The Impact of Neglect on Initial Adaptation to School

Abstract

This study examined the impact of child neglect during the first 4 years of life on adaptation to school during kindergarten and first grade in the context of neighborhood poverty (NP). Processes related to the development of school competencies were examined, including the mediational role of cognitive functioning and ego-resiliency (ER) in shaping children’s school outcomes. A total of 170 low-income urban children were followed prospectively for 2 years (ages 4–6). Results indicated that neglected children had significantly lower scores on kindergarten classroom behavior and first-grade academic performance than nonneglected children. Children’s cognitive performance at age 4, controlling for maternal intelligence quotient, mediated the relation between severity of neglect and children’s behavior in kindergarten as well as their academic performance in first grade. Moreover, severity of neglect was related to children’s ER at age 4. However, additional ecological adversity in the form of NP moderated the link between ER and classroom behavior, such that at lower levels of poverty, ER mediated the relation between severity of neglect and school adaptation. Conversely, when NP was extreme, the effects of ER were attenuated and ER ceased to predict behavioral performance in kindergarten. The implications of these findings for prevention and intervention are discussed.

Keywords

child neglect school adaptation ego-resiliency cognition neighborhood poverty developmental psychopathology

Adaptation to school is a central task of development for children, beginning in early childhood and extending through adolescence. Successful school adaptation requires a variety of competencies in children (Eccles, Lord, & Roeser, 1996; Pianta, 1999). At the most basic level, children need to demonstrate cognitive and linguistic competence in order to perform well academically in the classroom. They also need to exhibit acceptable levels of motivation and behavioral engagement in school activities. However, successful adaptation to the demands of the school environment also requires a number of nonacademic competencies. For example, the ability to regulate and manage one’s behavior is critical, as is the ability to adapt flexibly to new situations and to solve problems creatively when they arise. In addition, it is essential that children be able to assimilate into a new social environment characterized by extrafamilial adults and peers. Unfortunately, maltreated children are at risk for poor resolution of this developmental task, resulting in part from prior developmental failures. Child maltreatment has been associated with a wide range of negative sequelae, and poor academic outcomes and problematic school performance are among the documented consequences these children may face (see Cicchetti & Valentino, 2006, for a review).

Although maltreated children in general perform more poorly in school than nonmaltreated children, in terms of both motivational orientation and actual achievement, neglected children in particular appear to be at the highest risk of school maladaptation (Cicchetti & Valentino, 2006). Evidence of the increased risk for school difficulties can be seen even during the preschool years. For example, maltreated infants and toddlers are more likely to exhibit developmental delays by preschool, and children who experienced neglect demonstrate particularly high rates of developmental delay (Scarborough, Lloyd, & Barth, 2009; Scarborough & McCrae, 2010). In addition, neglected preschool-aged children have significantly lower scores on verbal comprehension and expression than other groups of children, including those who have been physically abused (Allen & Oliver, 1982; Culp, Watkins, & Lawrence, 1991). There is also evidence of lower intelligence quotient (IQ) scores among neglected children, even after controlling for the effects of poverty (Administration on Children and Families, 2005; Fishbein et al., 2009; Gowan, 1993; Perez & Widom, 1994; Polansky, Chalmers, Buttenweiser, & Williams, 1981).

Once neglected children enter school, signs of academic difficulty quickly emerge. In the longitudinal study of Egeland and his colleagues, neglected children were rated by their kindergarten teachers as having more difficulty comprehending day-to-day school work than children in the adequately reared group, and they had more difficulty handling the classroom environment (Erickson, Egeland, & Pianta, 1989). Neglected children also demonstrated poor work habits, and they had difficulty working independently. Academic difficulties increased during the early school years for these children, and by second grade all of the neglected children had been referred for special education services (Egeland, 1991). The pattern of poor academic performance among these neglected children continued through adolescence (Egeland, 1997). One limitation of this research, though, was the small sample size of neglected children upon which these findings were based.

Cross-sectional research by Eckenrode and his colleagues revealed school problems in a large sample of neglected children in grades Kindergarten through 12 (Eckenrode, Laird, & Doris, 1993; Kendall-Tackett & Eckenrode, 1996). Child neglect (without abuse) was more likely than physical or sexual abuse to be predictive of poor standardized test scores and lower grades in reading and math, even after accounting for the effects of gender and socioeconomic status. Child neglect, alone and in combination with other forms of maltreatment, also was predictive of more grade repetitions, more disciplinary referrals and suspensions, and lower scores in reading, math, and English compared with school performance among nonmaltreated children.

Other research has corroborated these deleterious effects of neglect. In another large cross-sectional study, Leiter and Johnsen (1994) reported that the most pervasive effects on academic performance were associated with child neglect and that these negative sequelae were over and above the effects of poverty. In a separate investigation, neglected children had more school absences and performed more poorly on academic achievement tests than nonmaltreated children (Wodarski, Kurtz, Gaudin, & Howing, 1990). Longitudinal investigations have yielded similar conclusions. For example, Perez and Widom (1994) followed maltreated children into adulthood and found that children with documented histories of neglect had lower IQ and reading scores, after socioeconomic status was controlled. Additionally, neglected and abused children completed fewer years of school, and reported more school difficulties, including repeating a grade, truancy, and being expelled or suspended, compared with nonmaltreated children (Perez & Widom, 1994). These school difficulties and lower IQ effects were particularly pronounced for females (Currie & Widom, 2010).

Although these studies shed light on the deleterious consequences associated with child neglect, questions still remain. For example, how does the severity of children’s neglect experiences impact their functioning? Severity of physical neglect has been shown to predict internalizing symptoms during the school-aged years (Manly, Kim, Rogosch, & Cicchetti, 2001). Does the severity of child neglect alter how children navigate the demands of school? Moreover, studies linking child neglect to school performance have revealed little about the processes through which neglect exerts its influence. An ecological transactional approach to understanding the effects of child neglect suggests both that (1) the effects of child neglect on subsequent functioning may be mediated by its effects on prior development and (2) adversity in other parts of the ecology may moderate the developmental cascades that link neglect to later school performance (Cicchetti & Valentino, 2006; Masten & Cicchetti, 2010). Understanding the processes by which exposure to early child neglect adversely influences children’s adaptation to school is critically important.

Neglect and School Performance: Mediating Processes

Child neglect adversely affects the resolution of important tasks of early development that may ultimately impact successful school entry and subsequent adaptation during kindergarten and first grade. One important developmental domain that is being consolidated during early childhood is cognitive performance. This domain is particularly important because cognitive abilities are strongly predictive of school achievement (Duncan et al., 2007). For example, the education literature provides ample evidence that indices of preschool intellectual ability (e.g., verbal skills and IQ) predict academic outcomes in early elementary school (see Tramontana, Hooper, & Selzer, 1988, for a review). Research on child neglect demonstrates that neglected children score lower on these indices than comparison children (Allen & Oliver, 1982; Gowan, 1993). Neglecting home environments may be characterized by parent–child interactions that undermine neglected children’s intellectual skill development (Cicchetti & Valentino, 2006; Lynch & Cicchetti, 1992; Toth & Cicchetti, 1996). Conversely, home environments characterized by parental involvement, openness, and support have been shown to predict higher scores on receptive vocabulary among African American kindergarteners (Luster, Reischl, Gassaway, & Gomaa, 1995). As a result, children coming from neglecting homes may be less ready for the demands of school because they have not been supported in the mastery of language and cognitive skills. Thus, neglect may have a negative impact on children’s school adaptation through the undermining of their cognitive development.

However, cognitive delays are not the sole obstacle to neglected children’s readiness for school. In addition to cognitive ability, Haskett, Nears, Ward, & McPherson (2006) identified that ego-resiliency (ER) and self-regulation can play a protective role in resilient development of maltreated children. The construct of ER involves flexible responsivity and positive adaptability in adjusting to situational demands with positive affect and self-confidence (Block & Block, 1980; Letzring, Block & Funder, 2005). In low-risk samples, ER has been shown to mediate the effects of parenting on academic achievement (Swanson, Valiente, Lemery-Chalfant, & O’Brien, 2011). Shonk and Cicchetti (2001) found that ER mediated relations between maltreatment and school functioning. Notably, a majority of the maltreated children in their sample had experienced neglect. When process models of the effects of maltreatment were examined, the effect of maltreatment on maladjustment in the classroom was fully mediated by social competencies and ER. Thus, maltreatment resulted in academic maladjustment through its impact on personal resources, such as children’s ability to adapt flexibly to the demands of school. It is likely that neglect played a significant role in this process, but the specific effects of neglect and ER were not examined. Neglect is expected to undermine ER, which would subsequently have a negative impact on children’s school adaptation.

