianta, 1989). As with cognitive abilities, for infants and young children in the child welfare system, limited evidence exists regarding language functioning. In a unique study of maltreatment and language development, maltreated toddlers showed a shorter mean length of utterance, a more limited expressive vocabulary, and shorter bouts of contingent discourse during play sessions than lower socioeconomic status-matched controls (Coster, Gersten, Beeghly, & Cicchetti, 1989). Similarly, a study of preschool foster children documented higher levels of language delay than what is found in demographically similar populations (Stock & Fisher, 2006). Data from NSCAW show that infants in child welfare have more compromised language skills than expected for children their age. Although no differences in language competence were associated with particular types of maltreatment or types of placement, language delays were found in 14% of young maltreated children (Administration for Children and Families, 2005), compared to prevalence rates ranging from 2.3% to 19% reported overall for preschoolers (Nelson, Nygrem, & Panoscha, 2006). In the socioemotional domain, most studies of infants in the child welfare system have focused on their attachment to caregivers. A long line of research on maltreated infants has documented their insecure attachment patterns, more specifically “disorganized” attachment

1393

behaviors (Carlson, Cicchetti, Barnett, & Braunwald, 1989; Cicchetti & Toth, 1995a, 1995b; Cicchetti, Rogosch, & Toth, 2006; Egeland & Sroufe, 1981). Similarly, infants in foster care have shown insecure attachments to their substitute caregivers (Dozier, Stovall, Albus, & Bates, 2001; Oosterman & Schuengel, 2008). Related evidence on parent–infant interaction in child welfare populations has suggested impaired relationships between maltreating parents and their children (Larrieu & Zeanah, 2004). Foster parents not evincing an emotional commitment to children have also been found to have problematic interactions with the infants in their care (Dozier et al., 2001). The literature on other indicators of the socioemotional functioning of infants in the child welfare system is far more limited. Although research suggests higher rates of behavioral problems in older maltreated children (e.g., Administration for Children and Families, 2009a, 2009b; Dodge, Pettit, & Bates, 1997), this construct rarely extends to infants and young children in the child welfare system. The limited evidence suggests higher rates of behavioral problems for young children in the child welfare system (Black et al., 2002; Lawrence, Carlson, & Egeland, 2006; Rogosch, Cicchetti, & Aber, 1995; Stahmer et al., 2005). However, the literature tracing the effects of the early caregiving environment of children in the child welfare system on this and other socioemotional outcomes (e.g., social skills) remains scarce. 3. Early home environments and young children in the child welfare system A plethora of research has documented links between the home environment and the functioning of children (e.g., Bradley, 2006; Linver, Brooks-Gunn, & Cabrera, 2004), however, few studies have focused on the characteristics of caregiving environments of young children in the child welfare system. Scholars have cited the need for increased attention to this area of research as these children are more likely to experience a myriad of environmental risks including larger households, less stability, and less cognitively stimulating and emotionally sensitive environments (Barth, Weigensberg, Fisher, Fetrow, & Green, 2008), which are all linked to detrimental outcomes (Bradley, Corwin, Burchinal, McAdoo, & Garcia Coll, 2001; Evans, 2004, 2006). 3.1. Number of children in the home Sociological researchers have long debated the well-known finding that larger family size (i.e., larger numbers of children in the home) is related to poorer child outcomes, including lower intellectual development and higher rates of psychopathology (Farrington, 2000; Guo & VanWey, 1999; Kylmanen, Hakko, Rasanen, & Riala, 2010). Despite this line of research, few studies have examined the influence of family size on the outcomes of children in the child welfare system. This is an important research direction because there is some evidence that increased family size is a risk factor for maltreatment. For example, in the most recent National Incidence Study (NIS-4; Sedlak et al., 2010), maltreatment was more likely in families with four or more children. Additionally, in a populationbased study, Wu et al. (2004) found that infant maltreatment was more likely in families with greater than 2 children. Regarding foster care, data from NSCAW suggest that infants are more likely than older children to be placed in out-of-home settings with larger numbers of children (Barth et al., 2008). Studies of foster care children suggest that the number of children in the foster home may be linked to detrimental experiences for children (e.g., Chamberlain, Price, Reid, Landsverk, Fisher, & Stoolmiller, 2006). Similarly, having more children in the home during reunification has been found to be associated with maltreatment recurrence and reentry into the child welfare system. Fuller (2005) found that children ages 0–12 years who

1394

B.J. Harden, J.V. Whittaker / Children and Youth Services Review 33 (2011) 1392–1403

were returned to homes in which four or more children were present, when compared with children who were returned to homes with three or fewer children, were three times as likely to experience maltreatment recurrence. Further, Barth et al. (2008) used data from NSCAW to examine reentry for children ages 5 to 12 years, and found that children in households with a larger number of children had a risk of reentry that was 23% more than children from households with two or fewer children. 3.2. Placement stability The child welfare goal of permanency was developed to reduce “foster care drift”, the patterns in which children move through multiple placements (Wulczyn, Barth, Yuan, Jones Harden, & Landsverk, 2006). In a study using NSCAW data, it was documented that children in foster care move an average of a little more than 3 times, with older children more likely to have multiple placements (Rubin, O'Reilly, Hafner, Luan, & Localio, 2007). Another group of researchers have suggested that onethird to one-half of children placed into foster care experience disruptions in care within the first 2 years (Berrick, Needell, Barth, & Johnson-Reid, 1998; Staff & Fein, 1995). The data on placement stability suggest an overwhelmingly negative developmental impact of multiple placement transitions (Jones Harden, 2004; Ryan & Testa, 2005; Wulczyn, Kogan, & Jones Harden, 2003). Multiple out-of-home placements can lead to insecure attachments with caregivers (Stovall & Dozier, 1998; Strijker, Knorth, & Knot-Dickscheit, 2008), and poor developmental outcomes, such as externalizing behavior problems (Newton, Litrownik, & Landsverk, 2000; Simmel, 2007). Children in the NSCAW study with multiple placements showed more compromised developmental outcomes across domains than children who experienced greater placement stability (Administration for Children and Families, 2003). Further, in a study designed to evaluate the impact of pre-adoptions stressors on 293 adopted foster children's short and long-term psychosocial functioning, Simmel (2007) found that one of the strongest risk factors for the display of behavior problems was placement in multiple foster homes. In contrast, studies suggest that children who experience stable out-of-home placements earlier in life, and who experience sensitive caregiving in these placements, are more likely to form secure attachment relationships with their caregivers, and show more positive social–emotional outcomes (Oosterman & Schuengel, 2008; Ponciano, 2010; Stovall-McClough & Dozier, 2004; Stovall & Dozier, 2000). 3.3. Cognitive stimulation and emotional support Numerous studies have addressed the influence of home environment quality on outcomes for children in the general population (Bradley, 2006; Bradley, Corwin, McAdoo, & Garcia Coll, 2001). Recent research has documented that the home environments of children from families with notable risk factors are particularly compromised and particularly disadvantageous for child outcomes (Watamura, Phillips, Morrisey, McCartney, & Bub, 2011). For example, the quality of the home environments of young children tends to diminish if they are reared in poverty (Bradley, Corwin, Burchinal, et al., 2001). Moreover, the link between the home environment and behavioral outcomes is stronger for children from impoverished families than for other groups (Bradley, Corwin, Burchinal, et al., 2001). In contrast, studies have shown high quality caregiving can protect at-risk children from poor developmental outcomes (e.g., Linver et al., 2002; Mistry et al., 2008; Whittaker, Jones Harden, See, Meisch, & Westbrook, 2011). Positive caregiving behaviors including cognitive stimulation and emotional availability are associated with children's positive development (Kim-Cohen, Moffitt, Caspi, & Taylor, 2004; Denham, Mitchell-Copeland, Strandberg, Auerbach, & Blair, 1997).

