Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Jun 1.
Published in final edited form as: J Fam Psychol. 2015 May 4;29(3):382–393. doi: 10.1037/fam0000087

Immigration and the Interplay of Parenting, Preschool Enrollment, and Young Children's Academic Skills

Arya Ansari 1,*, Robert Crosnoe 2
PMCID: PMC4461446  NIHMSID: NIHMS677856  PMID: 25938712

Abstract

This study tested a conceptual model of the reciprocal relations among parents’ support for early learning and children's academic skills and preschool enrollment. Structural equation modeling of data from 6,250 children (ages 2-5) and parents in the nationally representative Early Childhood Longitudinal Study-Birth Cohort (ECLS-B) revealed that parental support for early learning was associated with gains in children's academic skills, which, in turn, were associated with their likelihood of preschool attendance. Preschool experience then was associated with further gains in children's early academic competencies, which were then associated with increased parental support. These patterns varied by parents' nativity status. Specifically, foreign-born parents' support for early learning was directly linked with preschool enrollment and the association between the academic skills of children and parental support was also stronger for foreign-born parents. These immigration-related patterns were primarily driven by immigrant families who originated from Latin America, rather than Asia and did not vary by immigrants’ socioeconomic circumstances. Together, these results underscore the value of considering the synergistic relations between the home and school systems as well as “child effects” and population diversity in developmental research.

Keywords: Immigration, Parental Support for Early Learning, Preschool Enrollment, Child Effects, ECLS-B

Parenting and preschool are major foci of research on children's academic skills (Crosnoe et al., 2010; Lareau, 2004; Lerner, 2006; Winsler et al., 2008). This research is theoretically grounded, reflecting core themes of influential perspectives that highlight contexts of child development. It is also policy oriented, reflecting awareness that human capital interventions bring greater returns to investment when they target the settings of early childhood (Heckman, 2006). Just because the literatures on parenting and preschool are so extensive, however, does not mean that there is not more to learn. Indeed, returning to theory for guidance about how to approach these issues in new ways can facilitate the translation between research and action.

In this spirit, this study tests a conceptual model that links parents’ engagement in early academically focused parenting and their enrollment of children in preschool while also highlighting children's development of the academic skills that have been established as important precursors for achievement in the K-12 system. Drawing on central tenets of the developmental systems perspective (Lerner, 2006), this model captures the bidirectional interplay of parenting behavior and children's preschool enrollment. It also highlights “child effects”, or the potential for children to elicit parental investment and to “select” into preschool, with selection here referring to the tendency for children's skills to encourage parents to find opportunities for them. This interplay between parents and preschools and its relation to child elicitation are, in turn, expected to vary by families’ immigration statuses, which tap into their socioeconomic resources and familiarity with educational institutions in the U.S.

We test this conceptual model using nationally representative data on children and parents. Although the Early Childhood Longitudinal Study-Birth Cohort (ECLS-B) is somewhat limited relative to many community-based studies of parenting, especially in terms of measurement, it offers a valuable opportunity to capture dynamic and longitudinal processes while exploring diversity across and within segments of the population, which is not always possible in other datasets. The results of this study, therefore, can advance theory by demonstrating the insights to be gained by reversing some traditionally assumed and expected directions in the links among families, preschools, and children. It can also have more practical significance by pointing to potential targets for intervention during early childhood, which is increasingly viewed as critical to both development itself and to developmental disparities.

Parenting and Early Childhood Academic Development across Diverse Groups

The academic skills that children bring into kindergarten—their knowledge of letters and words and familiarity with numbers and basic math concepts—establish their positions in the academic hierarchy in ways that structure academic opportunities as they move through elementary school and beyond (Duncan et al., 2007). Differences in such skills reflect differences in early childhood experiences, and, thus, what happens in the years prior to the start of formal schooling can lead to disparities between children (Entwisle, Alexander, & Olson, 2005). Supporting the learning of young children, therefore, is an effective way to increase their educational attainment and break intergenerational cycles of inequality within families.

Reflecting the long-standing centrality of parenting in developmental theories of early childhood, much of the inquiry into early academic achievement has focused on parents (NICHD Early Child Care Network [ECCRN], 2005). In general, parents shape children's early learning by engaging in sensitive behaviors that give them confidence to explore (Mistry, Benner, Biesanz, Clark, & Howes, 2010; Gershoff, Aber, Raver, & Lennon, 2007). In recent years, more attention has been paid to the intentional ways that parents try to boost their children's school success, such as providing them with learning resources at home, engaging them in cognitively stimulating activities, and enrolling them in extracurricular activities (Cheadle, 2008; Crosnoe et al., 2010; Gershoff et al., 2007). These behaviors are rooted in socioeconomic backgrounds, in part because they often take money but also because individuals are socialized into parenting styles within their socioeconomic milieus that differ in how much they stress intensive approaches to investing in children's educational prospects. Whether such behavior indicates “good” parenting or not, it is typically rewarded by a U.S. educational system that is organized by the parenting ideals of White middle class families. That is, U.S. teachers often expect parents to conform to conventional ideas about academically focused parenting and invest more in children (often unconsciously) whose parents they perceive (rightly or wrongly) as meeting these expectations (Cheadle, 2008; Lareau, 2004; Lee & Bowen, 2006).

Although the ways in which parents manage their children's early learning are shaped by multiple systems that map onto race/ethnicity and socioeconomic status (Cheadle, 2008; Gershoff et al., 2007; Lareau, 2004), immigration is a critical lens through which to study these processes. As a group, immigrant parents are less likely to engage in the types of parenting that are part of this White middle class picture of family academic support (Suarez-Orozco & Suarez-Orozco, 2001). This discrepancy occurs despite their valuing schooling and wanting educational success for their children at levels comparable to U.S.-born parents and despite their clear engagement in other parenting behaviors that intended to support children's development. In this way, their children may be disadvantaged in school, regardless of whether these behaviors have any inherent value (Crosnoe & Turley, 2010).

A large portion of these immigration-related differences reflects socioeconomic disparities tied to immigration (Suarez-Orozco & Suarez-Orozco, 2001). Although many immigrant parents are socioeconomically advantaged, and although socioeconomic variability within the immigrant population is substantial, levels of family income and educational attainment are, on average, lower in this population (Cheadle, 2008; Hernandez, 2006). Given the ways in which socioeconomic advantages (and associated milieus) support some of the more active and intensive aspects of parental support for learning (e.g., income used to purchase activities and other opportunities, educational attainment emphasizing the need for cognitive stimulation), these socioeconomic disparities related to immigration have implications for such support (Suarez-Orozco & Suarez-Orozco, 2001). Yet, immigration-related disparities in parental support for learning are unlikely to be solely a function of socioeconomic circumstances. No matter how much education that immigrant parents have attained or how much money they have made, they tend to have less experience in the U.S. educational system than U.S.-born parents because most participated in schooling in their home country. Even immigrant parents with high levels of education, therefore, may enjoy the benefits of that human capital but not have as much experience with how U.S. schools work. As such, they may be less familiar with the written and unwritten rules of U.S. schools and associated norms about parenting than U.S.-educated parents of the same socioeconomic circumstances (Crosnoe, Ansari, Purtell, & Wu, forthcoming). This confluence between socioeconomic status and U.S.-based education experience means that immigrant families are a valuable group with which to understand variability in how academically supportive parenting develops over time and what its implications are.