Poverty as a Moderator of the Effects of Neglect on School Performance

Neighborhood poverty (NP) itself is a risk factor that can undermine healthy development, and the characteristics of impoverished communities—such as high rates of public assistance, low educational attainment, and low socioeconomic status—increase the risk of school failure and drop out (Ainsworth, 2002; Murry, Berkel, Gaylord-Harden, Copeland-Linder, & Nation, 2011; Nikulina, Widom, & Czaja, 2011). High-poverty neighborhoods create an environmental context that undermines successful development through increased stressors such as crime, limited housing, crowding, low-social capital, toxins, less access to resources, and a confluence of risk factors across multiple settings, including school (Blair & Raver, 2012; Evans & Kim, 2013). As a result, neglected children who live in impoverished neighborhoods may experience heightened risk of maladaptive outcomes (Bright & Jonson-Reid, 2008), and NP may interact with neglect in predicting academic achievement (Nikulina et al., 2011). An initial study assessing both neglect and poverty suggested that neglect overshadowed the effects of NP in predicting academic achievement, and the impact of poverty on academic success was only significant among nonmaltreated children (Nikulina et al., 2011). However, poverty has also been linked to self-regulation and other self-system processes (Evans & Rosenbaum, 2008). Since we are proposing that cognitive performance and ER may mediate the effects of child neglect, it is possible that there may be complex interactions involved in the processes that link child neglect and children’s functioning in school in the ecological context of high-poverty neighborhoods.

Summary

Given that neglect has received less attention in maltreatment research than other subtypes (Hildyard & Wolfe, 2002), yet neglect may have a detrimental impact on school performance, the central goal of the current study was to examine processes linking child neglect to poor initial adaptation to school. The available evidence provides empirical support for different elements of the mediational models that we propose to examine. First, child neglect has been linked to a number of indicators of poor adaptation to school. In addition, neglected children are likely to have problems with early development, including measurable deficits in their cognitive performance and ER.

Based on this evidence, the current investigation examined processes linking child neglect to successful school entry. Initially, we examined differences between neglected and nonneglect children. Subsequently, we examined the impact of severity of neglect in the context of NP. Specifically, we focused on the effects of neglect on children’s adaptation to school, along with hypothesized mediators of these effects, including cognitive performance and ER. The link between severity of early neglect and later school entry (classroom behaviors and academic performance) may unfold through children’s cognitive performance and ego-resilient functioning in early childhood. Moreover, from a developmental psychopathology framework in accord with an understanding that neglect occurs in a larger ecological context, the role of NP in shaping the pathways through which neglect impacts children’s adaptation to school was examined. Specifically, NP was examined as a moderator to the underlining mechanism between neglect and school performance. This study extends and expands upon existing research through a comprehensive assessment with multiple reporters, across domains, and multilevel analyses of the ecology of neglect and its longitudinal effects on school adaptation.

Research Questions and Hypotheses

Research Question 1: Extant literature suggests that neglected children have more difficulty in school adaptation, but it is not clear whether these school difficulties are evident at the time of school entry. Thus, the initial research question concerns whether neglected children will differ from nonneglected children in early school years. It is expected that neglected children will have significantly more problematic behaviors in kindergarten and lower grades in first grade than nonneglected children.

Research Question 2: Research on the ontogenic cognitive development of neglected children in the context of exosystem poverty suggests that cognitive delays may be a mechanism through which more severe neglect may undermine academic performance and that this mediation may be impacted by NP. It is expected that children’s cognitive functioning at age 4 will mediate the relation between early neglect and subsequent classroom behavior in kindergarten and academic performance in first grade. This mediation link will be moderated by NP such that in high-poverty contexts, poverty will exacerbate the effects of neglect on cognitive delays and will weaken the effect of cognitive abilities on school performance.

Research Question 3: Because school adaptation is also impacted by ER, the processes linking severity of neglect to school adaptation via ER were examined in the context of NP. It is expected that children’s ER at age 4 will mediate the relation between early neglect and subsequent school adaptation. The mediation between neglect and classroom behavior via ER will be moderated by poverty, such that the path from ER to kindergarten behavioral performance will be undermined by higher levels of poverty. Poverty will exacerbate the effects of neglect on ER and will weaken the effect of ER on school performance.

Method

Participants

Participants in this investigation included 170 high-risk urban children and their caregivers from upstate New York. All children were living with their biological mothers at the time of enrollment. Eighty-two percent of these families were unmarried. Fifty-six percent of the sample had less than a high school education. The sample was ethnically diverse, represented with African American, Latino, Caucasian, and biracial families (88% minority in the total sample; see Table 1). Children were recruited as preschoolers at the age of 4 (N = 170) and were followed at ages 5 and 6. Over the three phases of the study, retention rates were 92%. Across the time points of the study, 170 families were assessed at age 4; 160 families were assessed at age 5; and 156 families participated at age 6.

Table 1.

Demographic Characteristics.

Descriptors	Maltreated (%; N = 111)	Nonmaltreated (%; N = 59)	Test statistics, χ^2a
Gender of child (female)	52.3	52.5	.971(1) ns
Minority status (non-White)	89.2	84.7	.402(1) ns
Mother’s marital status (single)	85.6	78.0	.209(1) ns
History of public assistance	95.5	91.5	.209(1) ns
Completed high school	40.5	51.7	.165(1) ns

^aPearson chi-square.

Recruitment of maltreated children (N = 111) focused on children who had been identified through the Department of Human Services (DHS) as having documented histories of physical neglect. A DHS staff member who was assisting with the project identified eligible children and made the initial contact with families to ascertain interest in participation and to obtain permission to share their contact information with the project team.

To obtain a demographically matched comparison group, nonmaltreating families (N = 59) were selected randomly from the county recipients of Temporary Assistance for Needy Families (TANF). Because the majority of neglecting families referred to Child Protective Services (CPS) are socioeconomically disadvantaged and receiving public assistance (95.5% in the current study; Gaudin, 1999; Sedlak et al., 2010), utilization of TANF lists provided access to a demographically similar population. In the initial contact, consents were obtained to verify nonmaltreatment status by reviewing DHS central registry data. Medical records also were reviewed to confirm nonmaltreatment status.

Procedure

At the initial assessment, mothers signed informed consent and permission for their children’s participation, according to procedures approved by the university institutional review board. Participating families completed initial assessments of children’s functioning and the family environment, including demographic information on socioeconomic status, children’s race, and ethnicity. Families participated in a laboratory visit at children’s age of 4 that included a videotaped frustration task and assessment of cognitive performance. Parents signed consent for their children’s teachers to be contacted at the end of kindergarten and first grade to complete questionnaires regarding children’s behavior in the classroom, and school records were obtained.

Measures

Child Neglect and Abuse Histories

Permission to review DHS records was obtained from all participating families. Histories of maltreatment were coded using the Barnett, Manly, and Cicchetti (1993) Maltreatment Classification System, which captures detailed information from CPS records by obtaining systematic data from the narrative and investigation determinations contained in DHS records, rather than relying on CPS labels (Manly, 2005). This system captures information on the occurrence of multiple subtypes of maltreatment, including extensive information on severity, and the developmental period during which it occurred. For each report, presence and absence of each subtype is determined based on operational definitions of each category, and severity codes are determined within each subtype.

Neglect is coded when there is evidence that a caregiver has failed to exercise a minimum degree of care in meeting the child’s needs or failed to take adequate precautions to ensure the child’s safety. Neglect can range from frequently missed meals, unsanitary living conditions, and failure to provide adequate supervision to severe malnutrition, gross inattention to medical needs, or endangering the child in life-threatening situations. Neglect severity was scored on a 1–5 scale, with 5 indicative of events that were life threatening or likely to result in serious physical consequences and 1 indicative of neglect that was relatively mild but still rose to the level of attention by authorities. Adequate reliability for coding of maltreatment subtypes and severity was obtained with intraclass correlations ranging from 0.81 to 1.0 across subtypes.

Before the age of 4, 111 children were from maltreating families and 59 had no documented history of maltreatment. Within the maltreated group, 97 of the 111 children (87%) had reports of physical neglect, including reports of lack of supervision, failure to provide, and moral/legal maltreatment (Barnett, Manly, & Cicchetti, 1993). Neglect severity ranged from 1 to 5, with 70% of the neglected sample scoring in the moderately high range (3–4 on the 1–5 scale). Multiple subtype occurrence was present for 51% of the sample. Although the sample was originally recruited with a focus on physical neglect, other subtypes were also present before age 4; 42 children (38%) had reports of neglect with no other subtypes present and 69 children (62%) had combinations of other subtypes (13 children [12%] had neglect with physical abuse and emotional maltreatment, 39 children [35%] had neglect with emotional maltreatment, 3 children [3%] had neglect with physical abuse, 2 children [2%] had emotional maltreatment and physical abuse, 7 [6%] children had emotional maltreatment alone, and no children were sexually abused). The latter nine children had neglect severity equal to zero, and therefore were considered nonneglected in analyses of neglect, along with five children who had maltreatment in their families but no reports of maltreatment for the individual children. Subtype and severity were coded separately, and children had a range of severity ratings, regardless of whether they had a single subtype or multiple subtypes. Neglect severity ratings were higher than severity ratings of other maltreatment subtypes for the majority of multiple maltreated children. Of the children who experienced both physical abuse and physical neglect, 88% (14/16) had severity ratings of physical neglect that were greater than or equal to the severity of physical abuse. Of the children with physical neglect and emotional maltreatment, 86% (44/52) had higher neglect severity. Thus, neglect was the focal maltreatment subtype in analyses because of the frequency with which neglect occurred and because of these elevated neglect severity ratings relative to other subtypes of maltreatment. Children who were nonmaltreated and those who were maltreated but did not have neglect in their CPS records were classified as nonneglected by assigning a neglect severity score of zero for analyses of neglect severity. By age 6, an additional six children had experienced maltreatment. Because this investigation was designed to examine developmental consequences of early neglect prior to age 4, these children were considered nonmaltreated for analyses.

