© 2014 American Psychological Association 0893-3200/14/$ 12.00 http://dx.doi.org/10.1037/fam0000022

Journal of Family Psychology 2014, Vol. 28, No. 5, 718-727

Family and Neighborhood Disadvantage, Home Environment, and Children’s School Readiness Lieny Jeon, Cynthia K. Buettner, and Eunhye Hur The Ohio State University

The purpose of this study was to examine associations between family socioeconomic risk, neighborhood disadvantage, and children’s school readiness. A sample of 420 children from 48 early childcare programs yielded multi-informant data. The average age was 55.3 months (SD = 6.4), with 38% of children being Black, non-Hispanic, Hispanic, or other minority race (American Indian or Alaska Native, Asian, and Native Hawaiian or Pacific Islander). One third (32.4%) of the parents had annual incomes less than $30,000. We used multilevel structural equation modeling to test direct and indirect associations among family socioeconomic risk and neighborhood disadvantage and children’s cognitive and socialemotional development through home learning environment and parental depression. Children with a greater number of family socioeconomic risks and a higher level of neighborhood disadvantage dem­ onstrated lower scores on cognitive skills. The degree of family socioeconomic risk was indirectly associated with children’s cognitive ability through parents’ cognitive stimulation at home. Parents who had more family socioeconomic risks and neighborhood disadvantage reported more depressive symp­ toms, which, in turn, suggested children’s greater probability of having social-emotional problems. In other words, home learning environments explained associations between family socioeconomic disad­ vantage and children’s cognitive skills, while parental depression explained associations between family/neighborhood disadvantages and children’s social-emotional problems. Results suggest the im­ portance of intervention or prevention strategies for parents to improve cognitive stimulation at home and to reduce depressive symptoms. Keywords: socioeconomic disadvantage, neighborhood, depression, cognitive stimulation, school readiness

children’s vocabulary, literacy, language, and mathematical devel­ opment, and social-emotional competence. Social competence rep­ resents children’s behaviors that allow them to effectively manage and engage in social environment, including positive communica­ tion and interactions with other people (Raver & Zigler, 1997). Emotional competence can be defined as children’s ability to regulate their emotions in physiological or environmental needs (Campos, Mumme, Kermoian, & Campos, 1994). In the current study, home and neighborhood contexts are con­ sidered influential environments for children’s school readiness because preschool-aged children are more likely to face those contexts than other external environments. We model both direct and indirect associations between family and neighborhood disad­ vantage and preschool-aged children’s school readiness through two intervening variables: cognitive stimulation at home and pa­ rental depression. The associations among family/neighborhood disadvantage, cognitive stimulation at home, and child outcomes can be viewed from an investment perspective, and the associations among family/neighborhood disadvantage, parental depression, and child outcomes can be considered from a family stress per­ spective (Guo & Harris, 2000; Yeung, Linver, & Brooks-Gunn, 2002). The investment perspective emphasizes detrimental effects on children’s development in families via socioeconomic home and neighborhood risks as a result of a lack of energy, time, and financial resources to invest in cognitive stimulating learning environments at home. On the other hand, the family stress per-

Preschool-aged children’s cognitive skills and social-emotional competence have critical implications for success in the transition to formal schooling (Mistry, Benner, Biesanz, Clark, & Howes, 2010). Children who are more ready to enter school attain greater academic achievement and increased social-emotional adaptation, which coincide with positive behavioral outcomes (Welsh, Nix, Blair, Bierman, & Nelson, 2010), fewer crimes, and higher rates of employment in later life (Schweinhart et al., 2005). On the other hand, children who experience intellectual or social-emotional difficulties during the preschool period demonstrate lower grades, peer rejection, negative feedback from teachers, and lower levels of self-regulation (Welsh et al., 2010). Therefore, it is important to examine possible indicators that may be associated with children’s school readiness. The current study examines two important ele­ ments of children’s school readiness: cognitive skills, such as

This article was published Online First August 25, 2014. Lieny Jeon, Cynthia K. Buettner, and Eunhye Hur, Department of Human Sciences, The Ohio State University. We acknowledge that this article was developed from Lieny Jeon’s dissertation research, which was supported by the Lucile and Roland Kennedy Scholarship Fund in Home Economics. Correspondence concerning this article should be addressed to Lieny Jeon, Department of Human Sciences, The Ohio State University, 1787 Neil Avenue, 135 Campbell Hall, Columbus, OH 43210. E-mail: jeon [email protected] 718

CHILDREN’S SCHOOL READINESS

spective emphasizes the negative effects on children or adoles­ cents’ developmental abilities via parental distress or depression that were predicted from economic burdens.

Home Environment and Children’s School Readiness Exposure to family risk factors has been associated with young children’s school failure (e.g., Mistry et al., 2010; Whittaker, Harden, See, Meisch, & Westbrook, 2011). Socioeconomic status (SES) represents one of the most important family risk factors associated with children’s school readiness (Crosnoe et al., 2010). Family socioeconomic disadvantage results in fewer social and economic resources for young children and higher levels of intel­ lectual and social-emotional problems in children (RimmKaufman, Curby, Grimm, Brock, & Nathanson, 2009), reducing the likelihood that children will attain appropriate school readi­ ness. There are several dimensions that represent family SES, including household income, parental educational attainment, and family structure. Bronfenbrenner (1986) points out that household income is one of the most important contextual factors of family life in the United States, because financial resources are necessary to support parents and children’s health and wellbeing. Previous studies have indicated that children in poverty often face multiple sociodemographic risks and that the accumulation of risk factors impedes their school readiness (Duncan & Brooks-Gunn, 2000). In addition, children raised in low-income families demonstrate grad­ ual declines in their academic performance during early and mid­ dle childhood (Burchinal, Campbell, Bryant, Wasik, & Ramey, 1997), and studies have calculated that a $1,000 increase in annual income increases 2- to 5-year-old children’s cognitive achieve­ ment by 5% -6% of a standard deviation (Duncan, Morris, & Rodrigues, 2011). Parents’ educational attainment, a primary component of SES, also has been shown to have important influences on children’s literacy skills and social-emotional development (Walker et al., 2011). When parents are more educated, children typically dem­ onstrate greater school readiness (e.g., Johnson, Martin, BrooksGunn, & Petrill, 2008). Family structure, an additional dimension of family SES, has changed considerably over the past several decades due to the increased rates of divorce, nonmarital child­ bearing, cohabitation, and stepfamilies. As a result, children are increasingly living apart from a biological parent (Stewart, 2010). Hofferth (2006) has linked residence with biological and nonbiological married and unmarried parents to more behavioral prob­ lems in children aged 3-12. Children in all family types except the married-biological-parent family demonstrated higher levels of behavioral problems. Moreover, studies have described that chil­ dren with single mothers achieve lower cognitive achievement scores than those with married mothers (e.g., Mistry et al., 2010). Research has examined a variety of socioeconomic risks in the home; however, when the risk factors are highly correlated, it is difficult to simultaneously explain all individual risk factors in the analytic model (Burchinal, Roberts, Hooper, & Zeisel, 2000). Therefore, it is necessary to conceptualize a variety of family socioeconomic risks in an effective way. For example, Sameroff, Seifer, Zax, and Barocas (1987) attempted to understand the ef­ fects of accumulated family risk factors and developed the cumu­ lative risk index, which counts the number of risk factors in a family. They found that this single cumulative index was signifi­