Although this association has been found across all children (Bradley, Corwin, Burchinal, et al., 2001; Bradley, Corwin, McAdoo, et al., 2001), researchers have noted that caregiver's positive emotionality is a particularly significant protective factor for children experiencing high family risk (e.g., Ackerman et al., 1999). Children in the child welfare system who are already vulnerable from their experiences of maltreatment and other risk factors (e.g., placement instability) could be further compromised by the experience of a poor quality caregiving environment. Conversely, an environment high in cognitive stimulation and emotional support may serve as a protective factor for these children. A select number of studies have examined the relationship between environmental quality and child outcomes for children in the child welfare system (Cole, 2005; Jaffee, 2007; Stahmer et al., 2009). These studies suggest that organization of the home environment, cognitive stimulation, and emotional support, are related to positive child outcomes. For example, using a sample of infants and toddlers in the NSCAW data set, Jaffee (2007) examined the link between the caregiving environment and child outcomes, and found that improvements in the amount of sensitive, stimulating caregiving over the 18 month study period were related to better than expected language abilities and behavior. Furthermore, the relation between caregiving quality and changes in child outcomes was stronger for children who were removed from the care of their biological parents, compared to those who remained in their biological parents' care. In a similar study using a sample of NSCAW children 12–47 months old, Stahmer et al. (2009) found a positive relationship between the home environment and change in language and adaptive behavior standard scores over time. Although this small set of studies has begun to shed light on the links between environmental quality and children's outcomes for children in the child welfare system, gaps in the research remain. The limited evidence on foster family environments has documented the individual contributions of fewer children in the home, placement stability, or a high quality caregiving environment to child outcomes (e.g., Jaffee, 2007; Oosterman & Schuengel, 2008; Stahmer et al., 2009; Ponciano, 2010), but, in general, studies have not focused on the child welfare population in general. Furthermore, few studies have examined the interaction of environmental variables, and their impact on child outcomes. For example, is environmental quality more important for children who experience less placement stability? In the main, studies that have examined the association between environmental characteristics and child outcomes have focused on a limited number of child outcomes. For example, Jaffee (2007) examined the impact of changes in cognitive stimulation and emotional support on children's language abilities and problem behaviors, but did not examine children's cognitive outcomes or social skills, and did not examine child outcomes across the early childhood period. Moreover, linkages between more structural aspects of the caregiving environment that transcend placement type (e.g., number of children in household and placement stability) and outcomes for young children have rarely been made. Finally, although the developmental literature abounds regarding the impact of the infancy caregiving environment on later child outcomes (see Shonkoff & Phillips, 2000), and there is emerging consensus on the salience of infancy as an entry point into the child welfare system (e.g., Administration for Children and Families, 2006), limited empirical attention has been devoted to the contribution of infancy experiences for children in the child welfare system on their preschool outcomes. Lack of a nationally representative sample has limited the conclusions that can be drawn about infants in the child welfare system and their subsequent developmental outcomes. Because NSCAW is nationally representative, is longitudinal, and oversampled infants, its data set affords researchers a new opportunity to address specific gaps in the literature on infants and young children in the child welfare system.

B.J. Harden, J.V. Whittaker / Children and Youth Services Review 33 (2011) 1392–1403

Using NSCAW data, we examined the associations between the early caregiving environments experienced by children who entered the child welfare system as infants and their outcomes as preschoolers. Structural factors specific to the early caregiving environment (i.e., number of children in the home and number of placements) were investigated, as were other, more process characteristics of caregiving environments (i.e., level of cognitive stimulation and emotional support). The influence of these factors on preschool outcomes was investigated across four developmental domains: cognition; language; social skills; and behavioral problems. Four hypotheses were proposed: (1) higher numbers of children in the home would be associated with less positive child outcomes; (2) placement stability (e.g., fewer number of out-of-home placements) during infancy would positively affect preschool children's well-being; (3) a high quality early home environment (i.e., cognitive stimulation and emotional support) would positively affect preschool children's well-being; and (4) placement stability would moderate the effect of a quality caregiving environment on child outcomes such that the experience of a high quality caregiving environment would have a greater effect on child outcomes for children who experienced greater numbers of out-of-home placements. 4. Method 4.1. Data source Data from this study were from the National Survey of Child and Adolescent Well-Being (NSCAW). NSCAW is a longitudinal cohort study that includes over 6000 children and adolescents ages 0–15, who had contact with the child welfare system. Children in the study had either been investigated as victims of child abuse or neglect, or who had been in out-of-home care for approximately 1 year and whose placement had been preceded by an allegation of abuse or neglect. Children were selected from 92 primary sampling units (PSUs) in 97 counties across the United States. Several groups, including infants, were oversampled to ensure sufficient statistical power to study these groups. Data were collected in several waves from multiple sources. For the baseline data collection (Wave 1), there were in-person interviews or assessments with children, their primary caregivers (e.g., parents or other caregivers, such as foster parents and custodial kin caregivers), teachers (for school-aged children), and child welfare investigators. The Wave 2 data collection occurred 12 months post baseline, and included interviews with current caregivers and service caseworkers, and focused on the services received since baseline. Waves 3 and 4 data collection occurred 18 and 36 months after baseline, respectively, and included in-person interviews and assessments with children, their primary caregivers, their teachers (for school-aged children), and caseworkers. The response rate for Waves 2 and 3 was about 87%, and was 85% at Wave 4. 4.2. Participants Our analyses were on NSCAW data from baseline, Wave 2 (12month follow-up), Wave 3 (18-month follow-up), and Wave 4 (36month follow-up) child assessments and interviews with caregivers. Children who were 24 months old or younger at baseline were selected for these analyses, which yielded a sample of 1720 infants. At the Wave 4 data collection, children's mean age was 44.46 months (SD = 6.34). Children in this sample were primarily white (52.8%) and black or African American (29.7%). Most of the children lived at home with their primary caregiver (78.2%); however, 12.2% lived with relatives, and 9.2% lived with non-kin foster families. In terms of gender, the sample was almost equally divided (49.2% female) (for further descriptive information see Table 1).

1395

4.3. Measures 4.3.1. Caregiver and child background characteristics Caregivers were asked to respond to a series of questionnaires about family composition (e.g., number of children in home) and caregiver demographic characteristics (e.g., age, race/ethnicity, marital status, education, and employment). Developed for the NSCAW study, a questionnaire was also administered to caregivers to elicit the characteristics of the children, including race/ethnicity, gender, height, weight, and, for children younger than age 4, head circumference. The NSCAW Case Initiation Database was designed to verify children's demographic information and whether they were in out-of-home care. Additionally, caregivers were asked to report on their children's health—specifically, to rate their child's overall health on a scale from excellent to poor. 4.3.2. Child placement Throughout the study, child welfare system caseworkers provided information about the children's living situations. From these data several variables were derived, including the child's total number of placements at specific interview points. Placement data used in the current study are total number of placements at Wave 2 based on caseworker interview. 4.3.3. Quality of the home environment The quality and extent of stimulation and support in the home environment were measured with the Home Observation for Measurement of the Environment—Short Form (HOME-SF; Caldwell & Bradley, 1984). The HOME-SF was designed for the National Longitudinal Survey of Youth (NLSY) in consultation with the author of the original measure, Robert Bradley. The HOME-SF comprises both interview and observation items and yields two subscores, one measuring the cognitive stimulation of the child's environment and the other measuring the emotional relationship between the mother and the child. Good internal consistency has been reported for the HOME total score (r = 0.80), though coefficients for the subscales range from 0.30 to 0.80. Previous studies using the NLSY have demonstrated the reliability and construct validity of the HOME-SF. 4.3.4. Cognitive development Depending on the child's age, one of two measures was used to assess cognitive development. The cognitive functioning of children younger than age 4 was measured with the Battelle Developmental Inventory and Screening Test (BDI; Newborg, Stock, & Wnek, 1988). Although this measure assesses children's functioning in five domains, for this study only the cognitive domain was administered. The test–retest reliability ranges from 0.90 to 0.99. Concurrent validity with the Vineland Adaptive Behavior Scales (Sparrow, Balla, & Cicchetti, 1984) has been established (r =0.94; Newborg, Stock, Wnek, Guildubaldi, & Svinicki, 1984). The Kaufman Brief Intelligence Test (K-BIT; Kaufman & Kaufman, 1990) was used to measure the cognitive functioning of children aged 4 years or older. It entails a Vocabulary Subtest and a Matrices Subtest, as well as an IQ Composite; the composite measure was used for these analyses. The K-BIT yields standard scores for each child and was Table 1 Descriptive statistics for relevant covariates.