Demographics

The Demographics Interview (Cicchetti & Carlson, 1979) was utilized to obtain parent reports of information about race/ethnicity and socioeconomic status, including receipt of public assistance and education.

Maternal IQ

The Wechsler Abbreviated Scale of Intelligence (WASI; Wechsler, 1999) is a well-validated and standardized brief test of intelligence (Wechsler, 1999). Four subtests (vocabulary, similarities, block design, and matrix reasoning) that correlate .92 with the Wechsler Adult Intelligence Scale-III Full Scale IQ were collected with mothers. Mothers’ WASI full scale IQ was utilized to control for maternal intelligence on children’s cognitive performance.

NP

Poverty in the exosystem was assessed through the use of Census Tract Data (U.S. Census, 2000). Government census tracts provide units of analysis for which numerous types of data are readily available. We focused on data pertaining to economic hardship within census tracts or blocks. These data gave us approximations of NP that provide a backdrop for children’s more immediate experiences. These factors may moderate the effects of children’s adaptation as a result of the added stress they create. Although census tracts do not necessarily represent the neighborhood as it would be defined by the residents themselves (Coulton, Korbin, Su, & Chow, 1995), census tract data have been used successfully in a number of studies examining the functioning of high-risk samples (Aber, Jones, Brown, Chaudry, & Samples, 1998; Korbin, Coulton, Chard, Platt-Houston, & Su, 1998). A composite of NP was created following the work of Duncan and Aber (1997) using census variables including rates of high school graduates, home ownership, female-headed households with incomes below the poverty level, and per capita income.

Child Functioning

Cognitive Performance

Several measures of intellectual ability were utilized. The Wechsler Preschool and Primary Scale of Intelligence–Revised (WPPSI-R) is a well-validated and standardized test of intelligence (Wechsler, 1989). At the age 4 initial assessment, children completed four subtests that correlate .93 with the full-scale score: information, geometric design, block design, and similarities. The Peabody Picture Vocabulary Test (PPVT-III) and the Expressive Vocabulary test (EVT) were administered as measures of verbal intelligence. The PPVT-III measures children’s receptive vocabulary and has been standardized on a representative national sample. It correlates .90 with full scale IQ on the Wechsler Intelligence Scale for Children (Williams & Wang, 1997). The EVT measures expressive language abilities, and likewise has been standardized on a national sample. Correlations between the EVT and full scale IQ range from .66 to .84 over a range of tests of cognitive maturity (Williams, 1997). In the current study, these three factors were evaluated and estimated in a factor analysis and were shown to form a valid and reliable latent cognitive performance construct.

ER

The Barrier Box Task assessed the child’s reaction to a minor frustration (Erickson et al., 1989). Children were in a room with a research assistant who was working at a nearby table. In the center of the room was a large plexiglass box containing a variety of attractive toys, and the box was latched in such a way that a young child cannot open it. A few unattractive toys were left on the floor outside the box. For 10 minutes, the child was allowed to try to open the box, play with the available toys, or wander around the room. At the conclusion of the session, the assistant opened the box and allowed the child to play with the toys for several minutes. The manner in which the child responded to the initial frustration of impeded access to the attractive toys revealed the level of the child’s self-confidence and ability to act autonomously. The session was videotaped, and observers unfamiliar with the research hypotheses rated the following variables on 3- to 7-point scales: flexibility, positive affect, self-esteem, and creativity. These dimensions were selected because in defining ER, Block and his colleagues have stated that “as a result of this adaptive flexibility, individuals with a high level of resiliency are more likely to experience positive affect, and have higher levels of self-confidence and better psychological adjustment than individuals with a low level of resiliency” (Letzring et al., 2005, p. 398). Thus, flexibility, positive affect, self-esteem, and creativity reflect dimensions of ER, as well as behavioral engagement. This task has distinguished neglected and adequately reared children (Erickson et al., 1989).

Each videotape was viewed by two separate teams of trained observers and coded independently. The average correlation between pairs of coding teams for each Barrier Box dimension was as follows: flexibility, r = .57; positive affect, r = .68; self-esteem, r = .67; and creativity, r = .74. After independent ratings were obtained for each assessment, coding teams met to resolve any discrepancies in their scores and arrived at a consensus score for each Barrier Box dimension across the entire sample.

Classroom Behaviors

Teacher Ratings of Kindergarten Behavior

The Teacher–Child Rating Scale 2.1 (TCRS, Perkins & Hightower, 2002) was utilized by teachers to rate 32 items on children’s social, behavioral, and academic competencies in school. Four positive and four negative items load on the following primary scales: task orientation, behavior control, and peer social skills. Internal consistency ranges from .87 to .94, and test–retest reliability ranges from .66 to .94 across a number of samples (Perkins & Hightower, 2002). The internal consistency of these three scales in the current sample was very good (Cronbach’s α = .87 for task orientation, .86 for behavioral control, and .92 for peer social skills).

Report Card Effort Data

Quantitative indicators of children’s performance in school were obtained from school records for all participating children at the conclusion of kindergarten. Language arts and math effort scores from children’s report cards indicated behavioral engagement in kindergarten academic tasks. These scores were on a 1–4 scale, with higher scores indicative of greater effort and engagement.

Academic Performance

School Record Data in First Grade

Final grades in Language Arts and Math from report cards were obtained from school records for all participating children at the conclusion of first grade. Because some schools had variations in reporting of grades, all grade data were converted to a 1–4 scale with higher scores indicative of higher grades.

Analytic Plan

The hypothesized models were tested using structural equation modeling (SEM; M-plus Version 6.10; Muthen & Muthen, 1998–2010). Before modeling, the data were first inspected for missing data and outliers, including normality issues. Across the measures, the average percentage of missing data was 6% ranging from zero missing data on six measures and peaking at 28.8% on the kindergarten report card data (see Table 2 for sample sizes per variable and time of measurement). No missing data were present on maltreatment status. Missing data patterns over time (attrition) were analyzed at each wave with regard to the examined variables and were determined to be Missing at Random (MAR; Schafer & Graham, 2002). Thus, data missingness was analyzed under missing data theory using all available data via the full information maximum likelihood (FIML) estimation technique (Schafer & Graham, 2002).

Table 2.

Means, Standard Deviations, Sample Size, and Ranges of Model Variables.

Variables		M	SD	N	Minimum	Maximum
Gender		0.48	0.50	170	0	1
Minority		1.12	0.33	170	1	2
Neglect	Neg Pres	0.57	0.50	170	0	1
Neglect	Neg Sev	1.91	1.83	170	0	5
Cognitive performance	PPV	86.56	14.92	167	42	127
	Full IQ	82.19	12.71	166	52	117
	EVT	89.60	13.42	168	53	127
Classroom behaviors	Task Or	26.48	8.69	157	8	40
	Bev C	27.28	7.80	157	9	40
	Peer S	29.46	8.30	157	8	40
	K Lng	3.09	0.79	122	1	4
	K Math	3.19	0.74	121	1	4
Academic performance	Lang	2.91	0.87	130	1	4
Academic performance	Math	2.89	0.91	130	1	4
Mother IQ	WASI	84.27	13.34	147	55	118
Ego-resiliency	Flex	2.73	1.21	168	1	5
	Post af	3.21	0.94	168	1	5
	Self-Est	3.88	1.55	168	1	7
	Creat	2.24	1.20	168	1	5
Poverty	CToou	37.81	19.04	170	.8	96.1
	CTfh	48.37	16.89	170	7.4	81.7
	PerCap	12,838.90	5,149.98	170	7,317	28,627
	CThs	64.59	12.99	170	42.7	94.9

Note. Neg Pres = Neglect present/absent (age 4); Neg Sev = Severity physical neglect (age 4); Barrier Box: Flex = flexibility, Creat = creativity, Self-Est = self-esteem, Post af = positive affect; census tract: CToou = #owner-occupied units, CTfh = % families w/female householder no husband w/related kids under 18 below poverty level, PerCap = per capita income, CThs = high school graduate; kindergarten: Task Or = Teacher–Child Rating scale (TCRS) task orientation, K Lng = language arts effort mean, K Math = math effort mean, Bev C = TCRS behavior control, Peer S = TCRS peer social skills; first grade: Lang = language arts grades, Math = math grades; Wechsler Abbreviated Scale of Intelligence (WASI) full scale IQ for biomother; Age 4: PPV = Peabody Picture Vocabulary test (PPVT) standard score, Full IQ = Wechsler Preschool and Primary Scale of Intelligence (WPPSI) full scale IQ, Expressive Vocabulary test (EVT) = EVT standard score.