As highlighted by the developmental systems perspective, parents’ support for children's early learning and the family system overlap with extra-familial systems (Lerner, 2006). For example, the primary non-familial setting of early learning is preschool, with a substantial number of children attending formal early education programs prior to entering the K-12 system (Kids Count, 2013). Still, significant disparities in enrollment remain, including lower rates of participation among children from immigrant families due to the mechanisms discussed above (Crosnoe & Turley, 2010). These programs hold great promise, however, for promoting children's academic skills, especially for children from immigrant families (Winsler et al., 2008). Increasingly, developmental scientists are exploring the factors that influence parents’ preschool enrollment decisions as a means of identifying groups that should be targeted for outreach (Fuller, Holloway, & Liang, 1996). Several mechanisms have been identified that shape parents’ decisions regarding preschool enrollment including their socioeconomic circumstance, cultural orientations to non-parental care of young children, priorities for their children's education, and contextual opportunities and constraints (Meyers & Jordan, 2006; Coley, Votruba-Drzal, Collins, Miller, 2014). One understudied mechanism, however, is parenting, as both preschool enrollment and the academic benefits derived from it could be byproducts of parents’ efforts to support their children's development and to gain them a competitive edge in formal schooling (NICHD ECCRN & Duncan, 2003; Fuller, 2007).

Figure 1a presents this basic systems-based model, with parents’ support for their children's early learning promoting their academic skills over time and preschool enrollment serving as one potential channel for this association. Thus, immigration-related differences in parenting can influence immigration-related disparities in achievement directly or indirectly by influencing preschool enrollment. This model is quite simple, but it can be elaborated in important ways by drawing on other insights of developmental systems perspective.

Figure 1a.

Figure 1a

Open in a new tab

Basic systems-based early childhood model

Developmental Systems and Family-Preschool-Child Transactions

As discussed above, the systems perspective highlights the ways in which multiple systems work together to shape child development. Importantly, the emphasis is on two-way transactions (Lerner, 2006). Note that these systems also include the transactions between the child and external systems. In this way, ecological environments do not only act on children; children also elicit new environmental experiences (e.g., child effects; Bell, 1968). Theory, therefore, suggests that the model in Figure 1a can be extended to elucidate the dynamic transactions among parenting, preschool enrollment, and children's skill development.

First, consider that the two-way exchange between parents and preschools means that, in addition to the path from parenting to preschool enrollment already described, the path from preschool enrollment to parenting must also be considered. Indeed, parenting behavior might evolve as a function of children's enrollment in preschool—not only are children influenced by their preschool experiences, so too are their parents. Through the active or passive socialization of parents by teachers, parents may be more likely to engage in a cluster of parenting behaviors seen by U.S. educators as academically supportive. Importantly, these effects on parenting are additive and, thus, preschool may alter long-term trajectories of parenting (Benasich, Brooks-Gunn, & Clewell, 1992; Crosnoe et al., 2012; McCartney, Dearing, & Bub, 2007). Accordingly, early academically focused parenting could lead parents to enroll children in preschool while interactions with preschool can facilitate more academically focused parenting later on.

Second, children are likely to connect parenting and preschool in this two-way exchange, through the ways in which they are influenced by parents and preschools and how they select into different parenting and preschool environments. Parents’ choice of preschool might be driven by children's academic skills, and the lack of attention to this potential for children to elicit more investment (compensatory or enriching) from parents has been identified as a limitation of the literature on the impact of preschool in children's school success (Abner, Gordon, Kaestner, & Korenman, 2013). Additionally, care-related changes in children's early competencies can elicit changes in parenting (Jaeger & Weinraub, 1990). This argument is relevant to understanding why preschool could lead to more parental engagement in academically focused behaviors. Although the child effects literature is far shallower than the one on child outcomes of parenting, growing evidence suggests that normally developing children's academic abilities are associated with parenting (Ansari & Crosnoe, 2015; Crosnoe et al., 2012; Lugo-Gil & Tamis-LeMonda, 2008; Gershoff, Aber, & Clements, 2009). Whether these skills mediate the relations between the two core systems of child development is less clear. Thus, we propose that the skills that children develop early on as a result of academically focused parenting could increase motivation for parents to enroll children in preschool, and children with more developed academic skills (as a result of their preschool experience) could then elicit more academically focused parenting.

Just as immigration status is independently associated with parenting, preschool enrollment, and academic skills, it is also implicated in how they connect as transactional systems. Again, immigrant parents—despite their overall engagement in positive parenting—are generally less oriented to the kinds of academically focused parenting increasingly viewed as appropriate for young children by U.S. educators (Crosnoe, 2010; Lareau, 2004). This is partially because of the financial and time constraints as well as their relative lack of exposure to the White middle-class parenting norms that often organize U.S. schools. Consequently, external stimuli may play a role in overcoming these barriers. In other words, having a child whose early skills elicit adult investment and/or interacting with preschool teachers who encourage parents might make more of a difference for immigrants (especially those who are socioeconomically disadvantaged) than for non-immigrants (especially those who are socioeconomically advantaged), as the latter are more likely to be engaging in such behaviors already. Although no direct evidence supports this contention that elicitation will matter more to parents whose parenting is less variable across contexts and circumstances, related research has shown that socioeconomically disadvantaged women's school-focused parenting is more reactive to the elicitation of children, preschools, and other factors than more advantaged women, whose school-focused parenting is less likely to change in the face of new obstacles or supports (Crosnoe et al., 2012). This socioeconomic pattern is relevant to the immigrant population, which has high levels of socioeconomic disadvantage but also socioeconomic diversity, and it could be magnified if the lower level of familiarity with U.S. schools among immigrant parents not educated in the U.S. also means that they may be more reactive to factors that facilitate academically focused parenting.

Thus, examining socioeconomic heterogeneity within the immigrant population is important. We do so in two ways: 1) examining parent’ socioeconomic status as a stratifying mechanism within the immigrant population and 2) by comparing the two largest streams of immigration to the U.S.—Asia versus Latin America—that differ widely in the educational attainment of who does and does not migrate (with more educated Asians more likely to migrate than their less educated counterparts and the opposite pattern for Latin Americans; Feliciano, 2005). Indeed, the striking differences in these sources of immigration tap into key socioeconomic, cultural, and language differences (generally in favor of Asian immigrants) that may moderate the dynamic processes in our hypothesized model.

Conceptual Model and Hypotheses

Based on these systems-based transactional considerations, Figure 1b presents our elaborated conceptual model. The unidirectional link between parenting and preschool enrollment has been replaced with a two-way exchange. At the same time, the potential for children's skill development to influence—not just reflect—parenting and preschool enrollment has also been incorporated. Following this model, the three guiding hypotheses of this study are:

  • 1)

    Parents’ early academically focused parenting behavior will increase the odds that children will be enrolled in preschool, which, in turn, will lead to increases in parents’ engagement in these same behaviors over time.

