Maternal and Child Health Journal

, Volume 17, Issue 9, pp 1701–1711

Early Growth of Mexican–American Children: Lagging in Preliteracy Skills but not Social Development

Authors

  • Alma D. Guerrero
    • UCLA Center for Healthier Children, Families and Communities
    • Department of PediatricsDavid Geffen School of Medicine
    • Institute of Human DevelopmentUniversity of California, Berkeley
  • Lynna Chu
    • UCLA Center for Healthier Children, Families and Communities
  • Anthony Kim
    • Institute of Human DevelopmentUniversity of California, Berkeley
  • Todd Franke
    • UCLA Center for Healthier Children, Families and Communities
    • Department of Social Welfare
  • Margaret Bridges
    • Institute of Human DevelopmentUniversity of California, Berkeley
  • Alice Kuo
    • UCLA Center for Healthier Children, Families and Communities
    • Department of PediatricsDavid Geffen School of Medicine
    • UCLA School of Public Affairs
Article

DOI: 10.1007/s10995-012-1184-7

Cite this article as:
Guerrero, A.D., Fuller, B., Chu, L. et al. Matern Child Health J (2013) 17: 1701. doi:10.1007/s10995-012-1184-7

Abstract

Latino toddlers fall behind White peers at 24 months of age in oral language and interactive skills with their mothers in English or Spanish. But Latino children enter kindergarten with social skills that rival White peers, despite social-class disparities. We ask whether cognitive trajectories widen during the 24–48 month period, how these patterns differ for Latinos, especially Mexican–Americans, and whether similar gaps in social-emotional growth appear. We analyzed growth patterns for a nationally representative birth sample (n = 4,690) drawn in 2001, estimating levels of change in development from 24 to 48 months of age, focusing on Latino subgroups. The mean gap in cognitive processing for Mexican–American children, already wide at 24-months of age relative to Whites (three-fourths of a standard deviation), remained constant at 48 months. But differences in social-emotional status were statistically insignificant at both 24 and 48 months. Mexican–American mothers were observed to be equally warm and supportive relative to White peers during interaction tasks. Yet the former group engaged less frequently in cognitive facilitation, oral language, and preliteracy activities in the home. Growth in both cognitive and social domains was considerably lower in larger families, placing children raised in poor or Spanish-speaking homes within a large household at greater risk of delays. Pediatricians and practitioners must carefully gauge the social-emotional well-being of Latino children, in developmental surveillance activities. Growth in cognitive and social domains unfolds independently for children of Mexican heritage, even when raised in economically poor families.

Keywords

Latino childrenCognitive and social developmentClinical assessment

Introduction

Recent work reveals that Latino children display gaps in basic cognitive and communicative skills, relative to White peers, as early as 2 or 3 years of age [1]. These disparities, not discernible at 9–12 months of age, appear to stem from lower maternal education, weaker early reading practices, larger family size, and less exposure to quality preschool [2]. At the same time, Mexican–American children display robust social skills as they enter kindergarten, and even when raised in low-income families, as rated by teachers [3, 4]. This suggests that health practitioners and early educators should be careful to differentiate the levels and sources of Latino toddlers’ cognitive versus social-emotional growth, as well as risks to physical health.

Yet little is known about whether cognitive disparities stabilize or widen during the preschool years, from about 2 to 4 years of age. Two studies have shown that cognitive disparities appear quite early in the lives of Latino children, at least with regard to oral language and exposure to written symbols, each drawing from the federal Early Childhood Longitudinal Study (ECLS). The first found no discernible gaps between Latino and White infants at 12–18 months of age in terms of infants’ responsiveness to mothers and early communicative skills in observed laboratory tasks [1]. But by 24–36 months of age these same children showed disparities up to two-thirds of a standard deviation in productive language, word recognition, and communicative skills with their mothers, gauged through direct assessments and observed interaction tasks conducted in English or Spanish.

The second study, drawing from the ECLS kindergarten cohort, found that Mexican–American children, especially those of immigrant parents, displayed equally wide disparities in preliteracy and oral language skills, compared with Whites. Yet this study found only slight gaps in social-emotional well-being at entry to kindergarten [4]. The comparatively strong social development of Mexican–American children was verified in a third study, while modest yet significant disparities were apparent for kindergarten-age children from poor Latino families [3].

Much is known about home and parenting factors that advance young children’s early cognitive growth among majority populations [1, 5, 6]. But a paucity of research has asked whether developmental pathways may differ for children of Mexican heritage or other Latino subgroups [79]. This constrains how pediatricians and practitioners gauge development, along with the family-linked sources of developmental outcomes among Latino subgroups.