719

cantly associated with children’s IQ scores. Later, empirical stud­ ies found that children reared in families with a large number of negative influences do worse than children in families with few risk factors (e.g., Mistry et al., 2010). In the current study, a cumulative family risk index was created to overcome the limita­ tion of correlations between socioeconomic risk factors. Aside from a direct link between disadvantages in family and child outcomes, previous studies have also demonstrated that fam­ ily SES indirectly affects children’s lives through more proximal factors, such as the home learning environment (Mistry et al., 2010) and parents’ psychological wellbeing (Ertel, Rich-Edwards, & Koenen, 2011). First, a stimulating and linguistically enriched home environment supports children’s language and communica­ tion skills, and it also has implications for children’s social and behavioral functioning (Johnson et al., 2008). Children’s home learning environments are typically explained by access to a li­ brary card, the frequency of reading to the children, and availabil­ ity of reading materials and learning-oriented toys (Caldwell & Bradley, 1984). When parents are at greater socioeconomic risk, parents are more likely to have difficulties providing enriched learning environments for a child due to a lack of time and financial or social resources, which, in turn, may impede children’s cognitive and social skills (Mistry et al., 2010). For example, better educated mothers provided a greater number of books for their children, and their higher reading abilities allowed them to stim­ ulate children’s early reading, which, in turn, was associated with children’s better literacy skills (Johnson et al., 2008). Second, parental depression is one proxy that can explain the negative pathways from family socioeconomic risk to children’s school readiness (Whittaker et al., 2011). Parents’ depressive symptoms hamper mothers’ sensitivity and stimulating caregiving and are associated with negative interactions in the family (Goodman et al., 2011), leading to children’s internalizing and externalizing problems or problems in psychological functioning and deficits in school readiness (Pachter, Auinger, Palmer, & Weitzman, 2006). Parents who had greater socioeconomic cumulative risks were less likely to be emotionally healthy than those who were not at risk (Mistry et al., 2010) because these parents might already be emotionally exhausted by the socioeconomic burden at home. The current study posits pathways from family SES to child outcomes through parental depression and cognitive stimulation at home.

Neighborhood and Children’s School Readiness Living in disadvantaged neighborhoods, where a sizable pro­ portion of residents are poor or are less educated, has been asso­ ciated with children’s development beyond other markers of fam­ ily disadvantage (e.g., Dupere, Leventhal, Crosnoe, & Dion, 2010; Leventhal & Brooks-Gunn, 2000). For example, Xue, Leventhal, Brooks-Gunn, and Earls (2005) demonstrated that young children living in highly disadvantaged neighborhoods had more external­ izing and internalizing behavioral problems than those who were raised in more affluent neighborhoods, after adjusting for child and family demographics. Similarly, Vaden-Kieman et al. (2010) found that low neighborhood SES was associated with Head Start children’s cognitive and behavioral outcomes. Furthermore, stud­ ies have found significant long-term effects of concentrated dis­ advantage in neighborhoods on children’s language and cognitive skills (e.g., Lloyd & Hertzman, 2010; Lloyd, Li, & Hertzman,

JEON, BUETTNER, AND HUR

720

2010). It is thought that the associations between neighborhood disadvantage and negative child outcomes may be due to the influence of peers, a lack of role models and monitoring, less social networks for parents, and scarce resources in the neighbor­ hoods (Froiland, Powell, & Diamond, 2014). On the other hand, the presence of neighborhood resources such as libraries and daycare provides enriched opportunities for young children’s learning, and the availability of monitoring and role models pos­ itively influences children’s development (Kohen, Leventhal, Dahinten, & McIntosh, 2008). However, neighborhood-level expla­ nations may be less influential during the preschool years, because families still play the primary role in a child’s environment during this period. It is suggested that larger socioeconomic structures (e.g., neighborhood) affect an individual’s development through more proximal contexts such as families (e.g., Froiland et al., 2014; Vaden-Kieman et al., 2010). Neighborhood disadvantage impedes children’s development through decreasing the quality of learning experiences within the family and/or through increasing the likelihood of parental depres­ sion (Dupere et al., 2010). Adults living in more disadvantaged neighborhoods are more likely to be exposed to stressors, such as noise, pollution, crime, and disorder, all of which appear to be harmful to emotional health (Hill, Ross, & Angel, 2005). In addition, less-advantaged neighborhoods have fewer infrastruc­ tures, which may impede parents’ opportunities to access and provide literature resources for children (Neuman & Celano, 2001). These detrimental effects for parents, in turn, impact chil­ dren’s outcomes (Vaden-Kieman et al., 2010). Kohen et al. (2008) suggests that neighborhood SES may be associated with key components of family variables, such as parental mental health and parenting behaviors, in a similar manner to how family SES influences more proximal home environments. They found that even though there was no direct association between neighborhood SES and children’s verbal ability and behavior problems, there were indirect effects between neighborhood SES and the child outcomes through maternal depression and parenting styles after controlling for family SES (Kohen et al., 2008). This suggests that both neighborhood stmctures and family factors should be consid­ ered simultaneously as important factors in the process of how neighborhood environments are associated with children’s out­ comes. A number of studies have examined the relationships between neighborhood disadvantage and elementary schoolchil­ dren’s or adolescents’ development (e.g., Dearing et al., 2009), but relatively fewer studies examined the associations between neigh­ borhood contexts and preschool-aged children’s school readiness. We expect that neighborhood disadvantage works differently for preschool-aged children because they have a different exposure to extrafamilial influences (Duncan, Brooks-Gunn, & Klebanov, 1994; Vaden-Kiernan et al., 2010).

The Present Study Taken together, existing work will be extended by simultane­ ously examining family and neighborhood environments as factors related to children’s cognitive and social-emotional competence. Even though there are a few studies that examined the investment perspective and the family stress perspective in the context of family (Yeung et al., 2002) or neighborhood (Kohen et al., 2008), there is a lack of studies that simultaneously considered the rela­

tionships between family and neighborhood socioeconomic disad­ vantages and child outcomes via these two paths. In addition, previous studies focused primarily on household income in the investment perspective model or in the family stress model rather than considering multiple indicators of socioeconomic risk. We first hypothesized that children in more disadvantaged family and neighborhood environments will be more likely to exhibit low levels of cognitive skills and social-emotional problems. Second, we hypothesized that parents in more disadvantaged families and neighborhoods will be less likely to provide cognitive stimulation at home and more likely to exhibit depressive symptoms, which, in turn, will be associated with lower levels of children’s cognitive skills and a greater likelihood of having social-emotional prob­ lems.

