•  
  •  
 

Keywords

learning communities, self-selection, research methods, program assessment

Abstract

This article presents a method for addressing the self-selection bias of students who participate in learning communities (LCs). More specifically, this research utilizes equivalent comparison groups based on selected incoming characteristics of students, known as bootstraps, to account for self-selection bias. To address the differences in academic preparedness in the fall 2012 cohort, three stratified random samples of students were drawn from the non-LC population to match the LC cohort in mean ACT composite scores and mean high school percentile ranks. This process is called bootstrapping. The study suggests that LCs do impact student academic achievement and retention. The results indicate that LC students with similar entering characteristics to those of the bootstrap sample had higher rates for both GPA and retention than non-LC participants.