Maternal education (GRADE1T) Child age in months (CHAGE_B) Child health (CHHELTH)% Excellent Very good Good Fair Poor

n

Mean

SD

1088 1720

11.17 9.72

5.51 5.15

1: 2: 3: 4: 5:

46.74% 26.25% 17.46% 7.74% 1.80%

1396

B.J. Harden, J.V. Whittaker / Children and Youth Services Review 33 (2011) 1392–1403

Table 2 Descriptive statistics for predictors. n Total number of out of home placements through Wave 2 Categorical number of placements through Wave 2

Placement status Wave 1 Wave 3 Wave 4 Type of placement Wave 1

Wave 3

Wave 4

Number of children in the home Wave 1

Wave 3

Wave 4

Cognitive stimulation Wave 1 Wave 3 Wave 4 Emotional support Wave 1 Wave 3 Wave 4

standardized on a sample that matched the U.S. population in terms of gender, race, and ethnicity. Vocabulary internal consistency ranges from 0.89 to 0.98; Matrices internal consistency, from 0.74 to 0.95; and IQ Composite internal consistency, from 0.88 to 0.98. Reported test–retest reliability is high; content validity and concurrent validity with other intelligence tests have been established (Kaufman & Kaufman, 1990). 4.3.5. Language development Analyses of communication skills were based on children's Total Language score from the Preschool Language Scale-3 (PLS-3; Zimmerman, Steiner, & Pond, 1992). The scale measures prelinguistic and language skills in children from birth to age 6 and is composed of three subscales: Expressive Communication, Auditory Comprehension, and Total Language. Internal consistency for the total scale is, on average, high (M = 0.87); reported reliability ranges from 0.91 to 0.94. Concurrent validity has been established with several other measures of children's language development. 4.3.6. Behavioral problems The Child Behavior Checklist (CBCL; Achenbach, 1991, 1992) was used to assess children's behavioral problems. One of two versions was used, CBCL 2–3 or CBCL 4–18, depending on the child's age. This

1419 0 placements: 57.15% 1 placement: 21.49% 2+ placements: 21.35%

Mean

SD

0.77

1.10

1720 1493 1478

6.31 7.45 10.60

1.59 1.75 2.49

1720 1493 1478

7.05 7.24 8.79

1.80 1.84 2.03

Placed: 33.26% Not Placed: 66.74% Placed: 26.51% Not Placed: 73.49% Placed: 14.07% Not Placed: 73.49% In home: 64.45% Foster home: 19.62% Kin care setting: 12.93% In home: 74.16% Foster home: 14.64% Kin care setting: 11.20% In home: 87.13% Foster home: 6.60% Kin care setting: 6.27% 1 child: 34.30% 2 children: 26.86% 3 children: 17.62% 4 children: 9.77% ≥5 children 11.45% 1 child: 40.52%% 2 children: 22.50% 3 children: 18.26% 4 children: 9.24% ≥5 children 9.48% 1 child: 37.85% 2 children: 24.65% 3 children: 16.80% 4 children: 10.70% ≥5 children 10.00%

measure includes two broad problem scales (Externalizing Behaviors and Internalizing Behaviors), as well as a Total Problems scale. A raw and standardized score are derived for each problem scale, and a total behavioral problem score is computed. Our analyses are on the Total Problems score, the test–retest reliability of which is high (r = 0.91 for CBCL 2–3), with the authors reporting good internal consistency and stability. The CBCL 2–3 and the CBCL 4–18 are significantly related to scores on scales from other checklists (Achenbach, 1991). 4.3.7. Social skills The Parent Form of the Social Skills Rating System (Gresham & Elliott, 1990) was used to measure children's social skills. The Social Skills Scale comprises four subscales (Cooperation, Assertion, Responsibility, and Self-Control) and yields both a total score and scores for each subscale. The total score was used for these analyses. Internal consistency for the total scale ranges from 0.73 to 0.95; test–retest reliability, from 0.65 to 0.87. The total scale is moderately correlated with the Child Behavior Checklist—Parent Report Form (r = 0.58). 4.4. Analyses All statistical analyses were conducted with SUDAAN software (RTI International, 2007). Analyses were based on a subset of n = 1720

B.J. Harden, J.V. Whittaker / Children and Youth Services Review 33 (2011) 1392–1403 Table 3 Descriptive statistics for key outcome variables.

Table 5 Impact of number of children in the household on key outcomes.

Cognitive development (BD4_TDQ, b4 years) Cognitive development (KB4_NORM, 4+ years) Language development (CO4_TSTD) Behavior problems (TC4_TPT, b4 years) Behavior problems (BC4_TPT, 4+ years) Social skills (PS4_SCR)

0.319 0.380 0.398

0.933 (0.449) 3.872 (0.006) 5.368 (0.000)

Language development (CO4_TSTD) Time 1 1063 Time 1, Time 3 1063 Time 1, Time 3, Time 4 1013

0.091 0.093 0.118

1.463 (0.221) 0.230 (0.921) 2.538 (0.007)

Behavior problems (TC4_TPT, b 4 years) Time 1 925 Time 1, Time 3 925 Time 1, Time 3, Time 4 881

0.030 0.059 0.068

0.333 (0.855) 3.423 (0.012) 2.209 (0.031)

Behavior problems (BC4_TPT, 4+ years) Time 1 163 Time 1, Time 3 163 Time 1, Time 3, Time 4 155

0.410 0.470 0.499

2.470 (0.051) 3.213 (0.017) 2.874 (0.003)

Social skills (PS4_SCR) Time 1 Time 1, Time 3 Time 1, Time 3, Time 4

0.083 0.104 0.136

1.532 (0.201) 1.883 (0.121) 2.962 (0.002)

87.61 92.14 86.49 52.59 54.78 87.02

17.39 15.94 19.19 10.66 11.37 15.64

b2 + Placements

Wald F

5.823

2.575 (0.082)

Cognitive development (KB4_NORM, 4+ years) 131 0.309 − 1.226 − 0.001

0.065 (0.938)

Language development (CO4_TSTD) 881 0.085 2.985

5.683

2.583 (0.082)

Behavior problems (TC4_TPT, b 4 years) 763 0.024 1.993

0.570

0.629 (0.536)

Behavior problems (BC4_TPT, 4+ years) 140 0.455 − 1.055

5.122

0.971 (0.383)

Social skills (PS4_SCR) 760. 0.071

2.332

0.527 (0.592)

1.162

Cognitive development (KB4_NORM, Time 1 Time 1, Time 3 Time 1, Time 3, Time 4

1236 205 1447 1259 219 1247

Table 4 Impact of total number of placements on key outcomes. Placement

3.216 (0.017) 0.168 (0.954) 5.434 (0.000)

Predictors

The numbers of out-of-home placements were collapsed into a three-category variable (0, 1, and 2+). Results are summarized in

b1

0.150 0.151 0.192

SD

5.1. Total number of placements

Cognitive development (BD4_TDQ, b 4 years) 745 0.134 1.836

Wald F

Cognitive development (BD4_TDQ, b 4 years) Time 1 906 Time 1, Time 3 906 Time 1, Time 3, Time 4 861

Mean

5. Results

(R2)

(R2)

n

children aged 24 months or younger at baseline, and all employed the appropriate sample weights rescaled to sum to 1720 (i.e., NSCAW analysis weights). If data were missing for variables in a particular analysis, cases with missing data were excluded for that analysis. After computation of initial descriptive statistics for the covariates, predictors, and key outcomes (see Tables 1–3), focal analyses employed hierarchical general linear modeling techniques in which control variables were entered in earlier blocks and the key predictors were entered in subsequent blocks. For each analysis, the marginal contribution of the predictor variable, after controlling for the covariates, is documented as the Wald F value. The total model R2 is presented for each analysis as well, followed by the unstandardized slope (b). Control variables (covariates) were the child's race/ethnicity (black or African American, white, Hispanic, and “other”), gender (male and female), age in months, and health, as well as the mother's level of education. Predictors of specific interest were as follows: number of children in the home at each wave; total number of placements; and measures of cognitive stimulation and emotional support in the home environment at all waves. Two interactions were also tested as predictors: (1) measures of cognitive stimulation and out-of-home placement; and (2) measures of emotional support and out-of-home placement. Key outcome variables were as follows: measures of cognitive development for children under and at/above 4 years of age; language development for all children; total behavioral problems for children younger than 4 and for children aged 4 years or older; and social skills for all children.