  • 2)

    Children's academic skills will link early academically focused parenting to their eventual enrollment in preschool and then link their preschool enrollment to their later experience of more academically focused parenting.

  • 3)

    The two-way exchanges between parenting and preschool enrollment, and the role of children's academic skills in them, will be stronger for immigrant parents especially lower-SES and Latino/a immigrants.

Figure 1b.

Figure 1b

Open in a new tab

Elaborated conceptual model of the study

Method

Data and Sample

ECLS-B is a nationally representative sample of children (n = 10,700) born in the U.S. Children were sampled using a multistage, stratified, clustered design based on birth certificates in primary sampling units (counties), excluding children who died, those adopted after the issuance of the birth certificate, and those were born to mothers younger than 15 years of age. The children in the ECLS-B were followed longitudinally at 9-months, 2-, 4- (preschool), and 5-years of age (kindergarten) and data were collected from parent, caregiver, and teacher interviews as well as direct child assessments (for more information on sampling, see NCES, 2005). Per IES/NCES guidelines, we included all children who remained in the sample through kindergarten (n = 6,250; sample sizes rounded to the nearest 50 per IES/NCES guidelines). The racial/ethnic composition of this sample was 42% White, 20% Latino/a, 16% African-American, 10% Asian-American, and 12% other (see Table 1 for sample descriptives).

Table 1.

Sociodemographic Characteristics Among Sampled Children and Families

Age Foura
 Age Five
Variable % or M(SD) n % or M(SD) n
Child race/ethnicity
    White 41.6% 2600 - -
    African-American 16.0% 1000 - -
    Latino/a 20.0% 1250 - -
    Asian-American 10.4% 650 - -
    Other race/ethnicity 12.0% 750 - -
Gender (female) 49.6% 3100 - -
Assessment age (months) 52.84 (4.07) 6250 65.15 (3.74) 6250
Assessment month 3.80 (2.15) 6250 3.10 (1.47) 6250
Bayley Mental scale (9 months) 74.99 (9.97) 6250 - -
Bayley Mental scale (24 months) 125.70 (10.82) 6250 - -
Childcare at age two
    Parental care 60.0% 3150 - -
    Relative care 15.2% 800 - -
    Non-relative care 11.5% 600 - -
    Center-based care 13.3% 700 - -
Age of first non-parental care (months) 5.38 (5.29) 4300 - -
Parent nativity status
    Foreign-born 26.4% 1650 - -
    U.S.-born 73.6% 4600 - -
Parent highest education
    Less than high school 9.6% 600 8.9% 550
    High school or equivalent 21.6% 1350 23.4% 1450
    Some college 31.2% 1950 29.9% 1850
    Bachelors 17.6% 1100 18.4% 1150
    Graduate school 20.0% 1250 19.4% 1200
Parent marital status
    Married 68.9% 4200 68.3% 4200
    Separated/divorced 8.2% 500 9.7% 600
    Single 20.5% 1250 19.5% 1200
    Non-biological 2.4% 150 2.5% 150
Parents’ expectation for children's ed.
    High school degree or less 9.6% 600 8.7% 550
    Some college 13.6% 850 14.3% 900
    Bachelor's degree 41.6% 2600 39.7% 2500
    Some graduate school 35.2% 2200 37.3% 2350
Mothers’ age 32.70 (6.89) 6200 - -
Mothers’ employment status
    Full time 42.3% 2600 44.7% 2750
    Part time 18.7% 1150 18.7% 1150
    Unemployed 39.0% 2400 36.6% 2250
Per capital household incomeb 1.84 (0.95) 6250 1.89 (0.95) 6250
Household language
    Language majority 79.3% 4950 - -
    Language minority 20.7% 1300 - -
Number of siblings 1.47 (1.14) 6250 1.54 (1.15) 6250
Region
    Northeast 13.6% 850 13.5% 850
    Midwest 24.0% 1500 23.8% 1500
    West 26.4% 1650 26.2% 1650
    South 36.0% 2250 36.5% 2300
Urbanicity
    Urbanized areas 69.9% 4300 70.6% 4200
    Urban clusters 13.0% 800 12.6% 750
    Rural 17.1% 1050 16.8% 1000
Open in a new tab

Note.

a

Unless otherwise noted.

b

the per capita incomes across both time points correspond to roughly $7,700 per individual in the household.

Measures

Parental support of early learning

Three aspects of parenting were considered as part of a multi-dimensional construct. For investment, parents (at age 2) reported the number of soft toys, push toys, books, and CDs at home (see Table 2). Given that the items were on different scales and similar to past research using such items (Gershoff et al., 2007), each item was converted into quartiles. Their quartile scores were then summed into a composite for home resources (α = .63). During the kindergarten wave, parents reported the number of books they had at home and whether they had a computer, which we rescaled (1 = no, 4 = yes). Again, we created quartiles for the number of books and computed an overall composite of these two items. In both waves, parents answered questions from the Home Observation for Measurement of the Environment scale (HOME; Caldwell & Bradley, 1984). Specifically, they gave the frequency with which they sang songs, read books, told stories, and discussed books with their children in a ‘typical’ week. Items were scored from 1 (not at all) to 4 (every day) and summed to measure cognitive stimulation (α = .64). Third, at age 2, parents reported whether they had taken their children to a zoo, museum, or story hour and whether they had visited a library to borrow books or DVDs in the past month. Questions were scored as 0 (no) or 1 (yes) and summed to measure cultural or extracurricular activities. Similar questions were not asked at kindergarten. Instead, parents reported whether their children participated in seven activities outside of school: athletics, dance, music, drama, art, performing arts, and crafts.

Table 2.

Descriptive Statistics of Main Variables and Factor Loadings for Latent Variables

Descriptive Statistics Age Two or Foura
 Age Five
M (SD) n M (SD) n
Parental support for early learningb
    Home resourcesc 9.94 (3.08) 6250 5.38 (1.98) 6250
    Cognitive stimulation 11.60 (3.01) 6250 11.62 (2.57) 6250
    Cultural/extracurricular activities 1.12 (1.39) 6250 1.09 (1.30) 6250
Academic skills
    Math 29.53 (9.97) 5950 40.74 (10.96) 6200
    Reading 25.62 (10.54) 6000 39.51 (15.46) 6200
Age four preschool enrollment 0.62 (0.48) 6250 - -
Latent Factors Loadings B β B β
Parental support for early learning
    Home resources 1.00d .62 1.00 .63
    Cognitive stimulation .21*** .54 .28*** .41
    Cultural/extracurricular activities .31*** .43 .17*** .52
Academic skills
    Math 1.00 .86 1.00 .89
    Reading .98*** .89 .72*** .90
Open in a new tab

Note.

a

Parents’ support for early learning was drawn from the age two wave while children's academic skills were from the preschool year.

b

The correlation for the focal variables across time were as follows: home resources, r = .43; cognitive stimulation, r = .40; extracurricular activities, r = .26.

c

The possible range of home resources at age 2 was 0-16 while at age 5 it was 0-8.

d

According to SEM requirements, for each factor, one variable loading was set to 1.00.