To account for nonnormality present in the various variables, all of the model-based analyses were conducted using maximum likelihood estimation with robust standard errors (Yuan & Bentler, 2000). Subsequently, model fit indices were first determined for the measurement models before pursuing evaluation and interpretation of the structural models. Model fit refers to the ability of a proposed model to reproduce the observed data within a variance–covariance matrix. A good-fitting model is one that is reasonably consistent with the data and therefore does not necessitate respecification. For SEM models, a variety of global fit indices were used, including traditional model fit indices conforming to the following statistical criteria (McDonald & Ho, 2002) for the root mean square error of approximation (RMSEA < 0.08), for the test of close fit (p > .05), the comparative fit index (CFI > 0.95), Tucker–Lewis Index (TLI > 0.94), and the standardized root mean square residual (SRMR < 0.07). With respect to data nonnormality, traditional maximum likelihood methods assume the distributions of the continuous variables in the model are multivariate normal, an assumption that is often violated in mediation analyses (Shrout & Bolger, 2002). In order to remedy nonnormality issues, the product coefficient method, using a bootstrapping procedure with 5,000 sample replicates (MacKinnon, Fairchild, & Fritz, 2007), was used to test mediation paths (Preacher & Hayes, 2008; Shrout & Bolger, 2002). Moderated mediation was tested following model 3 described in Preacher, Rucker, and Hayes (2007).

Results

Table 3 summarizes Pearson and Spearman (for binary measures) bivariate correlations among variables included in the structural models. Table 2 summarizes means, standard deviations, sample size, and minimum and maximum values. At the bivariate level, presence of neglect was significantly associated with lower cognitive functioning, reduced flexibility and creativity dimensions of ER, decreased language arts effort in kindergarten, and lower language arts and math grades in first grade. Neglected children also were more likely to be living in neighborhoods with lower high school graduation rates and lower per capita income.

Table 3.

Correlations Among Model Variables.

Variables		1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22
1	Neg Pres
2	Neg Sev	.91**
3	Gender	.02	−.01
4	Minority	−.07	−.03	−.11
5	Flex	−.16*	−.15*	−.03	.01
6	Creat	−.17*	−.22**	−.16*	.08	.62**
7	Self-Est	−.10	−.14	.00	.04	.74**	.56**
8	Post af	−.01	−.07	.01	.20**	.43**	.45**	.58**
9	CToou	−.09	−.11	−.09	.35**	−.02	.05	.06	.08
10	CTfh	.10	.08	.06	−.35**	−.06	−.08	−.14	−.12	−.66**
11	PerCap	−.19*	−.18*	−.11	.46**	.04	.09	.08	.13	.66**	−.73**
12	CThs	−.18*	−.16*	−.11	.38**	.07	.09	.09	.12	.72**	−.74**	.87**
13	Task Or	−.14	−.09	−.18*	.04	.09	.10	.04	.03	−.01	−.06	.08	.07
14	K Lng	−.22*	−.20*	−.16	−.06	.18	.10	.10	.00	.17	−.16	.10	.20*	.58**
15	K Math	−.12	−.11	−.08	.12	.14	.06	.09	.04	.21*	−.15	.16	.26**	.51**	.82**
16	Bev C	−.07	−.09	−.19*	−.01	.02	.06	.03	.03	.08	−.01	.11	.09	.63**	.51**	.44**
17	Peer S	−.17*	−.17*	−.07	.00	.08	.05	.00	−.02	.01	−.04	.01	−.01	.68**	.45**	.32**	.63**
18	Lang	−.16	−.09	−.22**	.14	.11	.13	.00	.03	.15	−.09	.17*	.18*	.46**	.39**	.43**	.26**	.30**
19	Math	−.19*	−.14	−.11	.16	.10	.07	−.02	.06	.09	−.09	.11	.16	.41**	.37**	.38**	.26**	.25**	.81**
20	WASI	−.09	−.09	−.03	.36**	.17*	.23**	.05	.16	.25**	−.23**	.37**	.38**	.12	.08	.15	.06	.06	.37**	.27**
21	PPV	−.16*	−.12	.06	.30**	.03	.06	.03	.07	.15	−.18*	.21**	.22**	.29**	.08	.23*	.16*	.16*	.36**	.36**	.34**
22	Full IQ	−.19*	−.17*	−.07	.32**	.13	.18*	.14	.10	.22**	−.22**	.27**	.30**	.30**	.06	.11	.15	.20*	.42**	.44**	.49**	.65**
23	EVT	−.21**	−.20*	−.04	.31**	.07	.13	.06	.21**	.17*	−.18*	.27**	.29**	.24**	.03	.14	.07	.09	.49**	.43**	.42**	.69**	.67**

Note. Neg Pres = neglect present/absent (age 4; 1 = present = absent); Neg Sev = severity physical neglect (age 4); barrier box: Flex = flexibility, Creat = creativity, Self-Est = self-esteem, Post af = positive affect; census tract: CToou = #owner-occupied units, CTfh = % families w/female householder no husband w/related kids under 18 below poverty level, PerCap = per capita income, CThs = high school graduate; kindergarten: Task Or = Teacher–Child Rating scale (TCRS) task orientation, K Lng = language arts effort mean, K Math = math effort mean, Bev C = TCRS behavior control, Peer S = TCRS peer social skills; first grade: Lang = language arts grades, Math = math grades; Wechsler Abbreviated Scale of Intelligence (WASI) full scale IQ for biomother; age 4: PPV = Peabody Picture Vocabulary test (PPVT) standard score, Full IQ = Wechsler Preschool and Primary Scale of Intelligence (WPPSI) full scale IQ, EVT = Expressive Vocabulary test (EVT) standard score.

*p<.05, **p<.01

Presence/Absence of Neglect and School Outcomes

To examine differences between neglected and nonneglected children that would provide comparability with prior research on neglect, we began with a structural model that tested mean differences in kindergarten classroom behavior and first-grade academic performance by presence or absence of neglect. The structural model fit indices indicated good model fit (CFI = 0.958; TLI = 0.932; RMSEA = 0.078; close fit test p > .072; SRMR = 0.061). Results showed that neglected children had significantly lower factor scores on both outcomes, kindergarten classroom behavior (β = −.367; p < .05), and first-grade academic performance (β = −.239; p < .05).

Test of Measurement for Model 1

In order to obtain a common metric for the examined latent variables, their variance was fixed to 1.00 (Brown, 2006). In the first model (Figure 1), latent factors were constructed for cognitive performance and classroom behavior. The cognitive performance factor was measured by three indicators: PPVT-III, WPPSI-R IQ, and EVT scores at the child’s age of 4. The classroom behavior factor was composed of Kindergarten Report Card Effort (in language arts and math) and TCRS teacher rating measures from which three subscales were used (i.e., a total of five indicators.) The three TCRS teacher classroom rating subscales were (a) task orientation, (b) behavior control, and (c) peer social skills. The first-grade academic performance factor was derived from report card grades in language arts and math. This method of deriving latent factors allowed the model to capture common variance of cognitive performance and classroom behavior from different assessment methods and different measurement subscales and to avoid averaging subscales means that are less reliable (Yang, Nay, & Hole, 2010). The NP latent factor was created using the following four indices (see Figures 1 and 2 for factor loadings): rates of high school graduates, home ownership, female-headed households with incomes below the poverty level, and per capita income. The indices were rescaled, so high scores on the factors reflected higher level of poverty.

Figure 1.

Structural equation Model 1: Severity of neglect prior to age 4, kindergarten classroom behaviors, and first-grade academic performance, mediated by cognitive performance at age 4, including neighborhood poverty as a moderator, covarying maternal IQ., maternal high school education, minority status, receipt of public assistance. Note. Cognitive performance age 4: PPV = Peabody Picture Vocabulary test (PPVT) standard score, Wechsler Preschool and Primary Scale of Intelligence–intelligence quotient (WPPSI IQ = full scale IQ, Expressive Vocabulary test [EVT] = EVT standard score; classroom behaviors kindergarten: Task Or = Teacher–Child Rating Scale [TCRS] task orientation, Bev C = TCRS behavior control, Peer S = TCRS peer social skills, K Lng = language arts effort mean, K Math = math effort mean; academic performance first grade: Lang = language arts grades, Math = math grades; Neigh. Poverty = neighborhood poverty. The structural model fit indices indicated good model fit (comparative fit index [CFI] = 0.944; Tucker–Lewis Index [TLI]= 0.922; root mean square error of approximation [RMSEA] = 0.063; close fit test p > .123; standardized root mean square residual [SRMR] = 0.065). The following variables were used as covariates to control for family-level poverty (see Table 4): Maternal high school education, minority status, receipt of public assistance. *p < .05. **p < .01.

Figure 2.