In addition much of this early research focuses on young children’s cognitive growth, while failing to attend to social-emotional growth, a domain that bridges over to child health, mental health, and obesity risks [3, 6, 10]. Nor do we understand how cognitive or social-developmental gaps may emerge differently within homes and neighborhoods among Latino subgroups, including but not limited to Mexican-heritage children and families [11]. A strong focus on social-emotional growth may also point to protective factors or cultural assets from which practitioners can scaffold clinical and preschool interventions.

Building from recent findings appearing in this journal and Pediatrics [1, 2], we examine whether cognitive gaps widen among Latino subgroups and White peers; how disparities in social-emotional growth may open-up from 24 to 48 months of age; and identify factors that help to explain differing growth trajectories over this two-year period. These causal processes may involve prior child attributes (e.g., birth and health factors), parenting practices, and features of household structures that shape cognitive and social development during the preschool years.

We hypothesized, given this recent work that disparities in cognitive and preliteracy skills would persist or widen over the 24–48 month preschool period. At the same time, we hypothesized that gaps in social-emotional development would not widen during this period. The possible determinants of preschoolers’ development are drawn from prior work pertaining to all children or Latino subgroups, and we distinguish between determinants of, and social processes that may influence, growth in the domains of cognitive versus social-emotional domains of growth [3, 4, 11].1

Methods

Population

We draw from a nationally representative sample of births in 2001, the ECLS birth cohort, fielded by the National Center for Education Statistics (NCES) with support from the National Institutes of Health [13, 14]. Home visits were conducted, along with videotaped interaction tasks and direct child assessments when participating children were about 9, 24, and 48 months of age. We restricted our analytic sample to children who lived with their biological mother at 48 months, did not have any serious birth defects, participated in videotaped interaction tasks with their mother, and had not yet entered kindergarten by the 48-month home visit.

Of the 7,750 resulting cases with descriptive data on individual variables, 4,700 cases had complete data for multivariate estimation of cognitive processing and social-emotional growth from 24 to 48 months of age (Table 1), rounding to the nearest 50 under NCES reporting rules). Those cases with missing values for variables included in regression estimates were more likely to include mothers with lower levels of school attainment and situated in Spanish-speaking homes, compared with the 4,700 cases with complete data for all variables.
Table 1

Child and family samples for descriptive between-group analysis and explanatory models (means and standard deviations reported)

Demographic attributes at 24-month wave unless otherwise specified

Full sample for between-group descriptive analysis (n = 7,750)

Reduced sample with complete data for multivariate models (n = 4,700)

Missing data (n = 3050)

t statistic

Mother’s age (years)

32.4 (6.3)

32.4 (6.3)

32.3 (6.4)

−1.00

Child age (months at 48-months wave)

52.9 (4.2)

52.7 (4.0)

53.1 (4.4)

4.19***

Mother’s school attainment (%)

 Less than high school

22.9 %

18.9 %

28.8 %

10.26***

 High school diploma

22.7 %

22.9 %

22.5 %

−0.46

 Some college

24.4 %

26.0 %

22.0 %

−3.99***

 Bachelor’s degree or more

30.0 %

32.2 %

26.7 %

−5.19***

Father resident in the home (%)

81.1 %

81.4 %

80.6 %

−0.98

Dominant home language, non-English (%)

21.0 %

14.1 %

31.6 %

−18.91***

p < 0.05; ** p < 0.01; *** p < 0.001, October 15, 2012

Measures

Child Outcomes

The 48-month preliteracy battery, conducted directly with each child, covered oral language production, letter and word recognition. The math assessment at 48 months included items on number sense, counting, operations, geometry, and pattern comprehension [14]. We employed the Bayley mental scales for our 24-month control, which assesses early cognitive and language ability as manifest in memory, expressive and receptive vocabulary, reasoning and problem solving, and concept attainment. The 9-month version of the Bayley’s was also administered during this first home visit, and scores are included in our regression models to conservatively control on prior cognitive functioning.

All fieldwork, including the direct child assessments, was conducted in the mother’s home language at the 24- and 48-month data waves. One significant exception is the 48-month preliteracy battery, which was given only in English. An earlier paper details how the 24-month cognitive assessment is highly predictive of the 48-month battery [15], and our estimation models control on the former when estimating change scores between the two data waves. We utilized standardized scores derived from item-response theory (IRT), which adjusts for any systematic bias in scoring by field researchers.

A social-emotional index score was constructed from 21 parent-reported measures of specific behaviors at 48 months. Items include positive (i.e., child shares, pays attention, is eager) and negative (i.e., child has a bad temper, is unhappy, becomes angry) behaviors. Parents were asked to consider the child’s behavior in the last 3 months; responses were coded on a 5-point Likert scale (1, never to 5, very often). Negative behavior items were reverse-coded so that a higher score would indicate a more positive response. At 24 months a related index was constructed as a control variable using items from the Child Behavior Rating Scale, which includes items corresponding to the 48-month instrument [13].