Method Dataset The data used in the current study is from a larger study that evaluated the effectiveness of a Quality Rating and Improvement System (QRIS), which is a statewide strategy to assess the quality of childcare programs. A total of 48 full-time center-based early childcare and education programs were randomly selected from a list of all licensed childcare programs in a Midwestern state. After administrators confirmed participation in the study, approximately 10 children from each program were randomly selected for inclu­ sion in the study. Parents received a consent form, a questionnaire, and a return envelope via the administrators, and children whose parents consented to participate in direct assessments (e.g., literacy and math skills) were observed by our trained research assistants (n = 446). For the current study, 10 children were dropped from the sample of assessed children because their parents did not return a parent questionnaire that included demographics and home en­ vironment information (n = 436). The QRIS parent/child dataset was linked to 2006-2010 American Community Survey 5-year estimates data at the census tract level to characterize the neigh­ borhood environments. Children’s census tracts were identified via their addresses. Among the 436 children in the QRIS dataset, 15 children were dropped: 12 of the children’s addresses were miss­ ing, and there were 3 sibling pairs who had the same address. In the final analysis, one outlier on a vocabulary test (Peabody Picture Vocabulary Test Third Edition) was deleted, which yielded a total sample size of 420.

Participants Of the 420 preschool-aged children in the sample, mean age was 55.3 months with a 6.4 standard deviation (SD), and approximately half of the children were girls (46.8%). More than half of the children were White, non-Hispanic (61.9%); 20.9% were Black, non-Hispanic; 7.2% were Hispanic; and 10.6% were other race including multiracial, American Indian or Alaska Native, Asian, and Native Hawaiian or Pacific Islander. The average annual household income was from $50,001 to $75,000, with 32.4% of the parents having an annual income less than $30,000. Almost 60% of the parents had at least a high school degree. Additionally, 38.0% of respondents were single, separated, divorced, or wid­ owed. The rest of parents were married or cohabiting. Respondents

CHILDREN’S SCHOOL READINESS

to the parent questionnaire were primarily mothers (88.9%), 8.8% were fathers, 2.1% were grandparents, and one respondent was a legal guardian (0.2%).

Analytic Sample In total, 420 children from 354 different census tracts were included in the analytic sample. Comparisons of children in the analytic sample (n = 420) to the broader sample of the QRIS study (n = 436) revealed no significant differences in children’s age, gender, race/ethnicity, household income, parental educational at­ tainment, and marital status. Although the same number of pro­ grams was randomly selected from each level, the number of programs in the population was unevenly distributed across QRIS status. In this case, parameter estimates can be biased due to unequal selection probabilities (Stapleton, 2002). Weighting the sample, therefore, was necessary to better generalize our findings to the larger population of all preschool-aged children enrolled in licensed full-time child care centers in the state.

Measures Children’s cognitive skills. Trained research assistants mea­ sured children’s literacy and mathematical skills using three direct assessments: the Peabody Picture Vocabulary Test-Third Edition (Dunn & Dunn, 1997), the Phonological Awareness Literacy Screening-Prekindergarten (Invernizzi, Sullivan, Meier, & Swank, 2004), and a subtest from the Woodcock-Johnson Test of Achievement III (Woodcock, McGraw, & Mather, 2001). The Peabody Picture Vocabulary Test-Third Edition measures chil­ dren’s receptive vocabulary, which represents verbal ability (Dunn & Dunn, 1997). This is a norm-referenced test that reflects each child’s performance relative to the expected performance of chil­ dren in the population who are the same age, converting raw scores into standardized scores. The Phonological Awareness Literacy Screening-Prekindergarten is a measure of preschooler’s early phonological awareness and literacy skills that are predictive of future reading success (Invernizzi et al„ 2004). This measure tested alphabet recognition, letter sounds recognition, beginning sound awareness, print and word awareness, and rhyme awareness. A single variable was created to represent children’s phonological awareness by a sum of subtests (Cronbach’s alpha = .85). To measure children’s mathematical abilities, a subtest, Applied Prob­ lems, from the Woodcock-Johnson Test of Achievement HI (Woodcock et al., 2001) was used. The Applied Problems subtest scores on the ability to analyze and solve math problems. Children are asked to solve practical problems through deciding mathemat­ ical operation to be used, and conducting simple calculations. The Woodcock-Johnson Test of Achievement III provides an agestandardized score for each raw score. Children’s social-emotional competence. Children’s social and emotional competence was measured using parents’ reports on the Ages and Stages Questionnaire: Social-Emotional (Squires, Bricker, & Twombly, 2002). The Ages and Stages Questionnaire: Social-Emotional is a child-monitoring system for social and emo­ tional behaviors that was developed to identify a child who needs referral considerations (Squires et al., 2002). Parents responded to 35 questions on a 3-point scale regarding how frequently the children exhibit certain problems in behaviors (10 = most o f the

721

time, 5 = sometimes, and 0 = rarely or never). The Ages and Stages Questionnaire: Social-Emotional items represent seven ar­ eas: self-regulation, compliance, communication, adaptive func­ tioning, autonomy, affect, and interaction with people. The sum of 35 items was calculated (M = 38.3, SD = 26.8, range = 0-140) to make higher scores represent higher levels of problems in social-emotional competence. However, the continuous score was not considered in the analysis because it did not reflect a normal distribution. Non-normality is frequently found in a scale measur­ ing behavioral problems because there are more children who do not have behavioral programs in the general population (Squires et al., 2002). As suggested by Squires et al. (2002), a cutoff score was calculated to transform into a binary variable (1 = having socialemotional problems, 0 = not having social-emotional problems) using the semi-interquartile range (i.e., median + [quartile 3 quartilel]/2). The cutoff score in this study was 45. There were 118 children (28.8%) in a having social-emotional problems (i.e., recommended referral consideration) category. Family socioeconomic risk. Family socioeconomic risk was measured by counting the number of family socioeconomic risks as reported by parents. Three indicators were singled out as an index: single parent status (dummy coded as 1 = single, separated, divorced or widowed and 0 = married or cohabiting), parents’ educational level (dummy coded as 1 = less than an associate o f arts [AA] degree and 0 = at least an AA degree), and family income (dummy coded as 1 = annual income less than $30,000 and 0 = more than $ 30,001). We used an AA degree as a cutoff for the educational attainment because it is the indicator used to determine the level of higher education in the state where we collected the data (Lumina Foundation, 2013). In addition, the poverty line cutoff used in this study was $30,000 because it is near to the state’s eligibility for free and reduced lunch for a threeto four-person household. The Cronbach’s alpha for three indica­ tors in the current study was .77. Neighborhood disadvantage. Following previous work (Sampson, Morenoff, & Gannon-Rowley, 2002), five characteris­ tics of census tracts representing concentrated disadvantage were extracted from the U.S. Census Bureau data: the percentage below the poverty line, the unemployment ratio, the percentage of female-headed households with children, the percentage of people receiving public assistance, and the percentage of African Amer­ icans (e.g., Burchinal, Nelson, Carlson, & Brooks-Gunn, 2008; Sampson et al., 2002). In addition, the percentage of people receiving food stamps was newly added in this study (Cronbach’s alpha for 6 indicators = .75). Because this scale was established using 1990 U.S Census data, exploratory factor analysis was conducted to verify the single factor structure. Indicators were standardized and averaged to represent a measure of neighborhood disadvantage. Home environments. The quality of cognitive stimulation at home and parental depression were measured by parents’ re­ sponses to questions about home environments. First, the degree of cognitive stimulation was measured by the Home Observation of the Environment-Short Form, which was adapted from the Na­ tional Longitudinal Survey of Youth child survey for use as a parent-report form (Center for Human Resources Research, 1993). The Home Observation of the Environment-Short Form measures the quality of the home learning environment, that is, how well parents stimulate cognitive development by 10 items asking about

