Research That Matters (January 17 - 20, 2008)


Congressional Room A (Omni Shoreham)

Assessing the Mediated, Direct and Indirect Effects of Two Parenting Factors on the Academic Achievement of Mexican American Youths

Inna Altschul, PhD, University of Denver.

Background and Purpose: Increasing parental involvement with school has emerged as a popular intervention for improving students' academic achievement, particularly for disadvantaged youths. Empirical studies support the positive association between parental involvement and achievement (Jeynes, 2007). A second parenting factor, parental investment or allocation of resources to education (Mayer, 1997), has been hypothesized to account for the transmission of family economic advantage to child academic achievement but is less frequently assessed. This paper examines the relative contribution of parental involvement and parental investment to the academic achievement of Mexican American youth in the U.S. The research questions addressed are: does parental involvement have a greater influence on academic achievement than parental investment for this population of youth? What student characteristics mediate the effects of parental involvement and investment?

Method: A subset of data from the National Educational Longitudinal Study (NELS) representing Mexican American youth in the U.S. (n=2206) were used for these analyses. Structural Equation Models (SEM) representing proposed relationships between parenting factors, student characteristics, and test scores were assessed using MPlus 4.1 (Muthen and Muthen, 2006). Comparative Fit Index (CFI), Root Mean Square Error of Approximation (RMSEA) and Standardized Root Mean Square Residual (SRMR) were used to determine goodness of fit for all models. Measurement models for latent factors representing parental involvement and parental investment, each with three indicators, yielded excellent fit statistics. Two structural models were evaluated. The first assessed the direct effects of three parenting factors (parental involvement, allocation of resources to education, and help with homework) on 8th grade test scores, while the second SEM also included six student factors (positive school behaviors, academic and occupational expectations, language proficiency in English and Spanish, prior grades and being held back) as mediators. Both models yielded good fit indices suggesting a good fit to the data.

Results: Path coefficients for both structural models show that while parental investment (ß=0.383, p<0.001) and help with homework (ß=-0.183, p<0.001) had significant direct & indirect effects on test scores, parental involvement with school did not (ß=0.123, p>0.05). Parental involvement did have significant relationships with prior grades (ß=0.162, p<0.001), positive school behaviors (ß=0.142, p<0.01), and English proficiency (ß=-0.366, p<0.001), which are all significant predictors of test scores. Parental investment was found to have significant relationships with expectations (ß=0.361, p<0.001), positive school behaviors (ß=-0.152, p<0.05), and English proficiency (ß=0.862, p<0.001).

Conclusions and Implications: These findings suggest that for Mexican American youth parental investment in terms of allocation of resources to education is a stronger predictor of performance on standardized tests than parental involvement, while parental involvement is strongly related to school grades. The differential relationship of these two parenting factors with two academic achievement outcomes have important intervention implications; they suggest that parental involvement is a better target for interventions aiming to influence student grades, which are somewhat subjective measures of performance and account for student behavior. Conversely parental investment would be a more appropriate target for interventions aiming to improve performance on standardized tests, which is a more direct measure of cognitive abilities.