Predictors of Cognitive and Social-Emotional Status and Growth

Predictors of change in developmental outcomes, 24–48 months of age, fell into four domains: (1) basic family and maternal demographic characteristics, (2) maternal relationships, support, and mental health, (3) parenting practices and discipline, along with preliteracy activities, and (4) indicators of the mother’s acculturation status for Latina subgroups.

Demographic characteristics included the focal child’s gender and age in months, the mother’s age and employment status. Within the domain of maternal relationships, support, and mental health, eight variables were included: whether the father was resident in the home at 48 months, the ratio of children to adults in the home, mother’s closeness to her own mother (coded as a binary variable, defined as “close” if mother responded “close” or “very close” to own mother on the interview question), mother’s efficacy in raising the focal child (coded as “very difficult”, “average difficult”, or “less than average difficult”), mother’s mental health and depressive symptoms (from a standardized index score of 12 related items—i.e., “in the past week, how often… were you sad?” [13]), mother’s education level (coded into four categories: less than high school diploma, diploma, some college, 4-year degree or more), and household income (coded as a binary variable, greater or less than $50,000). All variables in this domain were assessed at the 24-month home visit to ensure that they were operating causally prior to child outcomes measured at 48 months.

We focused on 10 key factors pertaining to parental discipline and preliteracy practices that may affect the early growth of all young children, or may affect children of Latino heritage: frequency of punishment (measured as a binary variable that dichotomized a parent’s reported ways of responding to misbehavior), a harsh-discipline index (a composite score developed from the parent’s willingness to spank, hit, or yell at the child to exert control), frequency of reading to child (coded as less than 2 times a week, 3–6 times a week, or every day), frequency of library visits (coded as a binary variable with “1” indicating a library visit in the last month), the 9-month NCATS teaching-task score (measuring the extent to which parents praised or encouraged their child, responded to their child’s distress, expressed warm affect, and provided cognitive fostering), which stemmed from an observed mother–child interaction task, and positive communications and negative utterances observed via the videotaping of the “two bags task”, where the mother explains simple tasks to the child and encourages successful completion [13, 14].

We conducted a principal components analysis from the coded interaction and mother-affect items, scored by NCES, and derived two dimensions: maternal communications that encouraged the child to think through and complete a task (“cognitive facilitation”) and the presence of negative or frustrated utterances by the mother when the child failed to perform the task.

To assess possible socio-cultural differences among Latinos subgroups, we gauged the mother’s acculturation status in terms of primary language spoken in the home (Spanish, English, or other language), and whether the mother was foreign or native born.

Statistical Analyses

To determine levels of growth during the preschool years, weighted least-squares regressions were used to estimate children’s preliteracy, math concepts, and social-emotional scores at 48 months, taking into account the level at 24 months. That is, we estimate change in (IRT) scores pertaining to cognitive and social-emotional outcomes. We incorporated survey weights to account for oversampling of certain subgroups in the data sample and to approximate a nationally representative population. We then conducted the same analysis for Latino children only, again estimating levels of growth in both domains, examining how (1) basic family and maternal demographic factors, (2) maternal relationships, support, and mental health, and (3) parenting practices and discipline and mother–child interaction, along with preliteracy activities, help to explain growth for Mexican–American children, compared with other Latino subgroups.

Results

Differing Cognitive and Social-Emotional Trajectories

We begin by describing basic growth developmental patterns from 24 to 48 months of age (Table 2). The gap in Bayley mental scale (IRT) scores between Mexican–American and White children at 24 months is wide, 46.9 versus 53.4, respectively, or 0.77 standard deviation (SD). This disparity remains steady at 0.79 SD by 48 months of age for the cognitive and preliteracy-skills scores. This disparity is discernibly narrower for all other Latinos. The corresponding gap in math-concepts scores for Mexican–American children vis-à-vis White peers equals 0.68 at 48 months of age.
Table 2

Mean cognitive processing, preliteracy, and social-emotional scores at age 24 and 48 months (n = 4,700, standardized t scores and standard deviations reported)

 

24 months

 

48 months

IRT scale scores

SD

F value for mean differences (p value)

 

IRT scale scores

SD

F value for mean differences

(p value)

Bayley mental processing

   

Preliteracy

   

 Mexican–American

46.9

8.1

74.95 (0.000)

 Mexican–American

20.7

7.4

62.44 (0.000)

   

 Other Latinos

23.6

8.0

 Other Latinos

47.1

8.6

 White

27.0

8.9

   

Early math

   

 White

53.4

8.6

 Mexican–American

26.3

7.4

33.32 (0.000)

    

 Other Latinos

28.6

7.8

    

 White

31.1

8.3

Social-emotional

   

Social-emotional

   

 Mexican–American

40.1

7.2

6.15 (0.003)

 Mexican–American

80.2

7.9

0.83 (0.441)

 Other Latinos

38.6

7.5

 Other Latinos

81.4

7.9

 White

41.2

7.1

 White

81.0

7.6

October 6, 2012

Table 2 also shows the parent’s rating of their child’s social and emotional behavior. Mexican–American and White parents show quite similar ratings, and mean levels are statistically insignificant at both 24 and 48 months of age.2 This is consistent with prior work showing comparatively strong social development among Mexican–American children, even when raised in low-income families, relative to Whites [3, 4].