722

JEON, BUETTNER, AND HUR

literacy environments for a child. Sample items include, “About how often do you read stories to your child?” and “About how many children’s books does your child have?” The items were recoded into 0 if the child does not receive stimulation from the parent or 1 if the parent provides cognitive stimulation to the child based on the recoding criteria provided by National Longitudinal Survey of Youth. Ten items were summed to represent a total score of home cognitive stimulation (Cronbach’s alpha = .54). Adequate constructive validity has been documented in previous studies using the National Longitudinal Survey of Youth (e.g., Mott, 2004). Second, the short form of Center for Epidemiological Study of Depression Scale (Radloff, 1977) was used to measure parents’ depressive symptoms. Parents responded to a total of nine items describing the feelings that they had during the past week, using a 4-point scale (1 = rarely or none o f the time [less than 1 day], and 4 = most or all o f the time [5-7 days]). For example, the item that “You did not feel like eating; your appetite was poor” was asked. Nine items were summed to a single score of depression (Cron­ bach’s alpha = .76). Control variables. Child age, gender (dummy coded as 1 = girls and 0 = boys), and race/ethnicity (dummy coded into three dichotomous variables: White, non-Hispanic (reference category); Black, non-Hispanic; Hispanic; and Other race) were included as covariates.

Data Analytic Strategy As preliminary analyses, exploratory factor analysis (EFA) us­ ing an oblique rotation, which allows the factors to be correlated (Ford, MacCallum, & Tait, 1986), was conducted in Mplus 7.0 (Muthen & Muthen, 2012). We entered all family socioeconomic risk and neighborhood disadvantage indicators testing whether there are two distinct factors that separately represent family socioeconomic risk and neighborhood disadvantage— considering an association between family- and neighborhood-related charac­ teristics, as the literature suggests (e.g., Dupere et al., 2010). The number of factors was determined via the examination of eigen­ values greater than 1 and comparing the model fit (Ford et al., 1986). In addition, variables with factor loadings greater than .40 on a factor were considered significant and were used to reflect the factor (Ford et al., 1986). Another EFA was conducted to examine whether three assessments measuring children’s cognitive skills could be reduced to a single factor. We assessed the measurement model fit in the subsequent multilevel confirmatory factor analysis (CFA) by multiple goodness-of-fit indices: (a) a p value of chisquare statistic (x2) larger than .05 (however, a significant \ 2 is still acceptable until other fit indices are considered, because the y2 is sensitive for the sample size; Bollen, 1989), (b) a Comparative Fit Index (CFI) of .90 or higher, and (c) a root mean square error of approximation (RMSEA) less than .06 (Browne & Cudeck, 1993). In primary analyses, two-level multilevel structural equa­ tion modeling (MSEM) was conducted in Mplus 7.0 (Muthen & Muthen, 2012) to simultaneously test research questions. This allows examination of the direct and indirect associations among observed and latent variables while controlling for nested data structure. Even though independent variables, mediating variables, and outcome variables are all within-level variables in the model, MSEM should be considered to account for random effects from the nested data structure (Preacher, Zyphur, & Zhang, 2010). In the

current study, because the original data clearly have a multilevel structure with children (Level 1) nested within preschools (Level 2),*1 MSEM has an advantage over the traditional structural equa­ tion modeling in that it estimates more accurate standard errors and random effects from a clustered design (Preacher et al., 2010). The maximum likelihood with robust standard error estimator was used to adjust non-normality of some indicators and the nonindepen­ dence of observations due to the nested neighborhood census tracts. The sample weight was added at the between-program level in all analyses. The model fit was tested through comparing the log-likelihoods for nested models (i.e., null model and alternative model), because the MSEM with a categorical outcome does not produce model fit indices. Missing data (0.7%-5.5% of missing for each variable) was handled in the model using full information maximum likelihood estimation. This method is preferred because it offers less biased estimators over other traditional approaches (Acock, 2005).

Results Preliminary Analyses The EFA with family socioeconomic risk and neighborhood disadvantage indicators revealed that there were two eigenvalues greater than 1, indicating that the two-factor model was adequate. As we expected, the first factor consisted of three family socio­ economic risk indicators (single parent status, parents’ educational level, and family income; factor loadings between .72 and .99), and six neighborhood indicators loaded on the second factor (the unemployment ratio and the percentage below the poverty line, female-headed households with children, people receiving public assistance, minority concentration, and people receiving food stamps; factor loadings between .78 and .96) with an adequate model fit, x2 (19, N = 420) = 75.48, p < .01, RMSEA = .06, CFI = .96. Because there were two distinct factors reflecting family socioeconomic risk and neighborhood disadvantage, we calculated two separate single scores summing risk indicators and added them to MSEM. Another EFA using children’s cognitive outcomes suggested a one-factor model. In the following multi­ level CFA, three assessments measuring vocabulary, literacy skill, and math ability significantly reflected a single factor with an excellent model fit, \ 2 (3, N = 420) = 5.12, p > .05, RMSEA = .04, CFI = .99. Table 1 shows descriptive statistics and bivariate correlations for key measures.

Multilevel Structural Equation Modeling The hypothesized model was tested in MSEM. MSEM with intervening variables (i.e., parent-perceived cognitive stimulation at home and parental depression) fit the data better than the unconditional model and the direct association model without 1 In the current study, 15.7% of the children were nested within the neighborhood level. The observations per census tract group ranged from 1 to 4, with an average of 1.2. However, we conducted two-level MSEM (child level and childcare level) rather than three-level MSEM, which accounts for the neighborhood level, because there was no significant variation in child outcomes at the neighborhood level after accounting for the childcare variation.