***

p < .001.

Children's pre-academic skills

At age 2, children were assessed with the short form of the Bayley Scale of Infant Development (α = .92; Bayley, 1993), which tapped into children's problem solving, counting, comprehension, and receptive and expressive vocabulary skills. During later waves, children's reading/literacy skills were measured with a 37-item assessment developed specifically for the ECLS-B. Content covered letter recognition, early reading, phonological awareness, and knowledge of print. A 28-item math assessment was also developed, covering number sense, properties, measurement, and patterns. Both subscales have demonstrated strong reliability (α = .92; Najaran, Snow, Lennon, & Kinsey, 2010) and have been shown to be strong predictors of future academic success (Duncan et al., 2007). We used the IRT-based scores, which created a single underlying proficiency range across assessments that allowed for direct comparisons over time. Because the math and reading assessments were highly correlated (rs =.79-.82), we created a single latent factor for academic achievement.

Preschool enrollment

When children were 4, parents reported the type of preschool they attended. Approximately 62% reported that their children attended a formal preschool (e.g., center-based care, Head Start), which are generally associated with more positive child outcomes (Magnuson & Waldfogel, 2005) than the less formal childcare arrangements (e.g., relative, non-relative, parental) of the remaining children. Supplementary analyses (not shown) revealed no substantive differences in our general conclusions when comparing informal care, center-based care, and Head Start, which has a strong two-generation focus.

Immigrant groups

Parents also reported on their own (and the other parents’) place of birth, which we used to differentiate between the children of foreign-born parents (n = 1,650; 39% Latino/a, 38% Asian, 10% White, 8% mixed/other, 5% Black) and U.S.-born parents (n = 4,600). Given the adequate sample size, we also compared the experiences of Asian and Latin American immigrants as well as low- and high-SES immigrants proxied by college education, an important threshold for education-related differences in parenting (McLanahan, 2004).

Covariates

All analyses controlled for a full set of child and family factors to account for demographic variability as well as the potential for spuriousness in focal associations. The covariates included in our models tap into families’ cultural backgrounds, socioeconomic circumstances, and access to care as well as child characteristics associated with children's academic achievement, childcare selection, and parenting. For children, these factors included race/ethnicity, gender, scores on the Bayley mental scale at 9 and 24 months, age at assessment, assessment month of school, age of first non-parental care, and age 2 childcare type. Family factors included parent marital status, parents’ education, parent expectations for children's education, mothers’ age, mothers’ employment status, household income, household language, number of siblings, region, and urbanicity. Finally, we controlled for the respondents’ relationship with child even though 98% were biological mothers at baseline, with little change over time. Time-varying covariates were used when available.

Analytical Strategy

Structural equation modeling (SEM) in Mplus (Muthén & Muthén, 2013) was the primary methodological strategy for this study. Essentially, we tested a model that corresponded to Figure 1b, with preschool enrollment (age 4) as an intervening variable between the two parenting latent factors at ages 2 and 5. We also included a latent factor of children's academic skills that were predicted by age 2 parenting and used to predict preschool enrollment. In turn, preschool enrollment was used to predict changes in children's academic skills the following year. Finally, children's academic skills at age 5 were used to predict improvements in parenting. Other time-orderings of variables were examined in ancillary analyses and will be discussed.

All models included: 1) multiple imputation (50 datasets) to address missing data (less than 10%), 2) sampling weights to address non-response bias and to ensure that the sample was representative of the larger population, and 3) stratification and clustering variables to properly estimate standard errors. Model fit was evaluated with the chi-square statistic, Comparative Fit Index (CFI; recommendation, greater than .90) and root-mean square error of approximation (RMSEA; recommendation, less than .05; Hu & Bentler, 1999). Other fit indices were not available with our combination of imputation and complex survey design.

Although we controlled for potential confounds, the extent to which causal inference can be drawn from these models remains in question due to the possibility of unobserved confounds. To address the sensitivity of our findings, we utilized a post-hoc robustness check—the Impact Threshold for Confounding Variables (ITCV; Frank, 2000)—for all statistically significant direct effects. This statistic quantifies how strongly an unknown variable would have to be correlated with both the predictor and outcome to negate the observed association from our models.

Results

An initial step was to test the measurement model for the four latent variables of interest. As can be seen in Table 2, all factor loadings were significant at a minimum probability level of .001 and were comparable across time. The fit for this measurement model was good: CFI = .989, RMSEA = .029, x2 (df = 0) = 153.30, p < .001 (all degrees of freedom less than 25 have been rounded to 0 per IES/NCES regulations). We then estimated the full structural model with the latent factors, observed factors, and all time-varying covariates. This structural model also had good fit: CFI = .932, RMSEA = .008, and x2 (df = 450) = 622.70, p < .001.

Linking Parenting, Preschool Enrollment, and Academic Skills over Time

Results for the full model are presented in Figure 2. Beginning with the top portion of Figure 2, the results revealed no direct link between parenting at age 2 and preschool enrollment at age 4, once the full set of covariates, including prior childcare arrangements, were included. Similarly, no direct link between preschool enrollment and parenting at age 5 was revealed. Thus, the hypothesis of significant feedback between parental support for learning and preschool enrollment over the early childhood years was not supported, at least not at first glance. Shifting attention to the lower portion of Figure 2, however, sheds new light on this interplay. Specifically, we explored the possibility of indirect effects in the absence of initially observed direct effects. Doing so is important—and statistically possible—because it can reveal some potential mechanisms that would otherwise go undetected using the Baron and Kenny procedures (1986) for establishing mediation (for further discussion, see: Hayes, 2009).

Figure 2.

Figure 2

Open in a new tab

Observed model of parents’ support for school readiness, preschool enrollment, and children's academic skills. Standardized direct path coefficients are shown. Note: All variables in the figure above were regressed on the full set of covariates listed in Table 1. *p < .05, ** p < .01, *** p < .001

First, looking at the direct and indirect paths leading out from parental support of learning at age 2, this parenting factor was associated with children's academic skills at age 4 (β = .09, p < .01), which had important ramifications for the rest of the model in the form of indirect effects. For example, children's academic skills at age 4 were associated with their preschool attendance that same year (β = .19, p < .001, OR = 1.04). Thus, children with strong academic skills (+1 standard deviation) were 40% more likely to select into preschool. Because the INDIRECT command and bootstrapped modeling were not available, we used the Sobel test. As expected, parenting at age 2 was indirectly associated with preschool attendance through children's academic skills at age 4 (βindirect = .02, p < .01, OR = 1.02). As another example, academic skills were highly stable from age 4 to 5 (β = .69, p < .001) and, through this stability, parenting at age 2 was indirectly associated with age 5 academic skills (βindirect = .06, p < .001).

Second, looking at the direct and indirect paths leading out from preschool enrollment at age 4, this factor was associated with children's academic skills a year later (β = .05, p < .05). Children's academic skills at age 5, in turn, were associated with parenting that same year (β = .24, p < .001). Finally, preschool enrollment at age 4 was indirectly associated with parental support for early learning via its intermediary association with children's academic skills (βindirect = .01, p < .05).