Table 3 then reports descriptive differences among Mexican–Americans, other Latinos, and Whites for the attributes or maternal practices that may help to explain levels of growth from 24 to 48 months of age. For example, Mexican–American mothers were significantly younger than other groups. Birth fathers were less frequently residing in the homes of Mexican–American mothers, compared with Whites, while the ratio of children to adults in the home was generally similar across all ethnic groups, suggesting the presence of additional resident adults in many Latino homes.
Table 3

Between-group differences in preliteracy, math concepts, and social-emotional predictors at 9-, 24-, and 48 months (weighted means and standard deviations reported; significance tests relative to whites; for descriptives, n = 7,750, prior to sample reduction for multivariate estimation)

 

Ethnic groups

Latino childrena

Statistical differences viz. Whitesb

White

African American

Asian

Mexican–American

Other Latinos

Spanish speaking

English speaking

Mexican–American

Other Latinos

Latinos

4,950

1,250

250

1,000

500

900

600

Spanish speaking

English speaking

Demographics

Female (%)

50

47

49

48

45

45

52

−0.66

−1.57

−1.39

0.31

Child age (months at 48 months wave)

52.0 [3.5]

51.9 [4.4]

52.8 [7.5]

53.1 [3.5]

53.1 [3.5]

53.6 [3.4]

52.5 [4.1]

3.62***

3.32***

5.19***

1.39

Maternal relationships, support, mental health (at 24 months)

Mother’s age (years)

33.0 [5.3]

29.8 [6.4]

33.5 [1.1]

30.2 [5.2]

31.6 [5.1]

31.0 [4.9]

30.2 [5.5]

−6.53***

−3.37***

−5.72***

−5.63***

Father resident in home (%)

90

46

88

81

75

84

72

−4.08***

−4.02***

−2.82**

−5.94***

Mother employed (%)

 Full-time

33

43

38

35

43

37

41

0.67

2.15*

0.98

2.41*

 Part-time

26

17

16

14

19

15

14

−5.16***

−2.82**

−4.04***

−4.61***

Ratio of children: adult resident in the home

1.1 [0.5]

1.5 [1.0]

1.0 [1.1]

1.1 [0.6]

1.1 [0.8]

1.0 [0.5]

1.3 [0.9]

1.22

0.12

−1.56

2.16*

Mother’s closeness to own mother (% close or very close)

83

84

80

75

76

75

75

−2.75**

−1.91*

−2.57**

−2.47*

Mother’s efficacy in raising child (%)

 Very difficult

9

9

10

14

14

15

12

2.35*

1.46

2.40*

1.38

 Average difficult

29

35

29

27

27

24

31

−0.76

−0.44

−1.85

0.46

Mental health, depressive symptoms

 CESD score

0.3 [1.0]

0.2 [1.0]

0.2 [2.1]

0.3 [1.0]

0.1 [0.6]

0.2 [0.7]

0.3 [1.2]

−0.06

−2.44*

−1.77

0.61

Income ≥50,000 (%)

57

17

61

12

30

12

26

−18.34***

−6.63***

−17.89***

−7.95***

Mother’s education

 <High school

13

27

13

53

28

52

33

12.18***

3.65***

11.24***

5.56***

 High school

22

34

20

21

31

25

24

−0.74

2.41*

0.84

0.50

 Some college

28

27

17

19

24

14

31

−3.42***

−0.95

−5.72***

0.58

 Bachelor’s degree or more

37

12

50

7

17

9

12

−15.07***

−6.89***

−13.70***

−9.98***

Acculturation indicators (24 months)

Dominant language spoken at home (%)

 English speaking

99

97

63

39

39

 Spanish speaking

0.1

0.08

0

60

55

 Other

1

3

37

1

5

Mother, foreign born (%)

3

9

60

54

60

82

17

13.03***

9.59***

31.32***

4.12***

Parental discipline and preliteracy practices (at 24 months)

Mother punishes misbehavior (%)

58

64

57

56

51

57

52

−0.75

−1.87

−0.56

−1.60

Harsh discipline index score (1–3)

0.6 [0.7]

0.9 [0.8]

0.5 [1.3]

0.4 [0.6]

0.5 [0.6]

0.3 [0.5]

0.5 [0.6]

−3.75***

−1.80*

−4.27***

−1.13

Reading (%)