Structural equation model 2: Severity of neglect prior to age 4 on kindergarten classroom behaviors and first-grade academic performance via ego-resiliency, including neighborhood poverty by ego-resiliency interaction as a moderator covarying maternal high school education, minority status, receipt of public assistance. Note. Ego-resiliency: barrier box: Flex = flexibility, Post af = positive affect, Self-Est = self-esteem, Creat = creativity; classroom behaviors kindergarten: Task Or = Teacher–Child Rating Scale (TCRS) task orientation, Bev C = TCRS behavior control, Peer S = TCRS peer social skills, K Lng = language arts effort mean, K Math = math effort mean; poverty: census tract: CToou = #owner-occupied units, CTfh = % families w/female householder no husband w/related kids under 18 below poverty level, CThs = high school graduate, PerCap = per capita income; academic performance first grade: Lang = language arts grades, Math = math grades; The structural model fit indices were comparative fit index (CFI) = 0.941; Tucker–Lewis Index (TLI) = 0.932; root mean square error of approximation (RMSEA) = 0.050; close fit test p > .499; standardized root mean square residual (SRMR) = 0.077. The following variables were used as covariates to control for family-level poverty (see Table 5): maternal high school education, minority status, receipt of public assistance. *p < .05. **p < .01.

All traditional indices of global fit suggested good fit between the data and the model tested. Specifically, the RMSEA estimate was equal to 0.062, with the test of close fit (p = .118). Additionally, the CFI = 0. 949, the TLI= 0. 936, and the SRMR= 0. 067 reflected good model fit. The standardized factor loadings were in the average or larger range (>.4), while the measured variables’ loadings on the latent variables were all statistically significant (i.e., p < .001), signifying that latent variables were adequately measured by their indicators. After inspecting modification indices greater than 4, we allowed error covariance between measures in which we have expected common residual variance due to shared method error (i.e., kindergarten effort on language and math).

Structural Model 1—Mediation via Cognitive Performance

We then tested the link between severity of child neglect (measured prior to age 4) and child Kindergarten classroom behavior at age 5, followed by first-grade academic performance at age 6, via cognitive performance measured at age 4. Maternal IQ was included as a covariate because maternal cognitive functioning may be expected to be related to children’s cognitive functioning. Demographic variables were controlled in all of the SEM models by covarying out minority status (White vs. none), history of public assistance (yes vs. no), and maternal education (completed high school or not). The structural model fit indices indicated good model fit (CFI = 0.944; TLI = 0.922; RMSEA = 0.063; close fit test p > .123; SRMR = 0.065). Direct paths were defined from severity of child neglect as an exogenous variable to each of the three endogenous latent variables (see Figure 1). Results of the structural paths are presented in Table 4. There was a significant negative direct path from severity of neglect to the cognitive performance factor, after controlling for variance related to maternal IQ and demographic covariates, suggesting that more severe neglect prior to age 4 was associated with decreased cognitive performance at age 4. The links from cognitive performance at age 4 to kindergarten classroom behavior at age 5 and first-grade academic performance at age 6 were statistically significant. In addition, children who demonstrated more positive classroom behavior and more effort in kindergarten showed better academic performance in first grade. Thus, these findings support an indirect effect from neglect severity prior to age 4 to later classroom behavior and academic achievement via cognitive performance. In order to test the significance of the mediation path from neglect severity to classroom behavior in kindergarten and academic performance in first grade, the product coefficient method was used. Following 5,000 bootstrap replicates, analyses confirmed that the mediation paths to kindergarten and first-grade functioning were significant (α × β = −.043; confidence interval [CI] 95% [−0.107, −0.003]; α × β= −0.021; CI 95% [−0.056, −0.002], respectively). Thus, neglect severity was not significantly related to kindergarten classroom behaviors or first-grade academic performance directly, but only indirectly through cognitive performance. Moderation analyses by NP were probed for from neglect to cognitive performance and from cognitive performance to classroom behaviors and academic performance. Specifically, NP moderation of the link between child neglect and cognitive performance (B = −0.021; Z = −1.66; p = .095) as well as moderation of the link from cognitive performance to kindergarten classroom behavior at age 5 (B = −0.073; Z = −0.132; p = .835) and first-grade academic performance at age 6 (B = −0.008; Z = −1.237; p = .216) were statistically insignificant.

Table 4.

Cognitive Performance Model.

Direct effects	B (β)	SE	Estimation/SE
Cognitive performance
Classroom behaviors	0.042 (0.395)	0.013 (0.118)	3.159 (3.360)
First grade	0.021 (0.368)	0.007 (0.108)	2.853 (3.400)
Classroom behaviors
First grade	0.244 (0.390)	0.052 (0.080)	4.688 (4.872)
Neglect severity
Cognitive performance	−1.016 (−0.160)	0.488 (0.075)	−2.081* (−2.121*)
Classroom behaviors	0.005 (0.008)	0.049 (0.086)	0.094 (0.094)
First grade	0.020 (0.114)	0.027 (0.111)	1.427 (0.740)
Mothers’ WASI IQ
Cognitive performance	0.385 (0.452)	0.080 (0.094)	4.834 (4.794)
Classroom behaviors	−0.011 (−0.133)	0.009 (0.110)	−1.167 (−1.201)
First grade	0.006 (0.114)	0.006 (0.111)	1.427 (1.031)
High school education
Classroom behaviors	0.280 (0.128)	0.197 (0.087)	1.427 (0.917)
First grade	0.131 (0.097)	0.098 (0.072)	1.329 (0.917)
Minority status
Classroom behaviors	−0.289 (−0.087)	0.321 (0.097)	−0.898 (−0.897)
First grade	−0.056 (−0.028)	0.165 (0.081)	−0.339 (−0.340)
Public assistance
Classroom behaviors	−0.043 (−0.009)	0.497 (0.108)	−0.086 (−0.086)
First grade	−0.154 (−0.055)	0.175 (0.062)	0.882 (−0.880)
Indirect effects	a1 × b1	SE	95% confidence interval
Neglect severity → Cog → CB (Figure 1)	−0.043	0.026	[−0.107, −0.003]*
Neglect severity → Cog → First grade	−0.021	0.013	[−0.056, −0.002]*

Note. Cog = cognitive performance at age 4; CB = classroom behavior in kindergarten; WASI IQ = Wechsler Abbreviated Scale of Intelligence–intelligence quotient.

High school education (1 = high school graduate and 0 = less than 12th grade education); minority (White = 1 and Others = 0); public assistance (1 = with history and nonhistory = 0).

*p < .05. **p < .01.

Test of Measurement for Model 2

In the second model (Figure 2), latent factors were constructed for ER and classroom behavior. The ER factor was measured by four indicators from the Barrier Box procedure: flexibility, creativity, self-esteem, and positive affect at the child’s age of 4. The measurement strategies for the NP, classroom behavior factor, and first-grade academic performance factor at Model 1 were retained in Model 2. Covariates included maternal education, minority status, and public assistance. Accordingly, the model fit indicators for the measurement Model 2 demonstrated very good fit (RMSEA estimate = 0.046; with the test of close fit p = .603; CFI = 0.975; TLI = 0.967; SRMR = 0.067). The standardized factor loadings were large (>.5), while the measured variables’ loadings on the latent variables were all statistically significant (i.e., p < .001), signifying that latent variables were adequately measured by their indicators. After inspecting modification indices >4, error covariance between measures was allowed to be estimated when common residual variance due to shared method error was expected (i.e., kindergarten effort performance on language arts and math; TCRS task orientation with TCRS peer social skills; barrier box self-esteem; and barrier box positive affect).

Structural Model 2—Mediation via ER

The second model tested the link between severity of child neglect (measured prior to age 4) and child kindergarten classroom behavior at age 5 followed by first-grade academic performance at age 6, via ER measured at age 4. The structural model fit indices were CFI = 0.941; TLI = 0.932; RMSEA = 0.050; close fit test p > .499; SRMR = 0.077. Direct paths were defined from severity of child neglect as an exogenous variable to each of the three endogenous latent variables (see Figure 2). Results of the structural paths are presented in Table 5. There was a significant negative direct path from severity of neglect to the ER factor, suggesting that higher levels of severity of neglect prior to age 4 predicted lower factor scores on ER at age 4. There were no significant direct paths from ER at age 4 to classroom behavior at kindergarten (age 5) or academic performance in first grade. This lack of association between ER and later academic outcomes ruled out the possibility that ER served as a linear mediator in the link from severity of neglect to child Kindergarten and first-grade functioning. Similar to Model 1, the significant path from kindergarten to first-grade academic performance suggests stability in this longitudinal link.

Table 5.

Ego-Resiliency Model.

Direct effects	B	SE	Estimation/SE
Classroom behaviors
First grade	0.306	0.058	5.271**
Ego-resiliency
Classroom behaviors	0.069	0.097	0.713
Neighborhood poverty × ego-resiliency
Classroom behaviors	0.269	0.083	3.255**
Neglect severity
Ego-resiliency	−0.109	0.046	−2.385*
Classroom behaviors	0.026	0.080	0.325
First grade	−0.003	0.029	−0.098
High school education
Classroom behaviors	0.316	0.182	1.733
First grade	0.209	0.111	1.888
Minority status
Classroom behaviors	−0.246	0.351	−0.699
First grade	0.296	0.150	1.973*
Public assistance
Classroom behaviors	−0.020	0.466	−0.044
First grade	−0.199	0.206	−0.969
Indirect effects	a1 × (b1 + b3 × 2)	SE	95% confidence interval
Moderated mediation (Figure 2)	−0.066	0.031	−2.120*

Note. High school education (1 = high school graduate and 0 = less than 12th grade education); minority (White = 1 and Others = 0); public assistance (1 = with history and nonhistory = 0).