n

1397

Bolded coefficients are significant at the p levels delineated.

n

4+ years) 152 152 147

918 918 867

Table 4. After controlling for the covariates, no significant differences were found between the number of placements and any outcome. 5.2. Number of children in the household To assess the predictive value of number of children in the home at each wave, a categorical measure of the number of children in the household (1, 2, 3, 4, and 5+) was created. Results are summarized in Table 5. All analyses yielded a statistically significant association between number of children (at differing waves) and key outcomes. Starting with younger children, number of children in the household at Wave 1 and Wave 4 appeared to affect cognitive development, although the direction of the association changes. For Wave 1 (baseline), having more children in the household was positively associated with younger children's cognitive development, whereas at Wave 4 (36-month follow-up) the association was negative. For older children at Wave 3 and Wave 4, having more children appeared to affect cognitive development negatively. Only at Wave 4 did number of children in the household affect language development, the association being negative. Having more children in the household at Wave 3 and at Wave 4 was significantly associated with behavioral problems for both younger and older children. At Wave 4, having more children in the household negatively affected children's social skills. 5.3. Cognitive stimulation Table 6 summarizes the predictive value of cognitive stimulation at three waves. No impact on cognitive development was detected for younger children; for the older children, however, Wave 1 stimulation levels appeared to positively affect cognitive development. Wave 4 cognitive stimulation positively affected language development. For younger children at Waves 3 and 4, higher levels of Wave 1 cognitive stimulation were associated with lower behavior problems and higher

B.J. Harden, J.V. Whittaker / Children and Youth Services Review 33 (2011) 1392–1403

1398

Table 6 Impact of cognitive stimulation on key outcomes. Predictors

n

(R2)

bWave

1

Cognitive development (BD4_TDQ, b 4 years) Wave 1 906 0.133 0.359 Wave 1, Wave 3 845 0.127 0.310 Wave 1, Wave 3, 845 0.135 0.007 Wave 4 Cognitive development (KB4_NORM, 4+ years) Wave 1 152 0.322 2.747 Wave 1, Wave 3 145 0.351 2.653 Wave 1, Wave 3, 145 0.351 2.649 Wave 4 Language development (CO4_TSTD) Wave 1 1063 0.074 Wave 1, Wave 3 995 0.062 Wave 1, Wave 3, 995 0.106 Wave 4

0.894 0.918 0.279

Behavior problems (TC4_TPT, b 4 years) Wave 1 925 0.038 −0.831 Wave 1, Wave 3 864 0.052 − 1.210 Wave 1, Wave 3, 864 0.078 −0.880 Wave 4 Behavior problems (BC4_TPT, 4+ years) Wave 1 163 0.368 − 1.371 Wave 1, Wave 3 154 0.382 − 1.040 Wave 1, Wave 3, 154 0.431 −1.136 Wave 4 Social skills (PS4_SCR) Wave 1 918 Wave 1, Wave 3 851 Wave 1, Wave 3, 851 Wave 4

0.089 0.085 0.138

1.919 1.562 1.172

Table 7 Impact of emotional support on key outcomes.

bWave

3

bWave

Predictors

n

(R2)

bWave

1

0.651

0.224 (0.637) 2.050 (0.135) 2.283 (0.085)

0.185

7.283 (0.008) 5.814 (0.004) 4.077 (0.009)

Cognitive development (KB4_NORM, 4+ years) Wave 1 152 0.281 1.000 Wave 1, Wave 3 145 0.379 0.379 Wave 1, Wave 3, 145 0.390 0.595 Wave 4

1.993

1.075 (0.303) 0.458 (0.643) 5.740 (0.001)

Language development (CO4_TSTD) Wave 1 1063 0.072 0.361 Wave 1, Wave 3 995 0.077 − 0.301 Wave 1, Wave 3, 995 0.100 −0.391 Wave 4

− 0.771

3.778 (0.055) 6.849 (0.002) 5.922 (0.001)

Behavior problems (TC4_TPT, b 4 years) Wave 1 925 0.023 0.055 Wave 1, Wave 3 864 0.038 − 0.306 Wave 1, Wave 3, 864 0.047 − 0.256 Wave 4

− 1.704

2.029 (0.158) 1.861 (0.162) 2.686 (0.052)

Behavior problems (BC4_TPT, 4+ years) Wave 1 163 0.367 1.153 Wave 1, Wave 3 154 0.394 1.409 Wave 1, Wave 3, 154 0.430 1.155 Wave 4

1.784

9.477 (0.003) 4.745 (0.011) 10.270 (0.000)

1.130 1.105

0.107 − 0.115

0.572 0.852

Wald F

Cognitive development (BD4_TDQ, b 4 years) Wave 1 906 0.132 − 0.134 Wave 1, Wave 3 845 0.132 − 0.157 Wave 1, Wave 3, 845 0.132 − 0.150 Wave 4

0.956 0.747

0.608 0.910

4

0.473 0.123

Social skills (PS4_SCR) Wave 1 918 Wave 1, Wave 3 851 Wave 1, Wave 3, 851 Wave 4

0.061 0.094 0.177

0.347 − 0.217 −0.302

bWave

3

1.128 1.142

bWave

−0.069

1.244

0.522 (0.460) 7.409 (0.001) 6.683 (0.000)

1.776

0.311 (0.578) 1.846 (0.164) 3.493 (0.021)

− 0.553

0.036 (0.851) 1.602 (0.208) 1.848 (0.145)

− 1.722

2.034 (0.158) 2.779 (0.068) 4.233 (0.008)

2.463

0.189 (0.665) 7.676 (0.001) 16.207 (0.000)

1.640 1.347

− 0.760 −0.363

Wald F 0.050 (0.823) 1.912 (0.154) 1.694 (0.175)

3.223 2.953

0.637 0.774

4

1.642 1.041

Bolded coefficients are significant at the p levels delineated.

Bolded coefficients are significant at the p levels delineated.

levels of Wave 4 cognitive stimulation were associated with lower levels of behavior problems at Wave 4. Wave 4 cognitive stimulation was also negatively related to behavior problems for older children. Finally, Wave 1 cognitive stimulation was positively associated with social skills at Waves 1 and 3, and cognitive stimulation at Wave 4 was positively associated with social skills at Wave 4.

5.6. Interaction between emotional support and number of out-of-home placements

5.4. Emotional support Table 7 summarizes the predictive value of emotional support at three waves. No significant associations emerged between emotional support and cognitive development for younger children. For the older children, emotional support at Wave 3 contributed to cognitive development; at Wave 4 it contributed significantly to language development. Emotional support did not significantly affect younger children's behavioral problems. For older children's behavioral problems, the overall model with all three waves entered was significant; however, no one wave was a significant predictor. Emotional support at Waves 3 and 4 appeared to be associated with increases in social skills. 5.5. Interaction between cognitive stimulation and number of out-of-home placements To assess whether number of out-of-home placements moderated the impact of cognitive stimulation on key outcomes, interaction variables were created between the three cognitive stimulation variables and the number-of-placements variable through Wave 2 (Table 8). No significant effect of the interaction variable was found on any key outcome.