 <2 times/week

14

42

21

48

47

57

35

10.75***

8.03***

13.13***

5.72***

 3–6 times/week

24

33

33

28

26

26

29

1.27

0.46

0.54

1.23

 Everyday

62

26

46

24

27

18

36

−13.34***

−7.66***

−17.26***

−5.78***

Visit library (%)

35

29

40

26

21

20

33

−3.62***

−3.51**

−5.92***

−0.57

9-month NCATS teaching-task score

 Praises effort, encouragement

2.2 [1.8]

1.7 [1.8]

1.8 [3.4]

1.7 [1.5]

1.9 [1.6]

1.7 [1.4]

1.9 [1.8]

−3.62***

−1.57

−4.02***

−1.73

 Responds to child distress

4.0 [1.7]

3.8 [1.7]

4.2 [3.0]

3.8 [1.5]

3.8 [1.4]

3.7 [1.4]

4.0 [1.5]

−1.08

−0.89

−2.45*

−0.10

 Warm affect, emotional support

2.3 [1.0]

2.2 [1.1]

2.2 [2.1]

2.2 [1.0]

2.2 [1.0]

2.2 [1.0]

2.2 [1.0]

−1.80

−0.43

−1.54

−0.52

 Cognitive fostering, verbal

2.7 [0.5]

2.6 [0.7]

2.7 [1.4]

2.6 [0.7]

2.7 [0.6]

2.6 [0.6]

2.7 [0.7]

−1.87

−0.20

−1.63

−0.60

9-month two-bag

 Two bag cognitive

0.4 [1.5]

−0.8 [2.0]

0.2 [3.1]

−0.3 [1.3]

−0.2 [1.5]

−0.6 [1.4]

−0.1 [1.2]

−7.15***

−3.21**

−7.04***

−2.36*

 Two bag negative

0.2 [1.1]

0.1 [1.5]

0.04 [2.3]

−0.5 [1.1]

−0.2 [1.0]

−0.7 [1.0]

0.06 [1.2]

−7.98***

−3.00**

13.15***

−1.40

aNon-English, non-Spanish mother are excluded

bReporting t values * p < 0.05; ** p < 0.01, *** p < 0.0001

October 5, 2012

About one-third of Mexican–American mothers worked full-time, which is equivalent to Whites, but a greater percentage (43 %) of non-Mexican Latina mothers worked full-time. Both Mexican–American and other Latina mothers were much less likely to work part-time, when compared with Whites. About four-fifths of ‘other Latinos’ were originally from Central or South America. Almost three-quarters (74 %) of Mexican–American mothers had not gone beyond high school, compared with just 35 % of White mothers.

Home Environs, Maternal Practices

The practices of Mexican–American mothers differ from their White peers on several dimensions. Observed within the videotaped interaction tasks, White mothers articulated praise for the child’s efforts and engaged in cognitive fostering at significantly higher rates. Children in Mexican–American and other Latino homes read with their mothers much less frequently. Nearly half of Mexican–American and other Latino children were read to once a week or less.

Mexican–American mothers reported less strict or harsh parenting than Whites on average. We return to this pattern to explain the relatively robust social-emotional status of the former group. We saw no evidence of weaker mental health among Mexican–American mothers, compared with White peers, and other Latinas reported lower incidence of depressive symptoms than Whites. At the same time, a smaller share of Mexican–American mothers reported feeling close to their own mother, and they reported lower efficacy in raising their child, compared with White mothers.

Estimating Change in Child Outcomes, 24–48 Months of Age

Turning to the multivariate estimates of growth (Table 4), we first control for earlier levels of the child’s cognitive proficiency at 9 and 24 months, and we control for the age at which the child was assessed in the home. The 24-month cognitive proficiency scores, not surprisingly, helped significantly in predicting 48-month preliteracy and math concepts scores. These controls yield a model that estimates levels of change in child outcomes at 24 and 48 months of age.
Table 4

Ordinary least-squares regression estimation of children’s preliteracy, math concepts, and social-emotional scores at 48-months of age, controlling for 24-month status (unstandardized beta coefficients and standard errors reported)

n = 4,700

Preliteracy

Math concepts

Social-emotional

Estimate

SE

Estimate

SE

Estimate

SE

Earlier child status (controls)