*p < .05. **p < .01.

Based on the current study’s developmental psychopathology conceptualization and ecological framework (Lynch & Cicchetti, 1998), we hypothesized that NP may serve as a contextual factor that moderates the link of ER on kindergarten behavior. Results revealed that the link between ER and kindergarten classroom behavior was positively moderated by NP. A two-way interaction between ER and poverty (higher factor scores indicate less poverty) was statistically significant (p < .001). We probed the interaction using a simple slope method (Aiken & West, 1991) described in Selig and Preacher (2009). Specifically, examining factor scores of −1, 0, and +1 SD on the poverty mean score revealed that at lower poverty levels, the effect of ER on kindergarten behaviors was statistically significant. The simple slope for −1 SD below the mean level in NP was significant, B (β) = .375 (.1221), t = 3.0701, p = .0025. In contrast, at average poverty level (i.e., NP = 0), and above mean level of poverty (i.e., NP = 1 SD), the effect of ER on kindergarten was insignificant, B (β) = .108 (.089), t = 1.2136, p = .2267; B (β) = .159 (.1221), t = −1.3017, p = .1949, respectively. In addition, to further clarify the interaction direction, we conducted multiple group analyses using two groups: high and low NP based on median score computed from the sum of all of the NP indicators. Analyses revealed that the link between ER and kindergarten classroom behaviors was significant in the low NP (B = .473; Z = 3.580; p < .001) group and insignificant in the high NP group (B = −.102; Z = −0.980; p = .327).

Subsequently, we tested for moderated mediation (i.e., conditional indirect effects) using the product coefficient method described in Preacher et al.’s (2007) Model 3. Specifically, we wanted to test whether the association between ER and kindergarten performance is moderated by NP within the mediational path models. The conditional indirect test was significant, that is, β = −.066 (.031); z = −2.120; p < .05, supporting a significant moderated mediation effect. Thus, in neighborhoods with relatively less poverty, ER mediated the relation between neglect severity and school adaptation; however, in neighborhoods with extreme poverty, the impact of ER was no longer significant.

Discussion

Results from this study provide broad support for the hypotheses generated by our ecological transactional analysis of the effects of child neglect. First, there was evidence indicating that the effects of early experiences of child neglect on subsequent adaptation to school were mediated by prior development. Second, there was evidence that adversity occurring in other parts of the ecology (such as NP) provide an important context that can modify the cascading processes linking child neglect to school performance.

As predicted, children who experienced neglect were more likely to struggle as they entered school. Neglected children had more difficulty in attending to tasks and managing their behavior in kindergarten, and they were more likely to have lower grades as they moved through first grade. As has been reported for neglected children in later grades (Eckenrode et al., 1993; Kendall-Tackett & Eckenrode, 1996; Leiter & Johnsen, 1994; Perez & Widom, 1994; Pianta, Egeland, & Erickson, 1989; Wodarski et al., 1990), difficulties in school adaptation were evident at the time of school entry. This study examined the processes through which these school difficulties developed.

Children who experienced more severe neglect in early childhood were exhibiting more cognitive difficulties at age 4 than nonneglected children. More severe neglect was related to lower skills in receptive and expressive language and IQ. Severe neglect may lead to lower cognitive performance for several reasons. Severely neglected children may experience poor nutrition, lack of medical care, and unhealthy living environments that may undermine healthy development. Their experience of stress and exposure to additional risk factors—such as poverty—may have a negative impact on brain development and neuroendocrine regulatory systems (Alink, Cicchetti, Kim, & Rogosch, 2012; Cicchetti, Rogosch, Howe, & Toth, 2010; DeBellis, 2005; Shonkoff, Richter, van der Gaag, & Bhutta, 2012). In addition, parents who are less educated may provide fewer models of educational attainment and be less likely to provide developmentally appropriate stimulation (Slack, Holl, McDaniel, Yoo, & Bolger, 2004), which includes talking or reading to their children. However, it should be noted that the relation between neglect and cognitive performance in this study was above and beyond the impact of maternal IQ. Although the physiological and interactional processes linking neglect with cognitive performance were beyond the scope of this investigation, it was clear that children were demonstrating the negative impact of severe early neglect as seen in their poor cognitive performance by age 4.

This difficulty with cognitive performance predicted poor adjustment in kindergarten. More severely neglected children who had weaker language and cognitive abilities did not have the necessary resources to negotiate classroom tasks effectively. They were rated by teachers as being less engaged in math and language arts, less behaviorally controlled, and less adept at social interactions with peers. These classroom skills are essential for kindergarten success and predict academic achievement in the first grade. As seen in the significant path from kindergarten to first grade, these behavioral competencies in kindergarten set the stage for academic performance in the first grade. Children who can meet the initial behavioral demands of a formal school environment appear to be better prepared to meet the academic demands of this environment as they move forward. Cognitive performance at age 4 also was directly related to first-grade performance in math and language arts. Although not examined in the current investigation, neglect also may play a role in school success because of attendance problems or failure of neglecting families to enroll children in school or to provide preschool experiences. In our sample, attendance rates in kindergarten were as low as 24% and in first grade were as low as 64% for some children. However, kindergarten absences were not significantly related to first-grade academic performance; this issue should be further explored in future research.

Although school performance is likely to be impacted by the larger context of NP, the poverty indices in this study did not significantly moderate the relations between neglect, cognitive performance, and school adaptation in this sample. Within a low-income sample and controlling for family poverty indices, NP was not significantly associated with the mediational processes linking severity of neglect with classroom performance via cognitive performance at age 4. Further exploration of these processes with a larger sample and a broader socioeconomic spectrum is warranted to examine the role of the context of poverty on developing intellectual and preacademic skills.

A second model demonstrated the complex pathways by which ER influences the relation between neglect severity and school outcomes. In addition to possessing cognitive skills, children’s ability to resolve problems with flexibility, creativity, and self-confidence is critical in navigating entry into the school environment. In the current study, severity of early neglect was significantly associated with children’s ability to negotiate a problem-solving task successfully. Children without neglect or with less severe neglect approached the experimental task with more ER at age 4. However, the impact of ER on kindergarten behavior interacted with NP such that children with higher levels of ER were more successful in their kindergarten classroom behavior and task orientation, but only in the context of less extreme NP. It should be noted that all the families in the current sample were characterized by low socioeconomic status. The finding that greater ER was associated with behavioral competence in kindergarten for children from less impoverished communities is consistent with research showing the importance of self-regulation in promoting adaptive functioning among children living in poverty (Buckner, Mezzacappa, & Beardslee, 2003, 2009). However, for children living in highly impoverished neighborhoods, the ability of ER to promote positive classroom behaviors in kindergarten was weakened.

Along these lines, extremely adverse environments—such as severe neglect in the context of extreme poverty—may overpower developmental competencies that typically contribute to successful early adaptation in school (cf. Eisenberg, Chang, Ma, & Huang, 2009; Evans & Rosenbaum, 2008; Raver, 2012). This conclusion is consistent with Nikulina, Widom, and Czaja’s (2011) finding that benefits to academic achievement associated with low NP were only evident in the nonneglected control group. Neglected children in their sample did not benefit by a lack of NP. Presumably, the deleterious effects of neglect overwhelmed ameliorative effects associated with low NP. In a similar vein, the combination of severe neglect and extreme poverty appear to represent a greater combined challenge than ER is capable of handling. Children facing such extreme ecological adversity may require even greater support as they enter school in order to bring them eventually to a more favorable trajectory of school functioning. As in the first model, adaptive behaviors in kindergarten were predictive of academic achievement in first grade, thus emphasizing the importance of understanding the processes that shape early pathways to school performance.

These findings regarding the mediational role of ER are similar to those of Shonk and Cicchetti (2001) in terms of ER’s association with maltreatment and behavioral maladjustment in school. They found that ER mediated the link between maltreatment and behavioral maladjustment; it did not, however, mediate the relation between maltreatment and academic maladjustment. The current study provided a more detailed ecological analysis by examining the severity of neglect in the context of NP. The current findings indicated that at school entry, ER predicted behavioral competence in kindergarten for neglected children, but only when neighborhood economic adversity was less severe. These behavioral patterns were predictive of subsequent academic performance in first grade, although the measurement approach was different than that of Shonk and Cicchetti. This investigation advanced the work of Shonk and Cicchetti by including a 2-year longitudinal assessment, including assessment of poverty in the exosystem, and specifying dimensions within child maltreatment.

Limitations

This investigation was limited by a relatively small sample size. Although the recruitment and longitudinal follow-up of 111 maltreated and 59 nonmaltreated children was large enough to identify significant processes associated with severity of neglect, the sample size did not permit examination of gender effects that have been found to play a role in other longitudinal studies (Appleyard, Yang, & Runyan, 2010). Moreover, there were not enough subjects to identify pathways associated with subtypes other than neglect or with subtype combinations or to disentangle other dimensions of maltreatment (Barnett et al., 1993). However, the ability to examine the severity of neglect during early childhood represents an advance in the literature. Most studies of child maltreatment focus on the presence or absence this type of caregiving failure. Relatively few studies to date have looked at maltreatment as reflecting a continuum of problematic and dangerous behaviors ranging in severity. The few studies that have included the severity of maltreatment as an independent variable reveal that maltreatment severity makes unique contributions to children’s maladjustment (Manly et al., 2001). The current study adds to this literature demonstrating that not only is child neglect damaging to children’s opportunities for school success, but the developmental challenges increase as neglect becomes more severe. Future research should continue to examine processes by which neglect impacts school functioning.

