55

ure PI1-Family Struct

.08***

.14***

.12***

.18***

Missing

-.01

.04

.01

.03***

PI2-Checking-Up

-.03***

-.05***

-.04***

.01

PI3-Discussion

.16***

.14***

.16***

.10

PI4-Attendance

.04***

.04***

.05***

.07***

SES Quartile 2

.11***

.10***

.11***

.08***

SES Quartile 3

.19***

.18***

.20***

.14***

SES Quartile 4

.35***

.35***

.38***

.24***

Asian

-.05***

.06***

.00

.05***

Hispanic

-.16***

-.16***

-.17***

-.07***

Black

-.18***

-.21***

-.21***

-.09***

Native American

-.04***

-.04***

-.04***

-.02**

Gender

.06***

-.06***

.00

-.02

Rural

-.01

-.01

-.01

-.01

Urban

.03***

.00

.01

-.01

R^{2 }for Model

.25

.26

.29

.14

*p < .05; **p < .01; ***p < .001 ^{a }N=15,325

The regression coefficient for the 1^{st }wave follow-up math score was not statistically

significant. The parental checking-up variable also produced negative, statistically

significant standardized coefficients in Jeynes’ original study. In the prior study, the

standardized coefficient for reading was B = -.09 (p < .01); for math the score is B = -.11

(p< .001); and the composite math/reading variable was B = -.11 (p < .001).

In the current study, standardized regression coefficients for parents who discuss

school classes, activities, and things studied are the highest of the parental involvement

set ranging from B = .16 (p < .001) for both the reading and combined composite

Table 2. Effects (in Standard Deviation Units) for Parental Involvement in the Their Adolescents for Twelfth Graders in the ELS 2002 – 2004 Data Set N =

Education of 15,362

Academic Measure Intercept

Reading 45.09

Math 46.88

Composite 45.71

1^{st }

Wave Math^{a }31.30

math/reading, and B = .14 for the math score, (p < .001). These scores are also in the

same range as the original study. Jeynes’ standardized coefficients for the parents

discussion variable were B = .14 (p < .001) for reading scores; B = .09 (p < .01) for math