Impacts of a Problem-Based Instruction Approach to Economics on High school Students
Analysis plan
The analyses for this study will compare outcomes for treatment students and teachers with their control counterparts after completing the economics course, using conditional multilevel regression models. Additional terms will be used to account for the nesting of individuals within higher units of aggregation (see Goldstein 1987; Raudenbush and Bryk 2002; Murray 1998). A random effect for teachers is included to account for the nesting of student observations within teachers. Potential fixed effects will include treatment group, state (California or Arizona), baseline (pretest) measures of outcome variables, and other student- and teacher-level covariates. Exploratory analyses are also planned to investigate differences in problem-based economics program impacts by gender, race/ethnicity, and English language learner and non–English language learner status—with expectations of finding more pronounced positive impacts on students who traditionally exhibit lower levels of academic achievement.
The procedures described by Schochet (2008) will be used to account for multiple hypothesis tests involving the numerous outcome variables assessed in the study. Within each of the three outcome domains—teacher content knowledge in economics, student content knowledge in economics, and student problem-solving skills—confirmatory impact analyses will apply multiple comparison procedures to adjust for errors that can arise from testing multiple hypotheses.