To assess whether number of out-of-home placements moderated the impact of emotional support on key outcomes, interaction variables were created between the three emotional support variables and the number-of-placements variable through Wave 2 (Table 9). Out-ofhome placements moderated the effect of emotional support on younger children's behavioral problems. Analysis of the simple slopes indicated that the effect of emotional support on children's behavioral problems was stronger when children experienced more out-of-home placements (i.e., 2 or more) than when they experienced fewer. These findings suggest that emotional support affords some protection against the impact of high numbers of out-of-home placements. 6. Discussion Our analyses of the effects of the early caregiving environment on preschool children's developmental outcomes have produced a complex set of findings, underscoring the complexity of the experiences of children in the child welfare system. Overall, consistent with Bradley's (2006) formulation, the data underscore that the homes of young children are critical contextual influences that affect their developmental outcomes. As anticipated, a high-quality early home environment, in general, promoted the well-being of preschool children who had entered the child welfare system as infants. Children who lived with greater numbers of children incurred more compromised outcomes in the cognitive, language, behavioral, and social domains. Contrary to our hypothesis, no significant associations emerged between the total number of placements during infancy and developmental outcomes at preschool age. In fact,

B.J. Harden, J.V. Whittaker / Children and Youth Services Review 33 (2011) 1392–1403

1399

Table 8 Impact of interaction between cognitive stimulation and number of out-of-home placements. n

(R2)

b1

OOH

b2+

bcog.

OOH

b1

OOH × cog.

b2+ OOH

× cog.

Wald F

Cognitive development (BD4_TDQ, b 4 years) 745 0.139 − 6.424

3.948

0.170

1.392

0.335

0.242 (0.786)

Cognitive development (KB4_NORM, 4+ years) 131 0.356 17.060

29.868

3.192

− 2.513

−4.672

1.868 (0.161)

5.232

0.602

2.436

0.026

1.117 (0.332)

Language development (CO4_TSTD) 881 0.096

− 12.585

Behavior problems (TC4_TPT, b 4 years) 763 0.035

3.230

− 0.158

−0.726

− 0.252

0.074

0.045 (0.956)

Behavior problems (BC4_TPT, 4+ years) 140 0.469

3.022

− 22.385

−0.433

− 0.521

4.198

1.798 (0.172)

Social skills (PS4_SCR) 760 0.094

6.809

−1.391

1.652

− 0.842

0.697

0.629 (0.536)

Bolded coefficients are significant at the p levels delineated.

children who remained in the same home during infancy had more compromised developmental outcomes in every domain except behavioral problems. Both cognitive stimulation and emotional support in the home predicted higher cognitive and language scores, decreased behavioral problems, and increased social skills at preschool age. Although there was no significant moderating effect between placement and cognitive stimulation on children's outcomes, early out-of-home placement and lack of emotional support interacted to predict children's behavioral problems. Interestingly, variables reflecting the direct daily care young children experience (e.g., emotional support) more consistently and robustly affected child outcomes than did structural variables associated with child welfare services (e.g., number of placements).

2005). We therefore hypothesized that fewer placements would be associated with better developmental outcomes for preschool children. That our findings did not confirm this hypothesis calls for closer consideration of the children comprising this sample. First, the children in this sample were placed in various caregiving settings, including with birth parents, relatives, and non-kin foster parents. However, the overwhelming majority of the children lived with their birth parents. Thus, the lack of a relation between number of placements and child outcomes could be a reflection of the more vulnerable environments experienced by children who only had one placement (i.e., living with birth parents). There is a wealth of evidence pointing to the compromised caregiving environments provided by maltreating parents, and the associated adverse child outcomes (e.g., Azar, 2002; Scannapieco & Connell-Carrick, 2005). Additionally, much of the research on placement stability in child welfare has involved older children (Newton et al., 2000; Wulczyn et al., 2003). The negative effects of multiple transitions may not manifest themselves until children encounter the developmental demands of later childhood and express their emotional vulnerabilities in more overt ways that are perceived as more negative by caregivers (i.e., the reporters for children's social–emotional functioning). This may explain similar findings of the lack of effect of the number of placements on child outcomes in other research assessing preschool foster children. One research team found no effect of

6.1. Placement stability and child outcomes Permanency is a central tenet of current child welfare policy and programmatic approaches (Wulczyn et al., 2006). The emphasis on permanency emanated from a long line of research suggesting that multiple transitions were hazardous to children's well-being (Jones Harden, 2004; Newton et al., 2000; Ryan & Testa, 2005). Analyses conducted on the entire NSCAW sample, too, revealed a strong relationship between numbers of moves and more compromised developmental outcomes (Administration for Children and Families,

Table 9 Impact of interaction between emotional support and number of out of home placements. n

(R2)

b1

OOH

b2+

OOH

bemot.

b1

OOH × emot.

b2+

OOH × emot.

Wald F

Cognitive development (BD4_TDQ, b 4 years) 745 0.163 − 21.393

−11.015

− 0.783

3.538

2.480

3.334 (0.041)

Cognitive development (KB4_NORM, 4+ years) 131 0.328 17.844

36.890

1.600

− 2.660

− 4.917

1.724 (0.185)

− 17.230

−13.106

− 0.199

2.950

2.624

1.990 (0.143)

Behavior problems (TC4_TPT, b 4 years) 763 0.056 18.042

9.185

0.486

− 2.441

− 1.282

7.072 (0.002)

Behavior problems (BC4_TPT, 4+ years) 140 0.437 19.014

− 21.604

0.526

− 2.676

3.352

2.551 (0.084)

3.137

0.942

− 0.075

− 0.201

0.009 (0.991)

Language development (CO4_TSTD) 881 0.099

Social Skills (PS4_SCR) 760 0.870

1.545

Bolded coefficients are significant at the p levels delineated.

1400

B.J. Harden, J.V. Whittaker / Children and Youth Services Review 33 (2011) 1392–1403

placement history on developmental, cognitive, and neuropsychological functioning (Pears & Fisher, 2005a) as well as emotional understanding and theory of mind (Pears & Fisher, 2005b). It is important to note that the children in this sample moved a limited number of times (i.e., M = 1.87; SD = 1.37). There is some evidence which suggests that the threshold for the negative impact of placement change is high (children experiencing three to five placements; Newton et al., 2000; Ryan & Testa, 2005). Although the duration of the sampled children's placements is unknown, other research suggests that infants have a high initial movement rate (Wulczyn et al., 2003), with longer placements and permanency typically coming shortly later. Leaving the initial short-term placements, which are more characteristic of the infant experience, may be less traumatic than leaving long-term foster care placements. 6.2. Number of children and child outcomes In the general family literature, it appears that children who are reared in households with large numbers of children have more adverse outcomes, particularly if they are not the oldest child (e.g., Farrington, 2000). Although complex, the evidence from the current study suggests that larger numbers of children is associated with poorer cognitive, language, social, and behavioral outcomes for young children in the child welfare system. Thus, data from this study corroborate evidence about the negative impact of being in a family with large numbers of children on children's developmental outcomes (Barth et al., 2008; Chamberlain et al., 2006). The global impact of this experience for young children in the child welfare system suggests that this set of high-risk children may be vulnerable to the negative effects of the experience of multiple children in the household across developmental domains, whether they live with their birth families or are in foster care settings. 6.3. Home environment quality and child outcomes Our clearest and most robust findings related to the care that young children directly experience. We considered whether levels of cognitive stimulation and emotional support in the early caregiving environment may contribute to developmental outcomes. A plethora of evidence on non-child welfare samples documents the contributions of these aspects of the caregiving environment for cognitive, language, and socioemotional outcomes, especially for high-risk children (e.g., children reared in impoverished circumstances; Bradley, Corwin, McAdoo, et al., 2001). Consistent with the literature, both cognitive stimulation and emotional support predicted better developmental outcomes in each domain at some point along the early childhood developmental trajectory (Linver et al., 2004). This finding is consistent with those of studies on non-child welfare samples that suggest that both aspects of the home environment are important for high-risk children's development across domains (e.g., Chazan-Cohen et al., 2009). The quality of the home environment – whether through the provision of development-enriching materials or through nurturing interactions – promoted the generalized development of these children. Finally, the results of the interaction analyses indicated that the effect of cognitive stimulation on children's outcomes was not affected by the number of out of home placements. The strong relation between cognitively and linguistically enriched environments and children's outcomes, particularly in the cognitive and language domains, has been documented for children who have experienced a myriad of risk factors (e.g., Mistry et al., 2008). However, the potency of this environmental characteristic may be reduced if children have experienced multiple placements. On the other hand, the current data indicated that the number of placements moderated the impact of emotional support in the caregiving environment on children's behavioral outcomes. The affective quality