An additional limitation was the somewhat low interrater reliability for the ER measure. However, all ER data were double coded, and discrepancies were resolved by consensus led by an expert coder. Kindergarten report card effort scores were available for 121 of the 170 subjects. This missing data may have limited estimation accuracy and power in the models. However, given that the rate of missing data peaked at 28.8% on a single factor at only 1 point of time in the longitudinal study, and met missing criteria for MAR, the employment of a state-of-the-art missing data technique (FIML) should have yielded robust and trustworthy model estimations (Schafer & Graham, 2002).

Conclusion

Children who experience neglect are at risk of maladaptation in cognitive development, resilient and flexible problem solving, behavioral adjustment in kindergarten classrooms, and academic performance in first grade. Severity of neglect impacts initial school adaptation, at least in part through its impact on cognitive performance and ER. The context of NP poses additional challenges that may increase risk of school failure, especially for children who do not receive sufficient familial support. ER and cognitive abilities can play protective roles in buffering the effect of neglect on school adaptation. For children in extremely poor neighborhoods, however, this protective influence is overwhelmed by severe contextual challenges presented by poverty. Preventive approaches should focus on reducing the occurrence of child neglect as well as assisting families with promoting cognitive and coping skills when children have experienced neglect. Ameliorating negative environments for children in extreme poverty also is critical to reduce risks of school failure. Providing high-quality preschool and elementary programs can promote positive developmental trajectories for high-risk children (Blair & Raver, 2012). During times of economic downturn, risks for children increase while funding for preventive programs is often cut. Failing to invest in the development of young children’s cognitive and resilient coping skills may result in long-term consequences for their academic trajectory (Heckman, 2006). It is imperative that investments in promoting positive adaptation for high-risk children be a priority to prevent school difficulties.

Footnotes

Acknowledgments

We appreciate the support of the Monroe County Department of Human Services and the valuable time and contributions of all of the children, families, and teachers who participated in the project.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the National Institute of Mental Health (NIMH), the Administration on Child, Youth, and Families (ACYF) Children’s Bureau, and the U.S. Department of Education.

References

Aber

J. L.

Jones

S. M.

Brown

J. A.

Chaudry

Samples

(1998). Resolving conflict creatively: Evaluating the developmental effects of a school-based violence prevention program in neighborhood and classroom context. Development and Psychopathology, 10, 187–213.

Administration on Children and Families. (2005, with revision in 2010). Research Brief No. 10: From early involvement with child welfare services to school entry: Wave 5 follow-up of infants in the National Survey of Child and Adolescent Well-Being. Retrieved from http://www.acf.hhs.gov/programs/opre/abuse_neglect/nscaw/reports/early_involve_child_welfare/nscaw_rb10_2_col_rev_mbw.pdf

Aiken

L. S.

West

S. G.

(1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage.

Ainsworth

J. W.

(2002). Why does it take a village? The mediation of neighborhood effects on educational achievement. Social Forces, 81, 117–152.

Alink

L. R.

Cicchetti

Kim

Rogosch

F. A.

(2012). Longitudinal associations among child maltreatment, social functioning, and cortisol regulation. Developmental Psychology, 48, 224–236. doi:10.1037/a0024892

Allen

R. E.

Oliver

J. N.

(1982). The effects of child maltreatment on language development. Child Abuse and Neglect, 6, 299–305.

Appleyard

Yang

Runyan

(2010). Delineating the maladaptive pathways of child maltreatment: A mediated moderation analysis of the roles of self-perception and social support. Development and Psychopathology, 22, 337–352.

Barnett

Manly

J. T.

Cicchetti

(1993). Defining child maltreatment: The interface between policy and research. In Cicchetti

Toth

S. L.

(Eds.), Child abuse, child development, and social policy (pp. 7–73). Norwood, NJ: Ablex.

Blair

Raver

C. C.

(2012). Child development in the context of adversity: Experiential canalization of brain and behavior. American Psychologist, 67, 309–318.

10.

Block

J. H.

(1980). The role of ego-control and ego-resiliency in the organization of behavior. In Collins

W. A.

(Ed.), Development of cognition, affect, and social relations. The Minnesota Symposia on Child Psychology: Vol. 13 (pp. 39–101). Hillsdale, NJ: Lawrence Erlbaum.

11.

Bright

C. L.

Jonson-Reid

(2008). Onset of juvenile court involvement: Exploring gender-specific associations with maltreatment and poverty. Children and Youth Services Review, 30, 914–927.

12.

Brown

(2006). Confirmatory factor analysis for applied research. New York, NY: Guilford Press.

13.

Buckner

J. C.

Mezzacappa

Beardslee

W. R.

(2003). Characteristics of resilient youths living in poverty: The role of self-regulatory processes. Development and Psychopathology, 15, 139–162.

14.

Buckner

J. C.

Mezzacappa

Beardslee

W. R.

(2009). Self-regulation and its relations to adaptive functioning in low income youths. American Journal of Orthopsychiatry, 79, 19–30.

15.

Cicchetti

Carlson

(1979). Demographics interview (Unpublished measure, available from D. Cicchetti). University of Rochester, Mt. Hope Family Center, Rochester, NY.

16.

Cicchetti

Rogosch

F. A.

Howe

M. L.

Toth

S. L.

(2010). The effects of maltreatment and neuroendocrine regulation on memory performance. Child Development, 81, 1504–1519. doi:10.1111/j.1467-8624.2010.01488.x.

17.

Cicchetti

Valentino

(2006). An ecological-transactional perspective on child maltreatment: Failure of the average expectable environment and its influence on child development. In Cicchetti

Cohen

(Eds.), Development and Psychopathology: Risk, disorder, and adaptation (2nd ed., Vol. 3, pp. 129–201). New York, NY: John Wiley.

18.

Coulton

Korbin

Chow

(1995). Community level factors and child maltreatment rates. Child Development, 66, 1262–1276.

19.

Culp

R. E.

Watkins

R. V.

Lawrence

(1991). Maltreated children’s language and speech development: Abused, neglected, and abused and neglected. First Language, 11, 377–389.

20.

Currie

Widom

C. S.

(2010). Long-term consequences of child abuse and neglect on adult economic well-being. Child Maltreatment, 15, 111–120.

21.

DeBellis

M. D.

(2005). The psychobiology of neglect. Child Maltreatment, 10, 150–172.

22.

Duncan

Aber

J. L.

(1997). Neighborhood models and measures. In Brooks-Gunn

Duncan

Aber

J. L.

(Eds.), Neighborhood poverty: Context and consequences for children: Vol. 1. Six studies of children in families and neighborhoods (pp. 62–78). New York, NY: Russell Sage.

23.

Duncan

G. J.

Dowsett

C. J.

Claessens

Magnuson

Huston

A. C.

Klebanov

Japel

(2007). School readiness and later achievement. Developmental Psychology, 43, 1428–1446.

24.

Eccles

J. S.

Lord

Roeser

R. W.

(1996). Round holes, square pegs, rocky roads, and sore feet: A discussion of stage-environment fit theory applied to families and school. In Cicchetti

Toth

S. L.

(Eds.), Rochester symposium on developmental psychopathology: Vol. 7, Adolescence: Opportunities and challenges (pp. 47–92). Rochester, NY: University of Rochester Press.

25.

Eckenrode

Laird

Doris

(1993). School performance and disciplinary problems among abused and neglected children. Developmental Psychology, 29, 53–62.

26.

Egeland

(1991). A longitudinal study of high-risk families: Issues and findings. In Starr

R. H.

Wolfe

D. A.

(Eds.), The effects of child abuse and neglect: Issues and research (pp. 33–56). New York, NY: Guilford Press.

27.

Egeland

(1997). Mediators of the effects of child maltreatment on developmental adaptation in adolescence. In Cicchetti

Toth

S. L.

(Eds.), Rochester symposium on developmental psychopathology, volume 8: Developmental perspectives on trauma: Theory, research, and intervention (pp. 403–434). Rochester, NY: University of Rochester Press.

28.

Eisenberg

Chang

Huang

(2009). Relations of parenting style to Chinese children’s effortful control, ego resilience, and maladjustment. Development and Psychopathology, 21, 455–477.

29.

Erickson

Egeland

Pianta

(1989). The effects of maltreatment on the development of young children. In Cicchetti

Carlson

(Eds.), Child maltreatment: Theory and research on the causes and consequences of child abuse and neglect (pp. 647–684). New York, NY: Cambridge University Press.

30.

Evans

G. W.

Kim

(2013). Childhood poverty, chronic stress, self-regulation, and coping. Child Development Perspectives, 7, 43–48.

31.

