| Keith Zvoch
University of New Mexico
Parental involvement: A test of factorial invariance across and within family type
FINAL REPORT #1:
Data from the U.S. National Education Longitudinal Survey were examined to investigate postsecondary educational investment in two-parent families. Consistent with hypotheses derived from the logic of inclusive fitness theory, contrasting children with two genetic parents with children from stepparent households on a multivariate composite of investment indicators revealed that stepchildren receive significantly less parental support for pursuit of higher education. Univariate tests on the three measures comprising the multivariate composite indicated that relative to children with two genetic parents, stepchildren have parents who (1) delay the start of savings accounts for postsecondary education, (2) put aside less money for subsidizing the costs of higher education, and (3) expect to allocate fewer economic resources to support the first year of postsecondary schooling. Statistical control of child ability, resource availability, and number of family members sharing in parental resources was accomplished in a second multivariate analysis by using child achievement, familial socioeconomic status, and number of financial dependents in each family as covariates. Statistically equating genetic and stepparent families on these measures reduced, but did not eliminate, the investment differences.
FINAL REPORT #2: The study reported here used data from the base year component ofthe National Education Longitudinal Survey of 1988 (NELS:88) to assess school-level influence on adolescent educational expectations. Consistent with much of the school effects literature, the present study finds evidence for school-level variance, but in a domain tangential to the common outcome of interest. To assess whether aspects ofthe school environment could account for a portion of the variability in school mean expectations, data were analyzed with multilevel modeling techniques. Use of multilevel models enabled adjustment of school differences in individual-level intake characteristics and allowed for specification of school context and school practice indicators at the unit-level. Controlling for several theoretically-relevant individual background characteristics, three contextual features, school SES, school achievement, minority concentration, and one aspect of school practice, beginning teacher salary, accounted for a large portion of the variation in school mean expectation. School context and school practice were not as effective in accounting for variation in the school slopes relating individuai background characteristics to educational expectations. While results generally support a contextual interpretation of school effects, methodological limitations preclude strong inference against the importance of school practice.
Back to Funded Dissertation Grants Page |