of care received during infancy may, therefore, have the potential to compensate for negative child welfare experiences (e.g., placement disruption). This evidence lends support to research on other young child populations that emphasizes the import of the early social– emotional environment on young children's later outcomes (ChazanCohen et al., 2009; Mueller et al., 2010). 6.4. Limitations and research directions Although the comprehensiveness of the NSCAW data set is unprecedented, some limitations clearly qualify our findings. The goal of this study was to examine the caregiving contexts and associated developmental outcomes of children who entered the child welfare system as infants. Although evidence exists that the experiences of infants younger than 4 months of age are distinct from those of older infants, we could not answer this question because the infants in the NSCAW sample were older at baseline (M = 13 months). In addition, because the status of children in the child welfare system may change daily (e.g., child may move from birth family to transitional placement to relatives), we could not track children's whereabouts precisely; therefore, we made some assumptions and decisions about placement status that may incompletely reflect children's actual experiences. Future studies should examine distinct samples of infants (e.g., neonates vs. older infants) and conduct more refined analyses of children's movement trajectories through the child welfare system. Measurement challenges also merit special mention. Survey research typically precludes the use of relatively rich measures of infant and family functioning. Because all young children, regardless of where they resided, were included in the study, we could not capitalize on data on parental functioning. Questions about parental mental health and other parental characteristics were posed only to birth parents. In addition, this study would have benefited from data derived from assessments of parent–infant interaction and attachment, observational methods of infant socioemotional functioning (e.g., emotion regulation), and physiological techniques. The longitudinal nature of the study, moreover, required that we change the measure used to assess cognition, social skills, and behavioral problems. Due to a developmental shift that occurs around 4 years of age, two different measures were used in each of these three domains. Because the skills and behaviors informing each domain differed somewhat according to the age-specific version, we treated the two versions of each measure separately in the analyses. In future studies, assessments of parent functioning should be included in order to consider the influence of parent mental health and parenting processes on child outcomes. Additionally, more consistent and refined assessments of child behavior and functioning would enhance our understanding of the impact of these environmental variables on young children's development. The influence of the parent–child relationship on children's outcomes should also be assessed, as well as its role as a potential mediator between the impact of negative environmental characteristics on this population of children. Finally, future research should consider the cumulative impact of these environmental risk factors on children's outcomes. 6.5. Policy and program implications Overall, the results of the current study argue for child welfare intervention that extends beyond safety and permanency, and focuses on the well-being of children (Wulczyn et al., 2006). The evidence presented herein corroborates the rich corpus of evidence about environmental contributions to child development, specifically that the developmental trajectories of young children in the child welfare system can be altered by specific ecological experiences. Thus, the child welfare system should strive to foster development-promoting environments for the young children they serve.

B.J. Harden, J.V. Whittaker / Children and Youth Services Review 33 (2011) 1392–1403

Based on the current study, there are specific mechanisms that would allow for the enhancements of the caregiving environments of young children in the child welfare system. First, findings from this study indicate that the number of children in a home has a potent effect on children's outcome. Therefore, birth and foster families with large numbers of children should receive more intensive interventions, which support parents to meet the varying needs of their children. For example, these families may need a higher dosage of home visits, expanded case management services that addresses the needs of all the children in these households, or additional child care or respite care supports. Regarding foster parent recruitment and assessment, there may need to be more careful examination of their capacity to meet the developmental needs of large numbers of children prior to the placement of more children in their homes. Although there was no relation between number of placements and outcomes, the data do suggest that the child welfare system should maintain family stability for young children. Given that the majority of children in this study were in one home, that of their birth parents, it is important to consider what services the child welfare system provides that address the compromised environments in which these children live. Child welfare scholars and practitioners have longed called for increased funding for prevention services (Wulczyn et al., 2006), which could be utilized to intervene with maltreating families in regard to their abusive and neglectful behavior, and also to address the quality of the environments they provide to their young children. The integration of child welfare and early intervention services, as in the Early Head Start child welfare initiative (Administration for Children and Families, 2011), is a particularly promising approach to facilitating improved caregiving environments for this population of children. Child welfare services should promote enhanced parenting among birth and foster families alike. Parenting and parent–child interaction interventions, which scaffold caregivers to provide cognitively and linguistically enriched environments to their young children, are essential for the families of these children (Administration for Children and Families, 2011; Stahmer et al., 2005). Similarly, parenting interventions that facilitate parents' emotional availability and responsivity to these children can positively impact the development of core emotional processes, such as emotion expression and regulation and attachment, which have implications for their short- and long-term mental health (Cicchetti et al., 2006; Dozier et al., 2009; Izard, Fine, Mostow, Trentacosta, & Campbell, 2002). Importantly, these interventions should build on the burgeoning evidence base about effective parenting interventions for maltreated children and their families (Barth, 2009). 6.6. Concluding remarks In sum, the current study documents that the home environments of infants in the child welfare system can explain some of the variability in the early childhood outcomes of this population of children. Specifically, worse outcomes were observed for children who had experienced environments that had higher numbers of children, were less cognitive stimulating, and were less emotionally supportive. There was no relation between number of placements and child outcomes, though this may be attributable to the low number of placements that these infants experienced. Overall, this study suggests that a high-quality caregiving environment is an essential development-promoting context for children who are more vulnerable due to their age (i.e., infancy) and the negative parenting experiences that may have precipitated their child welfare involvement. Thus, child welfare practice should be targeted to improving the environments young children experience, including the provision of extensive supports to families with large numbers of children and services that address parents' capacity to create cognitively enriched and emotionally responsive environments.