Evans

G. W.

Rosenbaum

(2008). Self-regulation and the income-achievement gap. Early Childhood Research Quarterly, 23, 504–514.

32.

Fishbein

Warner

Krebs

Trevarthen

Flannery

Hammond

(2009). Differential relationships between personal and community stressors and children’s neurocognitive functioning. Child Maltreatment, 14, 299–315.

33.

Gaudin

J. M.

(1999). Child neglect: Short-term and long-term consequences. In Dubowitz

(Ed.), Neglected children: Research, practice, and policy (pp. 89–108). Thousand Oaks, CA: Sage.

34.

Gowan

(1993). Effects of neglect on the early development of children: Final report. Washington, DC: National Clearinghouse on Child Abuse and Neglect, National Center on Child Abuse and Neglect, Administration for Children and Families.

35.

Haskett

M. E.

Nears

Ward

C. S.

McPherson

A. V.

(2006). Diversity in adjustment of maltreated children: Factors associated with resilient functioning. Clinical Psychology Review, 26, 796–812.

36.

Heckman

J. J.

(2006). Skill formation and the economics of investing in disadvantaged children. Science, 312, 1900–1902.

37.

Hildyard

K. L.

Wolfe

D. A.

(2002). Child neglect: Developmental issues and outcomes. Child Abuse & Neglect, 26, 679–695.

38.

Kendall-Tackett

K. A.

Eckenrode

(1996). The effects of neglect on academic achievement and disciplinary problems: A developmental perspective. Child Abuse and Neglect, 20, 161–169.

39.

Korbin

J. E.

Coulton

C. J.

Chard

Platt-Houston

(1998). Impoverishment and child maltreatment in African American and European American neighborhoods. Development and Psychopathology, 10, 215–233.

40.

Leiter

Johnsen

M. C.

(1994). Child maltreatment and school performance. American Journal of Education, 102, 154–189.

41.

Letzring

T. D.

Block

Funder

D. C.

(2005). Ego-control and ego-resiliency: Generalization of self-report scales based on personality descriptions from acquaintances, clinicians, and the self. Journal of Research in Personality, 39, 395–422.

42.

Luster

Reischl

Gassaway

Gomaa

(1995, March). Factors related to early school success among African American children from low income families. Paper presented at the Biennial Meeting of the Society for Research in Child Development, Indianapolis, IN.

43.

Lynch

Cicchetti

(1992). Maltreated children’s reports of relatedness to their teachers. New Directions for Child Development, 57, 81–107.

44.

Lynch

Cicchetti

(1998). An ecological-transactional analysis of children and contexts: The longitudinal interplay among child maltreatment, community violence, and child’s symptomatology. Development and Psychopathology, 10, 235–257.

45.

MacKinnon

D. P.

Fairchild

A. J.

Fritz

M. S.

(2007). Mediation analysis. Annual Review Psychology, 58, 593.

46.

Manly

J. T.

(2005). Invited commentary: Advances in research definitions of child maltreatment. Child Abuse and Neglect: The International Journal, 29, 413–619.

47.

Manly

J. T.

Kim

J. E.

Rogosch

F. A.

Cicchetti

(2001). Dimensions of child maltreatment and children’s adjustment: Contributions of developmental timing and subtype. Development and Psychopathology, 13, 759–782.

48.

Masten

A. S.

Cicchetti

(2010). Developmental cascades. Development and Psychopathology, 22, 491–495.

49.

McDonald

R. P.

M. H.

(2002). Principles and practice in reporting statistical equation analyses. Psychological Methods, 7, 64–82.

50.

Murry

V. M.

Berkel

Gaylord-Harden

N. K.

Copeland-Linder

Nation

(2011). Neighborhood poverty and adolescent development. Journal of Research on Adolescence, 21, 114–128.

51.

Muthen

L. K.

Muthen

B. O.

(1998–2010). Mplus user’s guide: Sixth edition. Los Angeles, CA: Author.

52.

Nikulina

Widom

C. S.

Czaja

(2011). The role of childhood neglect and childhood poverty in predicting mental health, academic achievement and crime in adulthood. American Journal of Community Psychology, 48, 309–321.

53.

Perez

Widom

C. S.

(1994). Childhood victimization and long-term intellectual and academic outcomes. Child Abuse & Neglect, 18, 617–633.

54.

Perkins

P. E.

Hightower

A. D.

(2002). Teacher-child rating scale: Examiner’s manual. Rochester, NY: Children’s Institute.

55.

Pianta

Egeland

Erickson

(1989). The antecedents of child maltreatment: Results of the mother child interaction research project. In Cicchetti

Carlson

(Eds.), Child maltreatment: Theory and research on the causes and consequences of child abuse and neglect (pp. 203–252). New York, NY: Cambridge University Press.

56.

Pianta

R. C.

(1999). Enhancing relationships between children and teachers. Washington, DC: American Psychological Association.

57.

Polansky

N. A.

Chalmers

M. A.

Buttenweiser

Williams

D. P.

(1981). Damaged parents. Chicago, IL: University of Chicago Press.

58.

Preacher

K. J.

Hayes

A. F.

(2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879–891.

59.

Preacher

K. J.

Rucker

D. D.

Hayes

A. F.

(2007). Addressing moderated mediation hypotheses: Theory, methods, and prescriptions. Multivariate Behavioral Research, 42, 185–227.

60.

Raver

C. C.

(2012). Low-income children’s self-regulation in the classroom: Scientific inquiry for social change. American Psychologist, 67, 681–689.

61.

Scarborough

A. A.

Lloyd

C. E.

Barth

R. P.

(2009). Maltreated infants and toddlers: Predictors of developmental delay. Journal of Developmental and Behavioral Pediatrics, 30, 489–498.

62.

Scarborough

A. A.

McCrae

J. S.

(2010). School-age special education outcomes of infants and toddlers investigated for maltreatment. Cildren and Youth Services Review, 32, 80–88.

63.

Schafer

J. L.

Graham

J. W.

(2002). Missing data: Our view of the state of the art. Psychological Methods, 7, 147–177.

64.

Sedlak

A. J., Mettenburg, J., Basena, M., Petta, I., McPherson, K., Greene, A., & Li, S

(2010). Fourth national incidence study of child abuse and neglect (NIS–4): Report to congress. Washington, DC: U.S. Department of Health and Human Services, Administration for Children and Families.

65.

Selig

J. P.

Preacher

K. J.

(2009). Mediation models for longitudinal data in developmental research. Research in Human Development, 6, 144–164.

66.

Shonk

S. M.

Cicchetti

(2001). Maltreatment, competency deficits, and risk for academic and behavioral maladjustment. Developmental Psychology, 37, 3–17.

67.

Shonkoff

J. P.

Richter

van der Gaag

Bhutta

Z. A.

(2012). An integrated scientific framework for child survival and early childhood development. Pediatrics, 129, 1–13.

68.

Shrout

P. E.

Bolger

(2002). Mediation in experimental and nonexperimental studies: New procedures and recommendations. Psychological Methods, 7, 422–445.

69.

Slack

K. S.

Holl

J. L.

McDaniel

Yoo

Bolger

(2004). Understanding the risks of child neglect: An exploration of poverty and parenting characteristics. Child Maltreatment, 9, 395–408.

70.

Swanson

Valiente

Lemery-Chalfant

O’Brien

T. C.

(2011). Predicting early adolescents’ academic achievement, social competence, and physical health from parenting, ego resilience, and engagement coping. The Journal of Early Adolescence, 31, 548–576.

71.

Toth

S. L.

Cicchetti

(1996). The impact of relatedness with mother on school functioning in maltreated children. Journal of School Psychology, 34, 247–266.

72.

Tramontana

M. G.

Hooper

S. R.

Selzer

S. C.

(1988). Research on the preschool prediction of later academic achievement: A review. Developmental Review, 8, 89–146.

73.

U.S.

Census

. (2000). U.S. Census Bureau: Census 2000. Retrieved from http://www2.census.gov/census_2000/datasets/

74.

Wechsler

(1989). Wechsler preschool and primary scale of intelligence—Revised manual. New York, NY: The Psychological Corporation.

75.

Wechsler

(1999). Wechsler abbreviated scale of intelligence—Manual. New York, NY: The Psychological Corporation.

76.

Williams

K. T.

(1997). Expressive vocabulary test manual. Circle Pines, MN: American Guidance Service.

77.

Williams

K. T.

Wang

(1997). Technical references to the Peabody picture vocabulary test—Third edition (PPVT-III). Circle Pines, MN: American Guidance Service.

78.

Wodarski

J. S.

Kurtz

P. D.

Gaudin

J. M.

Howing

P. T.

(1990). Maltreatment and the school-age child: Major academic, socioemotional, and adaptive outcomes. Social Work, 35, 506–513.

79.

Yang

Nay

Hoyle

R. H.

(2010). Three approaches to using lengthy ordinal scales in structural equation models: Parceling, latent scoring, and shortening scales. Applied Psychological Measurement, 34, 122–142. doi:10.1177/0146621609338592

80.

Yuan

K. H.

Bentler

P. M.

(2000). Inferences on correlation coefficients in some classes of nonnormal distributions. Journal of Multivariate Analysis, 72, 230–248.