|
Yoonkyung Oh
University of Wisconsin-Madison
School-level contextual effects of parent involvement on children's achievement growth in reading and math in primary grades
FINAL REPORT:
The concept of parent involvement has been largely understood as the activity of private persons to serve their own children’s best interests. While involvement of parents - as a primary caretaker, supporter, and advocate for their child - is an individual attribute, but at the same time it should be also recognized as a collective-level, social organizational feature that may lead to improved performance of a school and students as a whole. I argue that parent involvement has the potential to create public benefits for the entire school community because involved parents can make a difference in the internal contexts of school organizations, in terms of social, resource, and accountability contexts. First, involved parents may contribute to improving the social relational contexts of schools, creating a more positive school climate, enhancing trust and collaboration among participants, and developing a sense of collective community (Bryk & Schneider, 2002; Coleman, 1988; Warren et al., 2009). Parents also have the potential to influence the resource contexts of schools by generating additional financial and human resources and by exercising their voices over resource-related issues (Brunner & Imazeki, 2003; Castro et al., 2004). Parent participation may also create school contexts in which administrators and teachers are more accountable and more responsive to the needs of the local community and the students they serve (Comer, 2005; Rogers, 2004; Shatkin & Gershberg, 2007; Wholstetter & Odden, 1992). Improvement in social, resource, and accountability contexts would, in turn, lead to the improved performance of students in the school. Focusing on the role of parents as a part of larger social and organizational contexts of schools, this study uses the ECLS-K to examine the contextual influences of parent involvement on children’s reading and math achievement during elementary grades. First, I use the Rasch measurement model to develop the individual-level latent constructs of parent involvement based on a set of relevant items. A three-level hierarchical linear model (with a level-1 measurement model) is then used to estimate the individual-level latent true scores of parent involvement adjusted for measurement error and to construct a more reliable measure of the unobserved true school means on parent involvement. In the third step of analysis, the estimated individual- and school-level true scores of parent involvement are incorporated into the piecewise hierarchical linear growth model to examine whether parent involvement constructed at the school level is related to children’s achievement growth after controlling for the effect of individual-level parent involvement. The study generally finds no significant between-school learning growth variations associated with the school means on parent involvement. One exception is that for learning growth between the end of 1st and 3rd grade, the effects of the school-level measure of school-based involvement are moderated by school SES composition. Its association with school achievement growth during this period was negative in high SES schools, but positive in low SES schools. The finding is consistent across reading and math.
Back to Funded Dissertation Grants Page |