1401

References Achenbach, T. M. (1991). Manual for the Child Behavior Checklist/4–18 and 1991 profile. Burlington, VT: University of Vermont, Department of Psychiatry. Achenbach, T. M. (1992). Manual for the Child Behavior Checklist/2–3 and 1992 profile. Burlington: University of Vermont, Department of Psychiatry. Ackerman, B., Izard, C., Schoff, K., Youngstrom, E., & Kogos, J. (1999). Contextual risk, caregiver emotionality, and the problem behaviors of 6- and 7- year old children from economically disadvantaged families. Child Development, 70(6), 1415–1427. Administration for Children and Families (2003). Report on the one-year-in-foster-care sample, National Survey of Child and Adolescent Well-Being. Washington, DC: Author. Administration for Children and Families (2005). Wave 1 report on the child welfare services sample, National Survey of Child and Adolescent Well-Being. Washington, DC: Author. Administration for Children and Families (2006). NSCAW Research Brief No. 4: Infants and toddlers in the child welfare system. Washington, DC: Author. Administration for Children and Families (2007). NSCAAW Research Brief No. 8: Early intervention services among infants and toddlers in child welfare. Washington, DC: Author. Administration for Children and Families (2009a). NSCAAW Research Brief No. 16: A summary of NSCAW findings. Washington, DC: Author. Administration for Children and Families (2009b). Child maltreatment. Washington, DC: Author. Administration for Children and Families (2010). The AFCARS report: Preliminary FY 2009 estimates. Washington, DC: Author. Administration for Children and Families (2011). ACF Information Memorandum No. ACYF-CB-IM-11-01. Azar, S. (2002). Parenting and child maltreatment. In M. Bornstein (Ed.), Handbook of parenting. Social and applied parenting, 4. (pp. 361–378) Mahwah, NJ: Erlbaum. Barth, R. (1997). Effects of age and race on the odds of adoption versus remaining in long-term out-of-home care. Child Welfare, 76(2), 285–308. Barth, R. (2009). Preventing child and abuse and neglect with parent training: Evidence and opportunities. The Future of Children, 19(2), 95–118. Barth, R., Weigensberg, E., Fisher, P., Fetrow, B., & Green, R. (2008). Reentry of elementary aged children following reunification from foster care. Children and Youth Services Review, 30(4), 353–364. Berrick, J., Needell, B., Barth, R., & Jonson-Reid, M. (1998). The tender years: Toward developmentally sensitive child welfare services for very young children. Oxford: New York. Black, M., Papas, M., Hussey, J., Hunter, W., Dubowitz, H., Kotch, J., et al. (2002). Behavior and development of preschool children born to adolescent mothers: Risk and 3generation households. Pediatrics, 109, 573–580. Bornstein, M., & Lamb, M. (1992). Development in infancy. New York: McGraw-Hill. Bradley, R. (2006). The home environment. In N. Watt, C. Ayoub, R. Bradley, J. Puma, & W. LeBoeuf (Eds.), The crisis in youth mental health: Critical issues and effective programs. Early intervention programs and policies, Vol. 4, . Westport, CT: Praeger/ Greenwood. Bradley, R. H., Corwin, R., Burchinal, M., McAdoo, H., & Garcia Coll, C. (2001a). The home environments of children in the United States: Part II. Relations with behavioral development through age 13. Child Development, 72(6), 1868–1886. Bradley, R. H., Corwin, R., McAdoo, H., & Garcia Coll, C. (2001b). The home environments of children in the United States: Part I. Variations by age, ethnicity, and poverty status. Child Development, 72(6), 1844–1867. Bremner, C., & Fogel, A. (2001). Handbook of infant development. Malden: Blackwell. Brown, J., Cohen, P., Johnson, J., & Salzinnger, S. (1998). A longitudinal analysis of risk factors for child maltreatment: Findings of a 17-year prospective study of officially recorded and self-reported child abuse and neglect. Child Abuse & Neglect, 22, 1065–1078. Burton, L., & Clark, S. (2005). Home place and housing in the lives of low-income, urban African American families. In V. McLoyd, N. Hill, & K. Dodge (Eds.), African American family life: Ecological and cultural diversity (pp. 166–188). New York: Guilford. Caldwell, B. M., & Bradley, R. H. (1984). Home observation for measurement of the environment. Little Rock: University of Arkansas at Little Rock. Carlson, V., Cicchetti, D., Barnett, D., & Braunwald, K. (1989). Disorganized/disoriented attachment relationships in maltreated infants. Developmental Psychology, 25, 525. Chamberlain, P., Price, J., Reid, J., Landsverk, J., Fisher, P., & Stoolmiller, M. (2006). Who disrupts from placement in foster and kinship care? Child Abuse and Neglect, 30(4), 409–424. Chazan-Cohen, R., Raikes, H., Brooks-Gunn, J., Ayoub, C., Pan, B., Kisker, E., et al. (2009). Low-income children's school readiness: Parent contributions over the first 5 years. Early Education and Development, 20(6), 958–977. Cicchetti, D., Rogosch, F., & Toth, S. (2006). Fostering secure attachments in infants in maltreating families through preventive interventions. Development and Psychopathology, 18, 623–649. Cicchetti, D., & Toth, S. (1995a). A developmental psychopathology perspective on child abuse and neglect. Journal of the American Academy of Child and Adolescent Psychiatry, 34, 541–565. Cicchetti, D., & Toth, S. (1995b). Child maltreatment and attachment organization: Implications for intervention. In S. Goldberg, R. Muir, & J. Kerr (Eds.), Attachment theory: Social, developmental and clinical perspectives (pp. 279–308). Hillsdale, NJ: Analytic Press. Cole, S. A. (2005). Foster caregiver motivation and infant attachment: How do reasons for fostering affect relationships? Child and Adolescent Social Work, 22(5–6), 441–457.

1402

B.J. Harden, J.V. Whittaker / Children and Youth Services Review 33 (2011) 1392–1403

Connell-Carrick, K., & Scannapieco, M. (2006). Ecological correlates of neglect in infants and toddlers. Journal of Interpersonal Violence, 21(3), 299–316. Coster, W., Gersten, M. S., Beeghly, M., & Cicchetti, D. (1989). Communicative functioning in maltreated toddlers. Developmental Psychology, 25, 1020–1029. Crosnoe, R., Leventhal, T., Wirth, R., Pierce, L., & Pianta, R. (2010). Family socioeconomic status and consistent environmental stimulation in early childhood. Child Development, 81(3), 972–987. Denham, S., Mitchell-Copeland, J., Strandberg, K., Auerbach, S., & Blair, K. (1997). Parental contributions to preschoolers' emotional competence: Direct and indirect effects. Motivation and Emotion, 21(1), 65–86. Dodge, K., Pettit, G., & Bates, J. (1997). How the experience of early physical abuse leads children to become chronically aggressive. In D. Cicchetti, & S. Toth (Eds.), Rochester symposium on developmental psychopathology. Developmental perspectives on trauma, 8. (pp. 263–288) Rochester, NY: University of Rochester Press. Dozier, M., Lindhiem, O., Lewis, E., Bick, J., Bernard, K., & Pelosa, E. (2009). Effects of a foster parent training program on young children's attachment behaviors: Preliminary evidence from a randomized clinical trial. Child and Adolescent Social Work Journal, 26, 321–332. Dozier, M., Stovall, K. C., Albus, K., & Bates, B. (2001). Attachment for infants in foster care: The role of caregiver state of mind. Child Development, 72, 1467–1477. Duncan, G., Ziol-Guest, K., & Kalil, A. (2010). Early childhood poverty and adult attainment, behavior, and health. Child Development, 81(1), 306–325. Egeland, B., & Sroufe, L. A. (1981). Attachment and early maltreatment. Child Development, 52, 44–52. Erickson, M., & Egeland, B. (2002). Child neglect. In J. Myers, L. Berliner, J. Briere, C. Hendrix, C. Jenny, & T. Reid (Eds.), The APSAC handbook on child maltreatment (pp. 3–20). Thousand Oaks, CA: Sage. Erickson, M., Egeland, B., & Pianta, R. (1989). The effects of maltreatment on the development of young children. In D. Cicchetti, & V. Carlson (Eds.), Child maltreatment (pp. 647–684). New York: Cambridge University Press. Evans, G. (2004). Environment of childhood poverty. The American Psychologist, 59(2), 77–92. Evans, G. (2006). Child development and the physical environment. Annual Review of Psychology, 57, 423–451. Evans, G., Ricciuti, H., Hope, S., Schoon, I., Bradley, R., Corwyn, R., et al. (2010). Crowding and cognitive development: The mediating role of maternal responsiveness among 36 month-old children. Environment and Behavior, 42(1), 135–148. Farrington, D. (2000). Psychosocial predictors of adult personality and adult convictions. Behavioral Sciences & the Law, 18, 605–622. Fuller, T. (2005). Child safety at reunification: A case-control study of maltreatment recurrence following return home from substitute care. Children and Youth Services Review, 27(12), 1293–1306. Gresham, F. M., & Elliott, S. N. (1990). Social Skills Rating System manual. Circle Pines, MN: American Guidance Service. Guo, G., & VanWey, L. (1999). Sibship size and intellectual development: Is the relationship causal? American Sociological Review, 64(2), 169–187. Izard, C., Fine, S., Mostow, A., Trentacosta, C., & Campbell, J. (2002). Emotion process in normal and abnormal development and preventive interventions. Development and Psychopathology, 14, 761–787. Jaffee, S. (2007). Sensitive, stimulating caregiving predicts cognitive and behavioral resilience in neurodevelopmentally at-risk infants. Development and Psychopathology, 19(3), 631–647. Jones Harden, B. (2004). Safety and stability for foster children: A developmental perspective. The Future of Children, 14(1), 39–47. Jones Harden, B. (2007). Infants in the child welfare system: A development framework for policy and practice. Washington, DC: Zero to Three. Kaufman, A., & Kaufman, N. (1990). Kaufman Brief Intelligence Test (K-BIT). Circle Pines, MN: American Guidance Service. Kim-Cohen, J., Moffit, T., Caspi, A., & Taylor, A. (2004). Genetic and environmental processes in young children's resilience and vulnerability to socioeconomic deprivation. Child Development, 75(3), 651–668. Knitzer, J., & Perry, D. (2009). Poverty and infant and toddler development: Facing the complex challenges. In C. Zeanah (Ed.), Handbook of infant mental health (pp. 135–152). New York: Guilford. Kylmanen, P., Hakko, H., Rasanen, P., & Riala, K. (2010). Is family size related to adolescence mental hospitalization? Psychiatry Research, 177(1–2), 188–191. Larrieu, J., & Zeanah, C. (2004). Treating parent–infant relationships in the contexts of maltreatment: An integrated systems approach. In A. Sameroff, S. McDonogh, & K. Rosenblum (Eds.), Treating parent–infant relationship problems: Strategies for intervention (pp. 243–266). New York: Guilford Press. Lawrence, C., Carlson, E., & Egeland, B. (2006). The impact of foster care on development. Development and Psychopathology, 18, 57–76. Leslie, L. K., Gordon, J. N., Ganger, W., & Gist, K. (2002). Developmental delay in young children in child welfare by initial placement type. Infant Mental Health Journal, 23 (5), 496–516. Linver, M., Brooks-Gunn, J., & Kohen, D. (2002). Family processes as pathways from income to young children's development. Developmental Psychology, 38(5), 719–734. Linver, M., Brooks-Gunn, J., & Cabrera, N. (2004). The HOME Inventory: Derivation of conceptually designed subscales. Parenting, 4(2–3). Mallet, S. (2004). Understanding home: A critical review of the literature. The Sociological Review, 52(1), 62–89. Mistry, R., Biesanz, J., Chien, N., Howes, C., & Benner, A. (2008). Socioeconomic status, parental investments, and the cognitive and behavioral outcomes of low-income children from immigrants and native households. Early Childhood Research Quarterly, 23(2), 193–212.

