Predicting Success In Preparing For High-Stakes Admissions Tests: A Moderated Mediation Analysis
Date of Award
Doctor of Philosophy (PhD)
Educational Psychology and Special Education
Dr. Karen M. Zabrucky
Dr. Nanette Commander
Dr. Hongli Li
Research on preparing for standardized college admissions tests such as the SAT has largely been limited to studies exploring the relative effect size of test preparation. In several analyses and meta-analyses, investigators have demonstrated a positive effect of test preparation (e.g., Briggs, 2005; Kulik, Bangert-Drowns & Kulik, 1984; Lilly & Montgomery, 2011; Powers & Rock, 1999). Moving beyond the fundamental question of whether SAT prep influences test scores, researchers have only recently begun to explore the individual factors that inform successful test preparation. In their regression analysis of the salient factors of successful SAT preparation, Appelrouth, Moore, & Zabrucky (2014) found significant effects of homework completion, instructional hours, practice and official testing, distribution of study, and timing of test preparation. The current study builds upon that research in constructing a functional model of the factors involved in successful SAT preparation. It was hypothesized that there would be direct and indirect relationships between the factors of test preparation, and that some of these relationships would be moderated by student characteristics such as gender and socioeconomic status. Archival data from 1,933 students, provided by a private tutoring company, were analyzed. Significant direct relations were reported between tutoring start time and the following variables: session distribution, individual tutoring hours, group tutoring hours, homework completion, number of official tests, number of practice tests and total SAT increase. Starting tutoring earlier junior year yielded a number of positive direct and indirect effects. Session distribution, individual and group tutoring hours, and official SAT and practice SAT tests all mediated the relationship between start time and SAT score increase. Though gender had no significant moderating effects, both school type and socioeconomic status moderated the relationship between start time and individual tutoring hours. School type also moderated the relationship between homework completion and score increase. The results of this analysis have implications for the thousands of high schools and educational entities that offer SAT coaching programs. By encouraging earlier program start times, adequate instructional hours, distribution of sessions and practice effects, administrators can create more effective SAT preparation programs to serve their college-bound students.
Appelrouth, Jed, "Predicting Success In Preparing For High-Stakes Admissions Tests: A Moderated Mediation Analysis." Dissertation, Georgia State University, 2015.