CHILDREN’S SCHOOL READINESS

723

Table 1

Descriptive Statistics and Bivariate Correlations

M (SE)I%

Variable 1. Family socioeconomic risk 2. Neighborhood disadvantage 3. Parental depression 4. Home cognitive stimulation 5. Cognitive skills PPVT standardized score PALS sum: Literacy skills WJ: Applied Problems 6. Social-emotional problem

1.31 (.07) 0.00 (.04) 3.45 (.20) 8.73 (.07) 102.01 (.70) 59.01 (1.62) 13.36 (.23) 28.71%

Range 0-4 -.97-3.5 0-21 3-10 60-140 1-108 0-24 O-l

n

Factor loadings

397 420 410 415 418 420 419 411

_

Bivariate correlations a .77 .91 .76 .54

— —

— .54*** .65*** .88*** —

.85

1

2

3

4

5

6

_ .37*’* .28*** -.40***

.19*** -.2 2 ’**

-.14**

—

-.49***

-.37**’

-.13**

.36***

.16*’*

.18***

— —

—

—

-

.17***

-.1 8 ’**

-.2 5 ’**

—

Note.

Descriptive statistics and measurement model fit were weighted. M = mean; SE = weighted standard error; PPVT = Peabody Picture Vocabulary Test-Third Edition; PALS = Phonological Awareness Literacy Screening-Prekindergarten; WJ = Woodcock-Johnson Test of Aclievement-III.

" > < . 01 . * * > < . 001.

intervening variables (log-likelihood = —4,687.94, —2 loglikelihood ratio test statistics = 249.78, d f = 28, p < .01, com­ paring to the direct association model). Although the bivariate associations between family/neighborhood socioeconomic risks and children’s cognitive and social-emotional outcomes were sig­ nificant, in MSEM, the direct associations were only supported betw een family or neighborhood socioeconomic risks and chil­ dren’s cognitive skills as illustrated in Figure 1. Children with a greater number of family socioeconomic risks had significantly lower scores on cognitive skills after controlling for neighborhood disadvantage, home environments, children’s age, gender, and race/ethnicity. In addition, children in more disadvantaged neigh­ borhoods demonstrated significantly lower levels of cognitive achievement after controlling for family socioeconomic risk, home environments, and covariates. The coefficient for the family so­ cioeconomic risk (p = —.21) was small to moderate and the coefficient for the neighborhood was small ((3 = —.13). To test indirect associations, first, the paths from fam ily/neigh­ borhood socioeconomic disadvantage to the intervening variables

were examined. Figure 1 shows that parents having a greater number o f family socioeconomic risks reported more depressive symptoms and lower levels o f parent-reported cognitive stimula­ tion at home than parents who had a fewer number o f family socioeconomic risks. In addition, neighborhood disadvantage was significantly associated with parental depression but was not as­ sociated with parent-reported cognitive stimulation. Second, al­ though parental depression was not significantly associated with children’s cognitive skills, parent-reported stimulating learning environments at home was significantly associated with better child cognitive abilities after controlling for family/neighborhood socioeconomic risks, parental depression, and covariates. In addi­ tion, children in better parent-reported stimulating home environ­ ments demonstrated less likelihood o f having social-emotional problems after adjusting for family/neighborhood socioeconomic risks, parental depression, and covariates (odds ratio [O/?] = .84). The OR suggested that the odds of having social-emotional prob­ lems decreased by 16% when the degree o f parent-reported cog­ nitive stimulation at home increased by one unit. On the other

Between-program (Child-care level) Within-program (Child level)

Figure 1. The results from the model of the associations between family/neighborhood disadvantage and child outcomes through home environments. Unstandardized path coefficients and standard errors are reported. A dashed line indicates nonsignificant effects. Child age, gender, and race/ethnicity were controlled for. f p < . 10. * p < .05. * > < .01. * * > < .001.

JEON, BUETTNER, AND HUR

724

advantage, home environments, and children’s school readiness. The direct associations between family/neighborhood disadvan­ tage and child outcomes were only supported for children’s cog­ nitive skills, not for social-emotional functioning. In general, there were indirect associations through parent-reported cognitive stim­ ulation at home and parental depression. However, the pathways through intervening variables differed by specific child outcome. Consistent with previous studies (e.g., Guo & Harris, 2000; Yeung et al., 2002), the investment perspective (i.e., socioeconomic risk —» learning environment at home —» child outcomes) was more strongly supported in the association between family socioeco­ nomic risk and children’s cognitive skills. This may be because socioeconomically disadvantaged parents lack resources for in­ vestments in the type of cognitive stimulating home environments that can be beneficial to children’s learning opportunities. Re­ search has found a greater relationship between learning materials (e.g., books, magazines, or recorders) and activities such as shared reading with children’s literacy or mathematical skills (Johnson et al., 2008) than with social-emotional development. On the other hand, the association between socioeconomic risk in home/neighborhoods and children’s social-emotional function­ ing may be better explained by the family stress model (i.e., socioeconomic risk —» parental depression —* child outcomes). Interestingly, neither family socioeconomic risk nor neighborhood disadvantage was directly associated with the probability of chil­ dren having social-emotional problems. Rather, disadvantages at home and in neighborhoods were indirectly associated with chil­ dren’s social-emotional development through parental depression. It is notable to find that the number of socioeconomic difficulties at home, rather than household income, per se, was associated with parental depression and children’s social-emotional competence. Parents who have a greater number of socioeconomic risks and who live in disadvantaged neighborhoods may feel more pressure

hand, children having more depressed parents demonstrated greater odds of having social-emotional problems {OR = 1.08), indicating that the odds of having social-emotional problems in­ creased by 8% with a one unit increase in parental depression after controlling for family/neighborhood risks, parent-reported cogni­ tive stimulation, and covariates. The indirect associations were calculated by the products of the paths from family/neighborhood socioeconomic risks to home environments, and the paths from home environments to child outcomes. As shown in Table 2, the cumulative family socioeco­ nomic risk was significantly associated with children’s cognitive skills through parent-reported cognitive stimulation at home after controlling for covariates. This indicates that parents with a greater number of family socioeconomic risks had less cognitive stimula­ tion at home, which, in turn, suggested lower levels of children’s cognitive achievement. Although there was no significant direct association between family socioeconomic risk and children’s social-emotional functioning, there were statistically significant indirect associations between family socioeconomic risk and social-emotional competence through parental depression after controlling for covariates. The 1.06 odds ratio suggested that when family socioeconomic risk increases by one unit, the odds of having social-emotional problems increases by 6% through paren­ tal depression. Finally, children in more disadvantaged neighbor­ hoods exhibited higher odds of having social-emotional problems through parental depression. In MSEM, 49.8% of the variance in children’s cognitive skills and 12% of the variance in socialemotional problems was explained by the independent variables.