To gauge the robustness of the direct paths in the focal model to threats from unobserved confounds, we calculated the ITCV analyses. The largest ITCV value was .10 for the association between children's academic skills and their preschool enrollment. Thus, controlling for an unobserved confound would negate our finding under the condition that the unaccounted factor correlated with both children's academic skills and preschool enrollment at a minimum of .32. Because ECLS-B had no variables that approached this correlation with both focal variables, we can have more confidence that the observed association was robust. Further, an unknown confound would have to be correlated with both parental support for learning at age 2 and children's academic skills at age 4 at .28 to reduce the observed association between the two to non-significance (ITCV = .08). Although the ITCV values were smaller for the associations between preschool enrollment and children's academic skills at age 5 (ITCV = .05) and between children's academic skills at age 5 and parental support for learning at age 5 (ITCV = .03), these results also seemed to be robust and suggest that an unknown confound would have to correlate with the variables of interest between .17 and .22 to wash out the significant associations between them that we observed in our models.

In sum, parental support for learning during early childhood was associated with academic skills that possibly selected children into preschool. In turn, that preschool experience was associated with further academic skill development that was then associated with increased parental support as children transitioned into school. These patterns, which were robust to unobserved confounds, support the plausibility that “child effects” were working as well as the proposition that preschool enrollment has implications beyond children's own skill development.

Variation by Nativity Status

The focal models (discussed above) revealed that immigrant parents were less likely to engage in learning support at age 2 compared to their native-born peers (β = −.11, p < .001), but they were more likely to engage in such behavior by the time their children entered elementary school (β = .05, p < .05). Parents’ nativity had no implications for their odds of enrolling their children in preschool or their children's academic skills. Recall, however, that our conceptual model was concerned with the potential for nativity to moderate the associations among parental support, children's skills, and preschool enrollment rather than the nativity-related differences in each. Thus, we conducted multigroup modeling by nativity status. We began with a test for measurement equivalence of the focal constructs across groups. Nested model comparisons revealed that no differences in measurement and so the latent factors loadings were held equal across immigrant and non-immigrant families in the moderation analyses discussed below.

Preliminary tests indicated that the set of focal parameters differed depending by parents’ nativity [χ2 = 16.22, p < .05], and so we tested a series of nested models to determine which ones did so. Results suggested that parental support for learning at age 2 was directly related to a greater likelihood of preschool enrollment [χ2 = 9.67, p < .01] for the children of immigrants (β = .11, p < .01, OR = 1.23) but not for other children (β = −.08, ns, OR = .92). If immigrant parents’ supportive behaviors were 1 standard deviation above the mean, they were 44% more likely to enroll their children in preschool. Moreover, the association between children's academic skills at age 5 and their parents’ support seen in the full sample was stronger for the children of immigrants (β = .35, p < .001) than for children with U.S.-born parents (β = .23, p < .001) [χ2 = 14.01, p < .001]. That is, at low-levels of academic skills (−1 standard deviation), U.S.-born parents provided greater support for their children's learning, but, at average and high levels (+1 standard deviation) of such skills, immigrant parents provided more support.

We also explored differences within the immigrant population according to their socioeconomic circumstance (proxied by parent education). In general, model results did not differ between low- and high-SES immigrant families, just as they did not differ by SES in the full sample. Considering the socioeconomic differences in immigration patterns between Latino/a and Asian Americans (Feliciano, 2005), we also conducted models comparing immigrant families according to their region of origin. Compared to Asian immigrants, Latino/a immigrants tend to be of lower SES backgrounds, but, even when taking into account these differences, the association between parental support and preschool enrollment was stronger among Latino/a immigrants (OR = 1.60, p < .001) than Asian immigrants (OR = .77, ns). Children's elicitation of parents’ support was comparable across groups, albeit slightly stronger for Latino/a immigrants (β= .38 versus .29, p < .001). No differences emerged when comparing Latino/a and Asian immigrants with varying educational histories.

In sum, these analyses revealed some immigration-related moderation in the conceptual model. First, immigrant parents’ support for children's learning was directly associated with their children being enrolled in preschool later on, an association that did not hold for U.S.-born parents. Second, children's academic skills appeared to elicit greater investments from immigrant parents than from U.S.-born parents. These processes did not vary according to immigrant families’ educational attainment, but they did vary by their country of origin.

Robustness Checks

The best test of the conceptual model would have ensured proper temporal spacing among the focal constructs posited to be in a predictor/outcome association, but that was not always possible with these data. In particular, preschool enrollment and the assessment of children's academic skills had some temporal overlap. Furthermore, children's age 5 academic skills and parents’ report of their supportive behaviors were measured at the same time point.

To explore whether such overlap posed problems, we re-tested the basic models using a latent factor of children's age 2 cognitive skills as a predictor of preschool enrollment (compared to their age 4 skills) and children's age 4 academic skills as a predictor of parenting the following year. Results (CFI = .898, RMSEA = .009, and x2 (df = 450) = 559.39, p < .001) suggested that age 2 skills were not associated with preschool enrollment, although the other parts of the model were comparable to the main findings reported above. The absence of this link between early skills and later enrolment might have reflected the two-year lag between the two assessments. Ideally, we would have had access to children's age 3 academic skills, but such information was not available. Results from these analyses do suggest, however, that more academically skilled children, even at age 4, experienced more support from parents during the subsequent year. Post-hoc tested revealed that the regression coefficients were not significantly different for children's age 4 (β = .19, p < .001) or 5 (β = .24, p < .001) academic skills when looking at changes in their parents’ educational support [χ2 = .50, ns], thereby lending confidence to our general conclusions while addressing concerns of reverse causality.

As another supplemental exploration, we re-estimated a models identical to the final model in Figure 2 with a more restricted sample that included only children who were assessed in the first two months (59% of the original sample, n = 3,700) and first month of preschool (30% of the original sample, n = 1,900). These steps were intended to address concerns regarding the temporal overlap between preschool entrance and academic skills. We argue that children who were assessed in the first month or two of school were less likely to have experienced gains in their academic skills as a result of preschool enrollment than children who were assessed later in the year. If the results held, therefore, we could be less concerned about temporal overlap. Overall, the results from these analyses (first month model: CFI = .915, RMSEA = .010, and x2 (df = 450) = 491.12, p < .001; first two months model: CFI = .928, RMSEA = .009, and x2 (df = 450) = 545.44, p < .001) were comparable to those from the main models in which we had controlled for the assessment month of school (the general approach to addressing such issues). Finally, to ensure that our assumption was valid—that children in preschool were unlikely to have experienced gains in their academic skills within the first month—we conducted models that reversed the paths from preschool enrollment to children's preschool entry academic skills (net of prior levels of cognitive abilities). Results from these analyses revealed that preschool enrollment was not associated with gains in children's academic skills in the first month of school (β = .01, ns), thereby, lending confidence to our conclusions.

