This study used the data collected from Lehigh University's peer tutoring program which focused on 20 courses involving problem-solving skills over three academic years, 2003-04 through 2005-06. An OLS model and generalized liner mixed-effects models were used to measure effect of participation in tutoring in terms of hours spent getting tutored on exogenous final grade on subset of sample. Exploratory data analysis was performed to examine correlations between students' academic performance and other attributes including previous term GPA, high school rank, SAT scores, and demographic characteristics.
Useful source code: finalproject.Rmd (incomplete version) Presentation Slides: peer-tutoring.pptx