Discussion Following the investment perspective and the family stress perspective, the aim of this study was to understand the links between cumulative family socioeconomic risk, neighborhood dis­

Table 2 Results From the Model of the Associations Between Family/Neighborhood Disadvantage and Child Outcomes Through Home Environments Cognitive skills B Covariates Child age Child sex Race/ethnicity Black Hispanic Other race Independent variables Family socioeconomic risk Neighborhood disadvantage Intervening variables Cognitive stimulation Parental depression Indirect associations Family socioeconomic risk —» Cognitive stimulation Family socioeconomic risk —» Parental depression Neighborhood disadvantage —* Cognitive stimulation Neighborhood disadvantage —* Parental depression Note. N = 420. OR = odds ratio. > < . 1 0 . > < . 0 5 . * > < .0 1 .

* * > < .0 0 1 .

.68*** 1.18

SE .06 .74

95% Cl [.55, .80] [-.26, 2.63]

Social-emotional problems

P .53 .07

B

SE

OR

95% Cl for OR

-.03* .11

.02 .23

0.97 1.11

[.94, .99] [.70, 1.75]

.36 .43 .38

2.98 2.21 2.36

[1.48,6.00] [0.94,5.15] [1.12,5.15]

1.09** .79* .86*

-4.48** -3.37* -0.29

1.42 1.36 .95

[-7.26, -0.45] [-6.03, -.71] [-2.15, 1.57]

-.01

-1.24** -1.29*

.40 .53

[-2.03, - .45] [-2.33, -.25]

-.21 -.13

-.0 8 .14

.10 .18

0.93 1.15

[.77, 1.12] [.81, 1.63]

1.04*** -0.14

.32 .09

[.40, 1.67] [-.31, .04]

.17 -.0 7

-.18* .08*

.09 .03

0.84 1.08

[.71, 1.00] [1.01, 1.15]

-.34** -.1 0 .00 -.11

.12 .07 .10 .08

[-.57, [-.24, [-.19, [-.26,

-.0 6 -.0 2 .00 -.01

.06* .06* .00 .06*

.03 .03 .02 .04

1.06 1.06 1 1.06

[1.00, 1.13] [1.00, 1.12] [.97, 1.03] [.99, 1.14]

-.12] .04] .19] .05]

-.2 2 -.1 1

CHILDREN’S SCHOOL READINESS

and may be emotionally vulnerable due to a lack of available physical and psychological resources (Duncan et al., 1994). That is, the pressure and stress from socioeconomic difficulties may be reflected in parents’ depressive symptoms. Parents’ depressed mood has been found to be a strong predictor of low levels of sensitivity and unhealthy interactions in the family (Petterson & Albers, 2001), which, in turn, may hamper children’s healthy social-emotional development (Whittaker et al., 2011). Even though researchers have adapted the family stress model into neighborhood contexts, only a few studies have found significant associations between neighborhood disadvantage and child out­ comes through parental depression (Kohen et al., 2008). The findings in the current study indicate that it is reasonable to consider the detrimental association between neighborhood disad­ vantage and children’s social-emotional outcome in a similar man­ ner to how the family stress perspective explains the association between family disadvantage and child outcomes through parental depression. The family stress model in a neighborhood context may be explained by collective efficacy theory, which emphasizes the importance of mutual trust and solidarity in neighborhoods (Sampson et al., 2002). According to the collective efficacy model, parents in disadvantaged neighborhoods have less trust in the social networks in their community, which, in turn, might be associated with social isolation and mental health problems such as depression and anxiety for both parents and children (Xue et al., 2005).

Limitations The present study has several limitations. First, the causal rela­ tionships cannot be guaranteed in this cross-sectional observa­ tional study. Because children are not randomly assigned to neigh­ borhoods, there are potential biases from selection and omitted or unexplained variables. For example, there is a possibility that neighborhood choice and children’s outcomes are affected by parents’ and family characteristics other than family socioeco­ nomic disadvantage, such as parental beliefs and attitudes, cultural norms, or perceived social support (e.g., Burchinal et al., 2008; Xue et al., 2005), which are not examined in the current study. Second, the present study used parents’ self-reported responses to measure household/parent characteristics (family socioeconomic risk, cognitive stimulation, and depression) and children’s socialemotional functioning. Parents may over-report on their home environments or on their children’s social-emotional skills. It is also possible that depressed parents perceived their children as having more social-emotional problems (Kashani, Orvaschel, Burk, & Reid, 1985). Thus, the results of the pathways from family socioeconomic risk to social-emotional development through parent-reported cognitive stimulation and depressive symptoms may be biased due to shared method variance. Third, neighbor­ hood disadvantage was characterized by census tract level data, which was matched with children’s home addresses. However, parents and children might experience multiple neighborhoods (e.g., location of childcare or job) in their everyday experience (Wheaton & Clarke, 2003), and it might change the influence of the neighborhoods where they live. In addition, the length of time living in the census tract could not be accounted for in this study. Further, although the census tract level data was the most repre­ sentative data available for this study, block group level data might

725

be more preferred in a statewide sample, especially for people who live in suburban and rural areas. Fourth, because we used the cumulative risk index, it is not possible to evaluate the relative importance of each risk indicator. However, using the cumulative index helps to reduce the correlations among the indicators and modeling with a nonsufficient sample size (Burchinal et al., 2000). Finally, the majority of participants in this study were European American and middle-class parents. Therefore, it limits generaliz­ ing our findings to a broader population. For example, in a sample of more disadvantaged families and children, it is possible that there is a different association and process between family/neighborhood socioeconomic risks and children’s school readiness be­ cause there might be a lack of resources (e.g., social support and childcare availability) for parents in a more disadvantaged area (Dupere et al., 2010). In addition, although the weighting strategy was utilized to generalize the findings to the targeted population (preschool-aged children in full-time child care centers in the state), the findings cannot be generalized to other states or to families who do not use childcare programs.

Implications for Future Research and Practice Despite limitations, the findings of the current study suggest several implications for future research and practice. The current study investigated the investment perspective and family stress perspective in family and neighborhood contexts. The mechanisms of these two perspectives (e.g., mediated mediation) are an inter­ esting topic for further exploration. For example, the effects of parental depression and cognitive stimulation, which were associ­ ated with family/neighborhood risk, on child outcomes might be mediated by parent-child interaction, the degree of household chaos, or emotional support. Another pathway of how neighbor­ hood disadvantage influences parental depression could be also examined; one example of possible mediators might be collective efficacy as discussed previously. Furthermore, longitudinal studies are needed to test the lasting effects of socioeconomic risks on child outcomes through parental depression and cognitive stimu­ lation. Although there was a significant direct association between neighborhood disadvantage and children’s cognitive skills, the indirect associations were small or nonsignificant. Future research should explore the predictors that mediate the relationships be­ tween neighborhood disadvantage and child outcomes. In practice, because family socioeconomic risk and neighbor­ hood disadvantage are generally associated with child outcomes through parental depression and the home learning environment, intervention programs targeting parents may reduce the negative effects of socioeconomic risks in family and neighborhood. Al­ though the pathways worked differently for specific outcomes, a package of interventions dealing with parents’ psychological well­ being, providing learning materials, and educating parents to cre­ ate cognitive stimulating learning environments (e.g., how to read books to a child) might be considered to increase children’s overall school readiness, including cognitive ability and social-emotional development. Direct investment in socioeconomic family/neigh­ borhood risks requires a large scale investment of public resources (Duncan et al., 1994). Interventions for parents, which can be implemented at the individual-level, might be one efficient way to prevent further difficulties in families and child outcomes. Because families are dynamic systems, it is important to intervene in family

JEON, BUETTNER, AND HUR

726

processes through integrated parenting programs rather than inter­ vening for one specific target problem.

