Losing Control: Conducting Studies with Comparison Groups

Studies in education often report the differences between participants' and non-participants' test scores, course grades, retention, and other criteria. When participants' average performance is higher, it can be difficult to attribute the improvements to participation. Comparing participants and non-participants on other measures can strengthen the argument that participation had a positive impact when the two groups are otherwise similar in relevant ways. Reviewing students' demographic characteristics, incoming ACT/SAT scores, previous grades, and placement results can establish points of comparison; specific statistics can also assist in identifying similarities and differences between groups.