49

52.29

47.10

48.83

47.14

49.70

44.47

(9.73)

(9.14)

(9.34)

(10.04)

(9.72)

(9.30)

51.98

47.51

49.51

47.32

49.93

44.68

(9.88)

(9.09)

(9.74)

(9.64)

(9.76)

(8.91)

52.28

47.12

49.11

47.04

49.80

44.20

(9.75)

(9.03)

(9.48)

(9.89)

(9.69)

(8.84)

45.44

35.16

38.11

34.33

38.97

31.97

(21.15)

(23.92)

(23.91)

(24.93)

(24.04)

(23.23)

10,024

489

337

463

1,506

603

current research, the mean math variable score for the adolescents from intact families is

married in the original study A two-way ANOVA F (1, 25003) = .61, p < .5, indicates

that the difference between differences across the two studies is not significant.

families and children from divorced parents has also remained relatively stable. In the

Over the past ten years, the achievement disparity between children from intact

Reading

Combined Composite 1^{st }Wave Math ** Sub-sample N

*All measures are composite test scores. ELS 2002 – 2004 Data Set. N=15,362 **First follow-up N=15,325

Math

Table 1. Means and Standard Deviations for Various Family Structures In the ELS 2002 – 2004

Data Set Academic Measure* Standardized Tests

Intact Family

Cohabitation

Widowed

Separated

Divorced

Never Married

2.59 points higher than the mean for children whose parents are divorced. In the original

study, the mean math score for children from intact families was 1.79 points higher than

the score for children from divorced parents. For the math variable, the difference in

unadjusted mean scores between children from intact families and children whose parents

are divorced in the new study is .8 points higher than the difference in unadjusted mean

scores between children from intact families and children whose parents are divorced in

the original study (2.59-1.79 = .8). A two-way ANOVA F (1, 27414) = 5.02, p < .05,

indicates that the difference between differences across the two studies is significant.