 24 months child behavior rating scale

0.06**

0.02

 24 months Bayley t score

0.24***

0.02

0.24***

0.02

 9 months Bayley t score

0.02

0.02

0.02

0.02

Maternal and family demographics

 48 months child age

0.80***

0.04

0.94***

0.04

0.19***

0.04

 48 months mother age

0.08*

0.04

0.08*

0.03

−0.06

0.03

 24 months resident father

−0.72

0.54

−0.39

0.51

−0.16

0.51

 Female

0.77*

0.32

0.08

0.3

3.18***

0.3

 African–American

1.25**

0.46

0.47

0.47

1.56**

0.56

 Asian

2.97***

0.9

1.61*

0.8

−1.42*

0.67

 Mexican–American

−0.66

0.84

−1.53*

0.7

1.40

0.89

 Other Latinos

0.6

0.85

−0.84

0.71

2.34**

0.73

 24 months mom work full-time

−0.69

0.36

−0.25

0.33

0.49

0.43

 24 months mom work part-time

−0.34

0.43

0.14

0.45

0.08

0.39

 24 months income >$50,000

2.00***

0.35

1.74***

0.36

−0.04

0.42

 24 months mom edu high school

1.31**

0.5

1.33**

0.44

0.13

0.48

 24 months mom edu college

2.00***

0.44

2.31***

0.43

1.44**

0.53

 24 months mom edu > bachelor

4.60***

0.55

3.99***

0.55

1.04

0.56

 Spanish speaking

−1.97*

0.85

0.08

0.71

−1.15

1.16

 Other language speaking

0.41

1.08

1.54

0.95

0.37

0.99

 Mother foreign born

0.45

0.74

1.21*

0.62

0.20

0.66

Family structure and childrearing efficacy

 24 months kid/adult ratio

−1.32***

0.26

−1.10***

0.26

−0.71**

0.23

 9 months mother to own mother

−0.29

0.43

0.22

0.35

1.02**

0.38

 24 months child difficult raise

−0.58

0.42

−0.81

0.49

−4.42***

0.65

 24 months child avg difficult

0.07

0.34

−0.10

0.31

−2.76***

0.31

 24 months mother depression

−0.08

0.13

−0.16

0.12

−0.51***

0.13

Maternal practices and mother–child interaction

 24 months punish misbehav.

0.03

0.34

0.07

0.26

0.78**

0.3

 24 months harsh discipline

−0.3

0.2

−0.14

0.23

−0.79***

0.22

 9 months praise effort (NCATS)

0.03

0.09

0.002

0.09

0.13*

0.07

 9 months distress (NCATS)

0.005

0.09

0.15

0.08

0.04

0.08

 9 months warm affect (NCATS)

0.002

0.15

−0.07

0.14

0.07

0.17

 9 months cognitive (NCATS)

0.44*

0.22

0.32

0.20

0.6*

0.27

 24 months reading 3–6 times/week

0.27

0.40

0.39

0.44

1.26**

0.46

 24 reading everyday

2.27***

0.38

1.21***

0.36

1.93***

0.39

 24 months visit library

0.92**

0.38

0.48

0.29

0.53

0.38

 Two-bag task, cognitive

0.44***

0.1

0.60***

0.09

0.57***

0.1

 Two-bag task, negative affect

0.05

0.14

-0.17

0.1

-0.05

0.12

 Intercept

−35.28

2.66

−38.70

2.41

65.71

2.97

p value < 0.05; ** p value < 0.01; *** p value < 0.001, October 8, 2012

Basic demographic features of mothers and homes help to account for levels of growth. For example, children with older mothers display higher levels of growth over the 24–48 period. Maternal education is a strong factor in accounting for growth in both cognitive and social-emotional domains, especially for preliteracy and math growth. Maternal employment was not a significant factor, regardless of whether mothers worked part- or full-time.

Table 4 also shows that neither Mexican nor other Latino children fall further behind White peers in basic cognitive and preliteracy skills. Mexican–American children do show significantly lower math-concepts knowledge at 48 months relative to Whites, again controlling for their 24-month Bayley’s cognitive score. Yet remember that these math-concepts scales do not gauge the same constructs tapped by the Bayley’s. We also see that the social-emotional status of other-Latino children is higher at 48 months, compared with White peers, after controlling for 24-month status. This is notable, given that other-Latino children still lag significantly behind Whites on cognitive scales, while their rate of social growth is surpassing White children.

Explaining Variability in Growth Among Latino Children

We then examined the extent to which the same sets of predictors might account for growth only among Latino children (Table 5). For this subsample we again see the influence of maternal education on growth in cognitive skills and social-emotional development. Mexican–American mothers, remember, display lower levels of school attainment relative to other Latinas. Growth in cognitive and preliteracy skills was slower in Spanish-speaking households. Together, lower maternal education and a greater propensity to speak Spanish in the home, were associated with lower growth in the cognitive skills of Mexican-heritage children.
Table 5

Latinos only—weighted least-squares regressions estimating preliteracy, math concepts, and social-emotional scores at 48-months, controlling for 24-month status (unstandardized beta coefficients and standard errors reported)

N = 700

Preliteracy

Math concepts

Social-emotional

Estimate

SE

Estimate

SE

Estimate

SE

Earlier child status (controls)

      

 24 months child behavior rating scale

0.08

0.04

 24 months Bayley t score

0.12**

0.04

0.145***

0.05

 9 months Bayley t score

0.01

0.04

0.02

0.03

Maternal and family demographics

      