Mueller, S., Maheu, F., Dozier, M., Peloso, E., Mandell, P., Leibenluft, E., et al. (2010). Early life stress is associated with impairment in cognitive control in adolescence: An fMRI study. Neuropsychologia, 48(10), 3037–3044. National Clearinghouse on Child Abuse and Neglect Information (2004). Child abuse and neglect fatalities: Statistics and interventions. Retrieved January 1, 2008, from. http://nccanch.acf.hhs.gov Nelson, H., Nygrem, P., & Panoscha, R. (2006). Screening for speech and language delay in preschool children: Systematic evidence review for U.S. Preventive Services Task Force. Pediatrics, 117, 298–319. Newborg, J., Stock, J. R., & Wnek, L. (1988). Battelle Developmental Inventory (BDI). Allen, TX: Riverside. Newborg, J., Stock, J. R., Wnek, L., Guildubaldi, J., & Svinicki, J. (1984). Battelle Developmental Inventory: With recalibrated technical data and norms: Examiner's manual. Itasca, IL: Riverside. Newton, R. R., Litrownik, A. J., & Landsverk, J. A. (2000). Children and youth in foster care: Disentangling the relationship between problem behaviors and number of placements. Child Abuse & Neglect, 24, 1363–1374. Oosterman, M., & Schuengel, C. (2008). Attachment in foster children associated with caregivers' sensitivity and behavioral problems. Infant Mental Health Journal, 29(6), 609–623. doi:10.1002/imhj.20198. Pears, K., & Fisher, P. (2005a). Developmental, cognitive, and neuropsychological functioning in preschool foster children: Association with prior maltreatment and placement history. Journal of Developmental and Behavioral Pediatrics, 26(2), 112–122. Pears, K., & Fisher, P. (2005b). Emotional understanding and theory of mind among maltreated children in foster care: Evidence of deficits. Development and Psychopathology, 17, 47–65. Ponciano, L. (2010). Attachment in foster care: The role of maternal sensitivity, adoption, and foster mother experience. Child & Adolescent Social Work Journal, 27 (2), 97–114. Rogosch, F., Cicchetti, D., & Aber, J. L. (1995). The role of child maltreatment in early deviations in cognitive nad affective processing abilities and later peer relationship problems. Development and Psychopathology, 7, 591–609. RTI International (2007). SUDAAN user's manual, release 9.0.1. Research Triangle Park, NC: Author. Rubin, D., O'Reilly, A., Hafner, L., Luan, X., & Localio, R. (2007). Placement stability and early behavioral outcomes among children in out of home care. In R. Haskins, M. B. Webb, & F. Wulczyn (Eds.), Child Protection: Using research to inform policy and practice. Washington, DC: Brookings Institution. Ryan, J., & Testa, M. (2005). Child maltreatment and juvenile delinquency: Investigating the role of placement and placement stability. Children and Youth Services Review, 22(9–10), 227–249. Scannapieco, M., & Connell-Carrick, K. (2005). Focus on the first years: Correlates of substantiation of child maltreatment for families with children 0–4. Children and Youth Services Review, 27(12), 1307–1323. Sedlak, A., Mettenburg, J., Basena, M., Petta, I., McPherson, K., Greene, Al, et al. (2010). Fourth national incidence study of child abuse and neglect. Washington, DC: Administration for Children and Families. Shonkoff, J., & Phillips, D. (Eds.). (2000). From neurons to neighborhoods: The science of early childhood development. Washington, DC: National Academy Press. Simmel, C. (2007). Risk and protective factors contributing to the longitudinal psychosocial well-being of adopted foster children. Journal of emotional and behavioral disorders, 15(4), 237–249. Sparrow, S. S., Balla, D. A., & Cicchetti, D. V. (1984). Vineland Adaptive Behavior Scales (VABS). Circle Pines, MN: American Guidance Service. Staff, I., & Fein, E. (1995). Stability and change: Initial findings in a study of treatment foster care placements. Children and Youth Services Review, 17(3), 379–389. Stahmer, A., Hurlburt, M., Horwitz, S., Landsverk, J., Zhang, J., & Leslie, L. (2009). Associations between intensity of child welfare involvement and child development among young children in child welfare. Child Abuse and Neglect, 33(9), 598–611. Stahmer, A., Leslie, L., Hurlburt, M., Barth, R., Webb, M. B., Landsverk, J., et al. (2005). Developmental and behavioral needs and service use for young children in child welfare. Pediatrics, 116(4), 891–900. Stahmer, A., Leslie, L., Landsverk, J., Zhang, J., & Rolls, J. (2007). Developmental services for young children in foster care: Assessment and service delivery. Journal of Social Service Research, 33(2), 27–38. Stock, C., & Fisher, P. (2006). Language delays among foster children: Implications for policy and practice. Child Welfare, 85(3), 445–461. Stovall, K., & Dozier, M. (1998). Infants in foster care: An attachment theory perspective. Adoption Quarterly, 2(1), 55–88. Stovall, K., & Dozier, M. (2000). The development of attachment in new relationships: Single subject analyses for 10 foster infants. Development and Psychopathology, 12 (2), 133–156. Stovall-McClough, K., & Dozier, M. (2004). Forming attachments in foster care: Infant attachment behaviors during the first 2 months of placement. Development and Psychopathology, 16(2), 253–271. Strijker, J., Knorth, E., & Knot-Dickscheit, J. (2008). Placement history of foster children: A study of placement history and outcomes in long-term family foster care. Child Welfare: Journal of Policy, Practice, and Program, 87(5), 107–124. Watamura, S., Phillips, D., Morrisey, T., McCartney, K., & Bub, K. (2011). Double jeopardy: Poorer social–emotional outcomes for children in the NICHD SECCYD experiencing home and child care environments that confer risk. Child Development, 82(1), 48–65. Whittaker, J., Jones Harden, B., See, H., Meisch, A., & Westbrook, T. (2011). Family risks and protective factors: Pathways to Early Head Start toddler's social–emotional functioning. Early Childhood Research Quarterly, 26(1), 74–86.

B.J. Harden, J.V. Whittaker / Children and Youth Services Review 33 (2011) 1392–1403 Wu, S., Ma, C., Carter, R., Ariet, M., Feaver, E., Resnick, M., et al. (2004). Risk factors for infant maltreatment: A population-based study. Child Abuse & Neglect, 28(12), 1253–1264. Wulczyn, F. (2004). Family reunification. The Future of Children, 14(1), 95–113. Wulczyn, F., Barth, R., Yuan, Y., Jones Harden, B., & Landsverk, J. (2006). Beyond common sense: Child welfare, child well-being, and the evidence for policy reform. New Brunswick, NJ: Transaction/Aldine de Gruyter.

1403

Wulczyn, F., Kogan, J., & Jones Harden, B. (2003). Placement stability and movement trajectories. The Social Service Review, 77, 212–236. Zimmerman, I. L., Steiner, V. G., & Pond, R. E. (1992). PLS-3: Preschool Language Scale-3. San Antonio, TX: The Psychological Corporation.