References Acock, A. C. (2005). Working with missing values. Journal of Marriage and Family, 67, 1012-1028. doi: 10.1111/j. 1741 -3737.2005.00191 .x Bollen, K. A. (1989). Structural equations with latent variables. New York, NY: Wiley. Bronfenbrenner, U. (1986). Ecology of the family as a context for human development: Research perspectives. Developmental Psychology, 22, 723-742. doi: 10.1037/0012-1649.22.6.723 Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen & J. S. Long (Eds.), Testing structural equation models (pp. 136-162). Newbury Park, CA: Sage. Burchinal, M. R., Campbell, F. A., Bryant, D. M„ Wasik, B. H., & Ramey, C. T. (1997). Early intervention and mediating processes in cognitive performance of children of low-income. Child Development, 68, 935954. doi: 10.2307/1132043 Burchinal, M. R., Nelson, L., Carlson, M., & Brooks-Gunn, J. (2008). Neighborhood characteristics, and child care type and quality. Early Education and Development, 19, 702-725. doi: 10.1080/ 10409280802375273 Burchinal, M. R., Roberts, J. E., Hooper, S., & Zeisel, S. A. (2000). Cumulative risk and early cognitive development: A comparison of statistical risk models. Developmental Psychology, 36, 793-807. doi: 10.1037/0012-1649.36.6.793 Caldwell, B., & Bradley, R. H. (1984). Home observation for measurement of the environment. Little Rock, AR: University of Arkansas at Little Rock. Campos, J. J., Mumme, D. L., Kermoian, R., & Campos, R. G. (1994). A functionalist perspective on the nature of emotion. Monographs o f the Society for Research in Child Development, 59, 284-303. doi: 10.1111/ j. 1540-5834.1994.tb01289.x Center for Human Resources Research. (1993). NLSY user’s guide. Co­ lumbus, OH: Ohio State University. Crosnoe, R., Leventhal, T., Wirth, R. J., Pierce, K. M., Pianta, R. C., & NICHD Early Child Care Research Network. (2010). Family socioeco­ nomic status and consistent environmental stimulation in early child­ hood. Child Development, 81, 972-987. doi:10.1111/j.1467-8624.2010 ,01446.x Dearing, E., Wimer, C., Simpkins, S. D., Lund, T., Bouffard, S. M., Caronongan, P., . . . Weiss, H. (2009). Do neighborhood and home contexts help explain why low-income children miss opportunities to participate in activities outside of school? Developmental Psychology, 45, 1545-1562. doi:10.1037/a0017359 Duncan, G. J., & Brooks-Gunn, J. (2000). Family poverty, welfare reform, and child development. Child Development, 71, 188—196. doi: 10.1111/ 1467-8624.00133 Duncan, G. J., Brooks-Gunn, J., & Klebanov, P. K. (1994). Economic deprivation and early childhood development. Child Development, 65, 296-318. doi: 10.2307/1131385 Duncan, G. J., Morris, P. A., & Rodrigues, C. (2011). Does money really matter? Estimating impacts of family income on young children’s achievement with data from random-assignment experiments. Develop­ mental Psychology, 47, 1263-1279. doi:10.1037/a0023875 Dunn, L. M., & Dunn, L. M. (1997). Peabody picture vocabulary test, third edition. Circle Pines, MN: American Guidance Services. Dupere, V„ Leventhal, T., Crosnoe, R., & Dion, E. (2010). Understanding the positive role of neighborhood socioeconomic advantage in achieve­ ment: The contribution of the home, child care, and school environ­ ments. Developmental Psychology, 46, 1227-1244. doi:10.1037/ a0020211 Ertel, K. A., Rich-Edwards, J. W., & Koenen, K. C. (2011). Maternal depression in the United States: Nationally representative rates and risks.

Journal o f Women’s Health, 20, 1609-1617. doi:10.1089/jwh.2010 .2657 Ford, J. K., MacCallum, R. C., & Tait, M. (1986). The application of exploratory factor analysis in applied psychology: A critical review and analysis. Personnel Psychology, 39, 291-314. doi: 10.1111/j.1744-6570 .1986.tb00583.x Froiland, J. M., Powell, D. R., & Diamond, K. E. (2014). Relations among neighborhood social networks, home literacy environments, and chil­ dren’s expressive vocabulary in suburban at-risk families. School Psy­ chology International, 35, 429-444. doi:10.1177/0143034313500415 Goodman, S. H., Rouse, M., Connell, A., Broth, M., Hall, C., & Heyward, D. (2011). Maternal depression and child psychopathology: A metaanalytic review. Clinical Child and Family Psychology Review, 14, 1-27. doi:10.1007/s 10567-010-0080-1 Guo, G., & Harris, K. M. (2000). The mechanisms mediating the effects of poverty on children’s intellectual development. Demography, 37, 431— 447. doi: 10.1353/dem.2000.0005 Hill, T. D„ Ross, C. E., & Angel, R. J. (2005). Neighborhood disorder, psychophysiological distress, and health. Journal o f Health and Social Behavior, 46, 170-186. doi: 10.1177/002214650504600204 Hofferth, S. L. (2006). Residential father family type and child well-being: Investment versus selection. Demography, 43, 53-77. doi: 10.1353/dem .2006.0006 Invemizzi, M., Sullivan, A., Meier, J., & Swank, L. (2004). Phonological awareness literacy screening: Pre-K teacher manual. Charlottesville, VA: University of Virginia. Johnson, A. D., Martin, A., Brooks-Gunn, J., & Petrill, S. A. (2008). Order in the house! Associations among household chaos, the home literacy environment, maternal reading ability, and children’s early reading. Merrill-Palmer Quarterly, 54, 445-472. doi:10.1353/mpq.0.0009 Kashani, J. H., Orvaschel, H., Burk, J. P., & Reid, J. C. (1985). Informant variance: The issue of parent-child disagreement. Journal o f the Amer­ ican Academy o f Child Psychiatry, 24, 437-441. doi: 10.1016/S00027138(09)60561-6 Kohen, D. E., Leventhal, T., Dahinten, V. S., & McIntosh, C. N. (2008). Neighborhood disadvantage: Pathways of effects for young children. Child Development, 79, 156-169. doi: 10.1111/j. 1467-8624.2007 ,01117.x Leventhal, T., & Brooks-Gunn, J. (2000). The neighborhoods they live in: The effects of neighborhood residence on child and adolescent. Psycho­ logical Bulletin, 126, 309-337. doi: 10.1037/0033-2909.126.2.309 Lloyd, J. E. V., & Hertzman, C. (2010). How neighborhoods matter for rural and urban children’s language and cognitive development at kin­ dergarten and grade 4. Journal o f Community Psychology, 38, 293-313. doi: 10.1002/jcop.20365 Lloyd, J. E. V., Li, L., & Hertzman, C. (2010). Early experiences matter: Lasting effect of concentrated disadvantage on children’s language and cognitive outcomes. Health & Place, 16, 371-380. doi: 10.1016/j .healthplace.2009.11.009 Lumina Foundation. (2013). A stronger Ohio through higher education. Retrieved from http://www.luminafoundation.org/stronger_nation_2013/ downloads/pdfs/ohio-brief-2013.pdf Mistry, R. S., Benner, A. D., Biesanz, J. C., Clark, S. L., & Howes, C. (2010). Family and social risk, and parental investments during the early childhood years as predictors of low-income children’s school readiness outcomes. Early Childhood Research Quarterly, 25, 432-449. doi: 10.1016/j .ecresq.2010.01.002 Mott, F. L. (2004). The utility of the HOME-SF Scale for child develop­ ment research in a large national longitudinal survey: The National Longitudinal Survey of Youth 1979 Cohort. Parenting: Science and Practice, 4, 259-270. doi:10.1080/15295192.2004.9681273 Muthen, L. K., & Muthen, B. O. (2012). Mplus user’s guide. Los Angeles, CA: Authors.