 48 months child age

0.65***

0.10

0.92***

0.08

0.16

0.10

 48 months mother age

0.17**

0.07

0.11

0.06

−0.08

0.08

 24 months resident father

−1.10

0.67

−1.04

0.74

−0.22

1.17

 Female

1.02

0.67

0.78

0.64

2.44**

0.86

 Other Latinos (Mexican–American, reference group)

1.36

0.80

0.77

0.68

0.96

1.11

 24 months mom work full-time

−1.35

0.79

−0.60

0.67

0.08

0.9

 24 months mom work part-time

−0.48

1.15

0.47

1.15

1.02

1.37

 24 months income >50000

0.93

0.9

0.93

1.08

−2.06

1.38

 24 months mom edu high school

2.16**

0.82

1.75*

0.85

0.36

1.03

 24 months mom edu college

1.97

1.01

2.50*

1.02

1.90

1.14

 24 months mom edu > bachelor

4.14*

1.71

2.82*

1.41

2.81**

1.28

 Spanish speaking

−2.58**

0.99

−0.49

0.88

−1.34

1.46

 Other speaking

−3.07

1.93

−0.67

2.15

0.63

1.95

 Mother foreign born

0.26

0.92

1.56

0.86

0.93

1.22

Family structure and childrearing efficacy

      

 24 months child/adult ratio

−0.97

0.54

−1.01

0.56

0.15

0.50

 9 months relations, own mother

−0.47

0.90

−0.39

0.76

0.3

0.96

 24 months child difficult raise

−0.24

1.06

−0.37

1.11

−2.18

1.31

 24 months child avg difficult

0.28

0.77

−0.05

0.70

−0.51

0.87

 24 months mother depression

−0.36

0.33

−0.43

0.30

−0.68***

0.21

Maternal practices and mother–child interaction

      

 24 months punish misbehav.

−0.63

0.73

−0.06

0.63

1.24

0.87

 24 months months harsh discipline

0.95

0.61

0.67

0.52

−1.52*

0.73

 9 months praise effort (NCATS)

−0.08

0.22

0.21

0.21

0.18

0.22

 9 months distress (NCATS)

−0.25

0.22

0.02

0.21

0.03

0.22

 9 months warm affect (NCATS)

0.06

0.3

−0.04

0.35

0.44

0.32

 9 months cognitive (NCATS)

0.26

0.49

0.64

0.47

0.77

0.57

 24 months reading 3–6 times/week

0.14

0.93

0.88

0.80

1.54

0.91

 24 reading everyday

0.93

0.85

0.97

0.95

1.88

1.11

 24 months visit library

0.70

0.86

0.81

0.88

0.01

1.04

 Two bag cognitive

0.66*

0.29

0.69**

0.26

0.61*

0.27

 Two bag negative

0.14

0.31

−0.2

0.34

0.19

0.39

 Intercept

−22.05***

7.23

−36.08***

6.43

65.77***

7.71

p value < 0.05; ** p value < 0.01; *** p value < 0.001

Again, we see that specific maternal practices and mental health exerted significant effects on the child’s social-emotional growth. Mothers reporting significant depressive symptoms acted to suppress children’s social development, as did mothers that exercised harsh discipline. On the other hand, children demonstrated stronger rates of cognitive and social-emotional growth when their mothers engaged in higher levels of cognitive facilitation as observed in the videotaped interaction tasks.

Discussion

These results further clarify how cognitive and social-emotional growth unfolds for diverse Latino preschoolers. Practitioners and researchers often assume that similar causal pathways yield similar effects in both developmental domains, and the role of poverty and social class operates in identical ways across cultural groups. This harks back to the postulate regarding an all encompassing “culture of poverty”, allegedly shared across the upbringing of children in all ethnic or linguistic groups [9, 10]. Instead, our findings help to distinguish particular developmental pathways that vary between cognitive and social-emotional domains among Latino children.

Many low-income Latino parents, especially Mexican–American families, appear to raise children who display weak cognitive and language skills, yet comparatively strong social competencies [4]. The cognitive-skills gap between Mexican–American and White children did not widen from 24 to 48 months of age—but it remained constant and wide at about three-quarters of a standard deviation. This chasm in early development will disadvantage many Mexican–American children in their early years of schooling, their peer relations and health risks downstream [3, 5, 8].

At the same time, low-income Mexican–American families manifest weaker maternal education levels, larger family size, lower rates of maternal employment, less frequent reading with toddlers and preschoolers, and higher levels of maternal depression, compared with White peers [1, 5]. These home environments often yield children with slower rates of language and cognitive development. Our findings extend previous studies and detail how the cognitive growth for Latino children continues to lag behind Whites at 24 and 48 months, sustaining the earlier disparity reported for children at about 24 months of age [2, 3].