Discussion

The view of children as situated within the family and school systems is not new, nor is the recognition that taking such a view can shed light on children's development (Lerner, 2006). Still, there remains much to be learned. Primarily, these contexts have often been treated as separate entities, both empirically and politically. A growing literature, however, suggests that they can act synergistically rather than independently (Bradley, McKelvey, & Whiteside-Mansell, 2011; Crosnoe et al., 2010, 2012). This study extended the literature by examining the connections between family and school systems during early childhood while also focusing on “child effects”. We also reversed some traditionally assumed associations, which revealed some important connections among parents’ support for children's learning, preschool enrollment, and children's academic skills. Finally, we incorporated population diversity by examining how macro-level systems shaped proximate transactions. The results of all of this activity bring up several topics of discussion that are relevant to future work and policy makers.

To begin, the findings of this study echo assertions from past research highlighting the potential role children can play in driving connections between the home and school (Bell, 1968; Crosnoe et al., 2012; Jaeger & Weinraub, 1990). In line with emerging developmental research (Ansari & Crosnoe; 2015; Gershoff et al., 2009; Lugo-Gil & Tamis-LeMonda, 2008), we found that the contexts of children's development and their early learning are transactional processes. Links between parenting factors and child factors not only reflect parents socializing children but also likely suggest that children can evoke investment from parents. Targeting children's academic skills through policy and intervention, therefore, could conceivably have ramifications beyond children's own development. These potential transactional relations between parents and children are particularly important given the perceptions that intra-family dynamics are more difficult to manipulate through large-scale policy initiatives.

We also found some evidence that appears to suggest that more academically skilled children could have been “selecting” themselves into formal preschool programs. Such selection—which is really about what children are evoking from adults, not what they are actively choosing themselves—has been suggested before (Abner et al., 2013) but is insufficiently documented. The early education literature has long been concerned with selection effects, but children's own prior abilities have not been examined as a selection mechanism, one that could potentially explain observed preschool effects on children's achievement. The findings from this study suggest that selection effects may go beyond the traditionally measured socioeconomic and demographic factors and that future work should pay more attention to children's own abilities as a confound in observed preschool effects. In making these claims, however, we should also point out that that evidence suggesting potential “child effects” did not hold when we spaced out children's academic skills and preschool enrollment over longer intervals. Conclusions about selection effects and child evocation, therefore, should be interpreted with caution until better data can be leveraged. Importantly, even accounting for such potential child selection processes, we still found that enrollment in formal early education programs was associated with a small increase in children's academic skills, in line with growing evidence pointing to the benefits of preschool (Magnuson & Waldfogel, 2005; Winsler et al., 2008). Even so, and similar to past work with the ECLS-B (Iruka et al., 2014; Votruba-Drzal et al., 2013), parents’ support for learning was more strongly associated with children's academic skills than preschool enrollment. Notably, these analyses only examined children's academic skills. We did not examine children's early social-behavioral skills as possible mediators, which are necessary for future life success and warrant empirical attention.

In line with the emphasis of the developmental systems perspective on the bidirectional association between systems (Lerner, 2006), this study revealed some support for reciprocal and dynamic interactions between the home and school. Parents’ support for early learning was associated with children's preschool enrollment while children's preschool enrollment was associated with parents’ later supportive behaviors, and both of these associations were related to children's academic skills. In line with emerging developmental research (Bradley et al., 2011; Crosnoe et al., 2010, 2012), these findings highlight the synergistic relations between the home and school systems and underscore the importance of taking a more holistic approach to children's development. By focusing on both contexts simultaneously over time, we can better understand the processes that shape children's early learning, above and beyond the contributions of one setting alone. Again, we stress that caution is warranted in drawing conclusions from these patterns. Despite our best efforts to address issues of temporal overlap with lagged measures of our outcome variables and sensitivity analyses, reverse causality is still a possibility. Thus, future studies need to pay careful attention to child-focused mechanisms that may link the home and school context.

Although the findings from this study suggest that there is some interplay between these contexts of children's development, they do not make clear which parents select children into preschool and, similarly, benefit more from their children's preschool enrollment and subsequent improvements in academic skills—those with initially more or less supportive behavior or those who have children with initially lower or higher skills. Future research should examine these possibilities because it might illustrate whether preschool serves as a cumulative advantage or operates in a compensatory manner, which is important for policy as it would reveal where investments have the greatest return.

As for population diversity, multi-group analyses revealed both similarities and differences by parents’ nativity. Specifically, compared to the children of U.S.-born parents, the children of immigrants experienced greater increases in parents’ support when they had better developed early academic skills, and they also were more likely to be enrolled in preschool in the context of having parental support. These findings have practical implications by revealing potentially relevant mechanisms that can serve as levers for future interventions and policy targeting immigrant populations. Similar to past research, therefore, this study supports the perspective that these micro-level processes are embedded within cultural contexts that have implications for the opportunities that children have and how parents react to and manage them (De Feyter & Winsler, 2009; Quintana et al., 2006; Suarez-Orozco & Suarez-Orozco, 2001). If we had not incorporated population diversity into our conceptual and analytical models and instead generalized the more parsimonious model to all populations, we would have overlooked key processes that were only true, or much stronger, among immigrant families.

Contrary to expectations, immigrant families’ socioeconomic circumstances did not appear to drive these patterns. They were stronger, however, in immigrants from Latin America, one of the lowest-SES migration streams in the U.S. context, but the lack of relation to family socioeconomic status suggests that these differences might have been more reflective of cultural differences. Thus, in line with recent attempts to disentangle differences among immigration, race/ethnicity, and socioeconomic status (De Feyter & Winsler, 2009; Leventhal et al., 2006), these findings reiterate the importance of considering heterogeneity within immigrant groups rather than simply focusing on between-group differences. Accounting for intra-group differences is imperative in advancing developmental theory. Our approach allowed us to understand the diversity within immigrant populations and its ramifications for the connections between the two core systems of child development. Unfortunately, ECLS-B did not have sufficient sample size to model these complex processes across the diverse origins of families, which requires continued attention.

The value of these general points of discussion that we have raised is, of course, predicated on the awareness of some limitations of what we have done, beyond those discussed so far. Primarily, although controlling for multiple family and child covariates reduces the possibility of spurious associations, it does not eliminate it. We did conduct sensitivity analyses, which suggested that our findings were fairly robust to unobserved confounds, but, without an experimental design, we cannot make strong causal inferences. Another concern is that we were not able to account for preschool quality (e.g., using the ECERS or Arnett protocols) because only a small random subsample of children had such data. One caveat to this limitation is that recent research with ECLS-B suggests that preschool quality, as measured in these data, does not mediate the associations between program type and children's early academic skills (Abner et al., 2013). Moreover, the validity of these quality measures in the ECLS-B have been put into question (Gordon et al., 2013). Furthermore, the sample of parent respondents was overwhelmingly mothers, and, consequently, we lacked information on fathers’ (or both parents’) support for children's early learning. Future work should consider this additional source of support to better understand the generalizability of our findings. Although the reliability of our parenting measures were above recommendations in the literature, they still had moderate reliability, and, therefore, the size of the documented associations are a conservative estimate. Finally, measures of children's academic skills might not have been as optimal for English language learners, although, even among a subsample of socioeconomically disadvantaged Latino/a immigrant families, we documented patterns qualitatively similar to those from the full sample.

