Non-Randomized Groups: Implications and Strategies using Propensity Score Analysis

Group comparisons are common in educational research but results from these comparisons can be biased when individuals are not randomly assigned to the groups in question. Researchers may be familiar with ANCOVA, but this approach is not well-suited for quasi-experimental design. This workshop will demonstrate potential issues with non-random assignment and provide statistical alternatives using propensity score analysis (matching, inverse probability of treatment weighting). Participants will engage with heuristic data in SPSS and R.