Yet we also found that the social-emotional growth of Latino children continues to rival levels observed for White peers at 24 and 48 months of age, even for Mexican–American children raised in lower-income families. More work is required to understand how materially poor Latino parents nurture socially adept preschoolers. One recent study details how Latina mothers are quite direct in their controlling utterances to young children, while set against a social backdrop of caring and frequent affection [16]. Latina mothers tend to show rich and highly responsive behavior toward toddlers on average, except for those mothers suffering from depression or facing severe economic hardship [8]. Other culturally specific practices, which we were unable to observe in the present study, may further advance the strong social-emotional growth of Latino preschoolers.

Our results illuminate differing developmental pathways, along with corresponding maternal and home practices, that help to explain children’s cognitive or social development—causal structures that at times differ for Mexican and other Latino families vis-à-vis White peers. And remember that cognitive gaps were widest for Mexican–American preschoolers, relative to White peers, while other Latino children fared better. This suggests that risk and protective factors vary in kind and level among differing Latino subgroups. We observed that Mexican–American mothers are less strict or harsh with their children, compared with White mothers. At the same time, Mexican–American mothers issued significantly less praise and cognitive facilitation with their child, while White and other Latino mothers showed no significant differences.

Mexican–American mothers also reported lower efficacy in raising a child who they saw as more difficult, on average, compared with other Latina mothers, whose perceptions related to child-rearing efficacy were indistinguishable from White peers. The mother’s level of depressive symptoms and harsh discipline style did suppress children’s social-emotional growth, a reminder that poverty and economic stress, most common among Mexican–American families, continue to undercut early child development.

Conclusions

Many health practitioners and early educators try to gauge and remedy cognitive or language delays that poor children in particular may display (whether expressed in Spanish or English). Our findings confirm that these gaps emerge early in life for Mexican–American toddlers and persist through 48 months of age. Such disparities are less pronounced for other Latino subgroups but do reach statistical significance vis-à-vis White peers. At the same time, even Mexican–American children from low-income families show robust social and emotional development, on average, likely rooted in warm and supportive households [11]. So, as child-health providers and preschool teachers, among other practitioners, observe Latino children’s early growth they should not assume social-emotional delays, even when language or cognitive skills lag somewhat behind. Pediatricians and practitioners might also take particular note of any social and emotional delays identified in Latino children and ensure appropriate follow-up evaluations, given their otherwise robust developmental outcomes in this domain on average.

Future research should address limitations of this study. Stronger measurement would enrich our understanding of how Latino children grow within particular cultural contexts. Assessments of deep cognitive, oral language, and preliteracy skills could be conducted in English and Spanish when English is not the dominant home language. And conventional measures of Latino children’s social skills, as those used in the ECLS, do not necessarily tap into culturally valued behaviors. Our research team recently validated the new Mexican–American Social Scales, built from earlier qualitative work inside homes [17]. But designers of national longitudinal studies typically rely on measures developed from majority populations. Much remains to be learned regarding how clinical practitioners and early educators measure and identify factors promoting the social-emotional growth of Latino children. Work on the effectiveness of preschools and home visiting programs offer important pieces to the contextual puzzle in which young Latino children are raised [18].

These findings suggest that practitioners might better encourage mothers to engage in practices that would close early gaps in basic cognitive and preliteracy skills—frequent engagement in rich oral language, solving problems together thorough supportive communication, reading together, and exposing children to written text and stories. Practitioners should also be more attuned to recommending these practices if home environs of Latino youngsters include low levels of maternal education, weaker cognitive facilitation, and larger family size, factors that we show slow cognitive growth. At the same time, practitioners and researchers alike should not conflate the family practices and factors that shape social and emotional growth—for these causal paths differ from the ways in which parents variably foster young children’s cognitive growth, each avenue unfolding in particular cultural contexts.

Footnotes
1

Culturally oriented developmental psychologists at times emphasize the interwoven nature of cognitive learning and the social context and utility of what’s learned by young children. That is, by separating the two domains, we are distracted by their interplay in how youngsters acquire knowledge that serves culturally appropriate social behaviors [12].

 
2

We examined correlations between parent and preschool-teacher reports of the focal child’s social behavior and emotional status for a second, smaller sample of children who had entered preschool. These paired correlations were moderate in magnitude (Pearson r’s equal between 0.20 and 0.35, depending on paired items). Focusing on parent reports allowed us to examine change over the 24- to 48-month period.

 

Acknowledgments

This study was funded in part by grant R40 MC 21517 to Dr. Guerrero from the U.S. Department of Health and Human Services, Health Resources and Services Administration, Maternal and Child Health Research Program. Berkeley team members were funded by this grant and the Institute of Human Development. Dissemination activities are supported by the McCormick Foundation.

Copyright information

© Springer Science+Business Media New York 2012