CHILDREN’S SCHOOL READINESS Neuman, S. B., & Celano, D. (2001). Access to print in low-income and middle-income communities: An ecological study of four neighbor­ hoods. Reading Research Quarterly, 36, 8-26. doi:10.1598/RRQ.36.1.1 Pachter, L. M., Auinger, P., Palmer, R., & Weitzman, M. (2006). Do parenting and the home environment, maternal depression, neighbor­ hood, and chronic poverty affect child behavioral problems differently in different racial-ethnic groups? Pediatrics, 117, 1329-1338. doi:10.1542/ peds.2005-1784 Petterson, S. M., & Albers, A. B. (2001). Effects of poverty and maternal depression on early child development. Child Development, 72, 17941813. doi: 10. I l l 1/1467-8624.00379 Preacher, K. J., Zyphur, M. J., & Zhang, Z. (2010). A general multilevel SEM framework for assessing multilevel mediation. Psychological Methods, 15, 209-233. doi:10.1037/a0020141 Radloff, L. S. (1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385-401. doi:10.1177/014662167700100306 Raver, C. C., & Zigler, E. F. (1997). Social competence: An untapped dimension in evaluating head start’s success. Early Childhood Research Quarterly, 12, 363-385. doi:10.1016/S0885-2006(97)90017-X Rimm-Kaufman, S. E., Curby, T. W., Grimm, K. J., Brock, L. L., & Nathanson, L. (2009). The contribution of children’s self-regulation and classroom quality to children’s adaptive behaviors in the kindergarten classroom. Developmental Psychology, 45, 958-972. doi:10.1037/ a0015861 Sameroff, A., Seifer, R., Zax, M., & Barocas, R. (1987). Early indicators of developmental risk: Rochester Longitudinal Study. Schizophrenia Bulletin, 13, 383-394. doi:10.1093/schbul/13.3.383 Sampson, R. J., Morenoff, J. D., & Gannon-Rowley, T. (2002). Assessing “neighborhood effects” : Social processes and new directions in research. Annual Review o f Sociology, 28, 443-478. doi:10.1146/annurev.soc.28 .110601.141114 Schweinhart, L. J., Monde, J., Xiang, Z., Barnett, W. S., Belfield, C. R., & Nores, M. (2005). Lifetime effects: The High/Scope Perry Preschool study through age 40. Ypsilanti, MI: High/Scope Press. Squires, J., Bricker, D., & Twombly, L. (2002). The ASQ.SE user’s guide fo r the Ages & Stages Questionnaires: Social-Emotional. Baltimore, MD: Brookes. Stapleton, L. M. (2002). The incorporation of sample weights into multi­ level structural equation models. Structural Equation Modeling, 9, 475502. doi: 10.1207/S15328007SEM0904_2

727

Stewart, S. D. (2010). Children with nonresident parents: Living arrange­ ments, visitation, and child support. Journal o f Marriage and Family, 72, 1078-1091. doi: 10.11 11/j. 1741-3737.2010.00751.X Vaden-Kieman, M., D’Elio, M., O ’Brien, R., Tarullo, L., Zill, N., & Hubbell-McKey, R. (2010).Neighborhoods as a developmental context: A multilevel analysis of neighborhood effects on Head Start families and children. American Journal o f Community Psychology, 45, 49-67. doi:

10.1007/s10464-009-9279-z Walker, S. P., Wachs, T. D., Grantham-McGregor, S., Black, M. M., Nelson, C. A., Huffman, S. L . , . . . Richter, L. (2011). Inequality in early childhood: Risk and protective factors for early child development. The Lancet, 378, 1325-1338. doi: 10.1016/SO140-6736( 11 )60555-2 Welsh, J. A., Nix, R. L., Blair, C., Bierman, K. L., & Nelson, K. E. (2010). The development of cognitive skills and gains in academic school readiness for children from low-income families. Journal o f Educational Psychology, 102, 43-53. doi:10.1037/a0016738 Wheaton, B., & Clarke, P. (2003). Space meets time: Integrating temporal and contextual influences on mental health in early adulthood. American Sociological Review, 68, 680-706. doi:10.2307/1519758 Whittaker, J. E. V., Harden, B. J., See, H. M., Meisch, A. D., & Westbrook, T. P. R. (2011). Family risks and protective factors: Pathways to Early Head Start toddlers’ social-emotional functioning. Early Childhood Research Quarterly, 26, 74-86. doi:10.1016/j.ecresq.2010.04.007 Woodcock, R. W., McGraw, K. S., & Mather, N. (2001). WoodcockJohnson III Tests o f Achievement. Itasca, IL: Riverside. Xue, Y., Leventhal, T., Brooks-Gunn, J., & Earls, F. J. (2005). Neighbor­ hood residence and mental health problems of 5- to 11-year-olds. Ar­ chives o f General Psychiatry, 62, 554-563. doi:10.1001/archpsyc.62.5 .554 Yeung, W. J., Linver, M. R., & Brooks-Gunn, J. (2002). How money matters for young children’s development: Parental investment and family processes. Child Development, 73, 1861-1879. doi:10.1111/ 1467-8624.t01-l-00511

Received January 9, 2014 Revision received July 11, 2014 Accepted July 27, 2014 ■

Copyright of Journal of Family Psychology is the property of American Psychological Association and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.