Addressing these limitations is important to the understanding of the synergistic relations between home and educational systems in early childhood. The findings from this study do, however, build on developmental systems theory (Lerner, 2006). They provide a new conceptualization of the transactional relations among parental support for early learning, preschool enrollment, and children's academic skill formation during this critical developmental period and highlight how children are drivers of such transactions. This conceptualization has practical value in that it suggests that positive program effects for children, if intense enough, may be translated into changes in parenting and parents’ support for children's early learning.

Acknowledgments

The authors acknowledge the support of grants from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (R01 HD055359-01, PI: Robert Crosnoe; R24 HD42849, PI: Mark Hayward; T32 HD007081-35, PI: Kelly Raley). Opinions reflect those of the authors and not necessarily the opinions of the granting agencies.

Contributor Information

Arya Ansari, Population Research Center, University of Texas at Austin, 305 East 23rd Street, G1800, Austin, TX 78712.

Robert Crosnoe, Population Research Center, University of Texas at Austin, 305 East 23rd Street, G1800, Austin, TX 78712 (crosnoe@austin.utexas.edu).

References

  1. Abner KS, Gordon RA, Kaestner R, Korenman S. Does childcare quality mediate associations between type of care and development? Journal of Marriage and Family. 2013;75:1203–1217. doi: 10.1111/jomf.12055. doi: 10.1111/jomf.12055. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Ansari A, Crosnoe R. Children’s elicitation of changes in parenting during the early childhood years. Early Childhood Research Quarterly. 32:139–149. doi: 10.1016/j.ecresq.2015.03.005. (in press). doi.org/10.1016/j.ecresq.2015.03.005. Retrieved from http://dx.doi.org/10.1016/j.ecresq.2015.03.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Baron RM, Kenny DA. The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology. 1986;51:1173–1182. doi: 10.1037//0022-3514.51.6.1173. doi: 10.1037/0022-3514.51.6.1173. [DOI] [PubMed] [Google Scholar]
  4. Bayley N. Bayley Scales of Infant Development. 2nd ed. Psychological Corporation; San Antonio, TX: 1993. [Google Scholar]
  5. Bell RQ. A reinterpretation of the direction of effects in studies of socialization. Psychological Review. 1968;75:81–95. doi: 10.1037/h0025583. doi: 10.1037/h0025583. [DOI] [PubMed] [Google Scholar]
  6. Benasich AA, Brooks-Gunn J, Clewell BC. How do mothers benefit from early intervention programs? Journal of Applied Developmental Psychology. 1992;13:311–362. doi.org/10.1016/0193-3973(92)90035-G. [Google Scholar]
  7. Bradley RH, McKelvey LM, Whiteside-Mansell L. Does the quality of stimulation and support in the home environment moderate the effect of early education programs?. Child Development. 2011;82:2110–2122. doi: 10.1111/j.1467-8624.2011.01659.x. doi: 10.1111/j.1467-8624.2011.01659.x. [DOI] [PubMed] [Google Scholar]
  8. Caldwell BM, Bradley RH. Home observation for measurement of the environment. University of Arkansas at Little Rock; Little Rock: 1984. [Google Scholar]
  9. Cheadle JE. Educational investment, family context, and children’s math and reading growth from kindergarten through third grade. Sociology of Education. 2008;81:1–31. doi: 10.1177/003804070808100101. [Google Scholar]
  10. Coley RL, Votruba-Drzal E, Collins MA, Miller P. Selection into early education and care settings: Differences by developmental period. Early Childhood Research Quarterly. 2014;29:319–332. doi:10.1016/j.ecresq.2014.03.006. [Google Scholar]
  11. Crosnoe R, Ansari A, Purtell KM, Wu N. Latin American immigration, maternal education, and approaches to managing children’s schooling in the U.S. Journal of Marriage and Family. doi: 10.1111/jomf.12250. forthcoming. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Crosnoe R, Augustine JM, Huston AC. Children’s early child care and their mothers’ later involvement with schools. Child Development. 2012;83:758–772. doi: 10.1111/j.1467-8624.2011.01726.x. doi: 10.1111/j.1467-8624.2011.01726.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Crosnoe R, Leventhal T, Wirth RJ, Pierce KM, Pianta RC. Family socioeconomic status and consistent environmental stimulation in early childhood. Child Development. 2010;81:972–987. doi: 10.1111/j.1467-8624.2010.01446.x. doi: 10.1111/j.1467-8624.2010.01446.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Crosnoe R, Turley RNL. K-12 educational outcomes of immigrant youth. The Future of Children. 2011;21:129–152. doi: 10.1353/foc.2011.0008. doi: 10.1353/foc.2011.0008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Crosnoe R. Two generation strategies and involving immigrant parents in children’s education. Urban Institute; Washington, DC: 2010. [Google Scholar]
  16. De Feyter JJ, Winsler A. The early developmental competencies and school readiness of low-income, immigrant children: Influences of generation, race/ethnicity, and national origins. Early Childhood Research Quarterly. 2009;24:411–431. doi: 10.1016/j.ecresq.2009.07.004. [Google Scholar]
  17. Duncan GJ, Dowsett CJ, Claessens A, Magnuson K, Huston AC, Klebanov P, Japel C. School readiness and later achievement. Developmental Psychology. 2007;43:1428. doi: 10.1037/0012-1649.43.6.1428. doi: 10.1037/0012-1649.43.6.1428. [DOI] [PubMed] [Google Scholar]
  18. Entwisle DR, Alexander KL, Olson LS. First grade and educational attainment by age 22: A new story. American Journal of Sociology. 2005;110:1458–1502. doi: 10.1086/428444. [Google Scholar]
  19. Feliciano C. Educational selectivity in US immigration: how do immigrants compare to those left behind? Demography. 2005;42:131–152. doi: 10.1353/dem.2005.0001. doi: 10.1353/dem.2005.0001. [DOI] [PubMed] [Google Scholar]
  20. Frank KA. Impact of a confounding variable on a regression equation. Sociological Methods and Research. 2000;29:147–194. doi: 10.1177/004912410002900200. [Google Scholar]
  21. Fuller B, Holloway SD, Liang X. Family selection of child care centers: The influence of household support, ethnicity, and parental practices. Child Development. 1996;67:3320–3337. doi: 10.1111/j.1467-8624.1996.tb01916.x. [Google Scholar]
  22. Fuller B. Standardized childhood: The political and cultural struggle over early education. Stanford University Press; 2007. [Google Scholar]
  23. Gershoff ET, Aber JL, Clements M. Parent learning support and child reading ability: A cross-lagged panel analysis for developmental transactions. In: Sameroff A, editor. The transactional model of development: How children and contexts shape each other. American Psychological Association; Washington, DC: 2009. pp. 203–220. [Google Scholar]
  24. Gershoff ET, Aber JL, Raver CC, Lennon MC. Income is not enough: Incorporating material hardship into models of income associations with parenting and child development. Child Development. 2007;78:70–95. doi: 10.1111/j.1467-8624.2007.00986.x. doi: 10.1111/j.1467-8624.2007.00986.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Gordon RA, Fujimoto K, Kaestner R, Korenman S, Abner K. An assessment of the validity of the ECERS–R with implications for measures of child care quality and relations to child development. Developmental Psychology. 2013;49(1236):146–160. doi: 10.1037/a0027899. doi:10.1037/a0027899. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Hayes AF. Beyond Baron and Kenny: Statistical mediation analysis in the new millennium. Communication Monographs. 2009;76:408–420. doi:10.1080/03637750903310360. [Google Scholar]
  27. Heckman JJ. Skill formation and the economics of investing in disadvantaged children. Science. 2006;312:1900–1902. doi: 10.1126/science.1128898. doi: 10.1126/science.1128898. [DOI] [PubMed] [Google Scholar]
  28. Hernandez D. Young Hispanic children in the U.S.: A demographic portrait based on Census 2000: A report to the National Task Force on Early Childhood Education for Hispanics. National Task Force on Early Childhood Education for Hispanics; Tempe, AZ: 2006. [Google Scholar]
  29. Hu LT, Bentler PM. Fit indices in covariance structure modeling: Sensitivity to underparameterized model misspecification. Psychological Methods. 1998;3:424–453. [Google Scholar]
  30. Iruka IU, Gardner-Neblett N, Matthews JS, Winn DMC. Preschool to kindergarten transition patterns for African American boys. Early Childhood Research Quarterly. 2014;29:106–117. doi: 10.1016/j.ecresq.2013.11.004. [Google Scholar]
  31. Jaeger E, Weinraub M. Early non-maternal care and infant attachment: In search of process. In: McCartney K, editor. Child care and maternal employment: A social ecology approach. Jossey-Bass; San Francisco: 1990. pp. 71–90. [Google Scholar]
  32. Kids Count State trends in child well-being. 2013 Retrieved from: http://www.aecf.org/resources/the-2013-kids-count-data-book/ [Google Scholar]
  33. Lareau A. Unequal childhoods: Class, race, and family life. University of California Press; Berkeley, CA: 2004. [Google Scholar]
  34. Lee JS, Bowen NK. Parent involvement, cultural capital, and the achievement gap among elementary school children. American Educational Research Journal. 2006;43:193–218. [Google Scholar]
  35. Lerner RM. Developmental science, developmental systems, and contemporary theories of human development. In: Lerner RM, editor. Theoretical models of human development: Handbook of child psychology. Vol. 1. Wiley; Hoboken, NJ: 2006. pp. 1–17. [Google Scholar]
  36. Leventhal T, Xue Y, Brook-Gunn J. Immigrant differences in school-age children’s verbal trajectories: A look at four racial/ethnic groups. Child Development. 2006;77:1359–1374. doi: 10.1111/j.1467-8624.2006.00940.x. doi: 10.1111/j.1467-8624.2006.00940.x. [DOI] [PubMed] [Google Scholar]
  37. Lugo-Gil J, Tamis-LeMonda CS. Family resources and parenting quality: Links to children’s cognitive development across the first 3 years. Child Development. 2008;79:1065–1085. doi: 10.1111/j.1467-8624.2008.01176.x. doi: 10.1111/j.1467-8624.2008.01176.x. [DOI] [PubMed] [Google Scholar]
  38. Magnuson KA, Waldfogel J. Early childhood care and education: Effects on ethnic and racial gaps in school readiness. Future of Children. 2005;15:169–196. doi: 10.1353/foc.2005.0005. doi: 10.1353/foc.2005.0005. [DOI] [PubMed] [Google Scholar]
  39. Meyers MK, Jordan LP. Choice and accommodation in parental child care decisions. Community Development. 2006;37:53–70. doi: 10.1080/15575330609490207. [Google Scholar]
  40. McLanahan S. Diverging destinies: How children are faring under the second demographic transition. Demography. 2004;41:607–627. doi: 10.1353/dem.2004.0033. doi: 10.1353/dem.2004.0033. [DOI] [PubMed] [Google Scholar]
  41. Mistry RS, Benner AD, Biesanz JC, Clark SL, Howes C. 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. 2010;25:432–449. doi:10.1016/j.ecresq.2010.01.002. [Google Scholar]
  42. Muthén LK, Muthén BO. Mplus user’s guide. Seventh Edition. Muthén & Muthén.; Los Angeles, CA: 1998-2013. [Google Scholar]
  43. McCartney K, Dearing E, Taylor BA, Bub KL. Quality child care supports the achievement of low-income children: Direct and indirect pathways through caregiving and the home environment. Journal of Applied Developmental Psychology. 2007;28:411–426. doi: 10.1016/j.appdev.2007.06.010. doi:10.1016/j.appdev.2007.06.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Najarian M, Snow K, Lennon J, Kinsey S. Early childhood longitudinal study, birth cohort (ECLS-B), preschool–kindergarten 2007 psychometric report (NCES 2010-009) NCES, Institute of Education Sciences, US Department of Education.; Washington, DC: 2010. [Google Scholar]
  45. NICHD Early Child Care Research Network Duration and developmental timing of poverty and children's cognitive and social development from birth through third grade. Child Development. 2005:795–810. doi: 10.1111/j.1467-8624.2005.00878.x. doi: 10.1111/j.1467-8624.2005.00878.x. [DOI] [PubMed] [Google Scholar]
  46. NICHD Early Child Care Research Network. Duncan GJ. Modeling the impacts of child care quality on children's preschool cognitive development. Child Development. 2003;74:1454–1475. doi: 10.1111/1467-8624.00617. doi: 10.1111/1467-8624.00617. [DOI] [PubMed] [Google Scholar]
  47. Quintana SM, Aboud FE, Chao RK, Contreras‐Grau J, Cross WE, Hudley C, Vietze DL. Race, ethnicity, and culture in child development: Contemporary research and future directions. Child Development. 2006;77:1129–1141. doi: 10.1111/j.1467-8624.2006.00951.x. doi: 10.1111/j.1467-8624.2006.00951.x. [DOI] [PubMed] [Google Scholar]
  48. Suárez-Orozco C, Suárez-Orozco M. Children of immigration. Harvard University Press; Cambridge, MA: 2001. [Google Scholar]
  49. Votruba-Drzal E, Coley RL, Koury AS, Miller P. Center-based child care and cognitive skills development: Importance of timing and household resources. Journal of Educational Psychology. 2013;105:821–838. doi: 10.1037/a0032951. [Google Scholar]
  50. Winsler A, Tran H, Hartman SC, Madigan AL, Manfra L, Bleiker C. School readiness gains made by ethnically diverse children in poverty attending center-based childcare and public school pre-kindergarten programs. Early Childhood Research Quarterly. 2008;23:314–329. doi:10.1016/j.ecresq.2008.02.003. [Google Scholar]

RESOURCES