Virtually every university in the US has a mathematics “placement” policy. These policies are needed because high school mathematics curriculum across different schools is highly non-uniform, resulting in large variations in students’ skills. The main goal, for each entry-level mathematics class, is to identify students who have the needed prerequisite skills to take the class and students who need to enroll in a lower-level class first. But do these policies work? And how can we tell?
Towson University also has a placement policy, one that has changed over the years from threshold-based remedial placement to fixed-question set diagnostic exam placing students in introductory courses of various levels (for example, College Algebra or Precalculus or Calculus I) to adaptive diagnostic exam with the same goal of selecting the right level of mathematics course. A colleague asked me for a good way to evaluate performance of a placement policy: even after adjusting the cut-off scores, the results, at least anecdotally, did not seem to show improvement.
It turned out that there are very few rigorous studies of mathematics placement at 4-year universities; large community college systems are doing a much better job, but they focus mainly on remedial placement. My colleagues and I started looking at the university data, carefully thinking about methodology. The first paper below is the result of a project that involved 5 students. It used regression discontinuity analysis to evaluate the policy based on Math SAT threshold. In the second paper, we examine student outcome metrics that are used in the literature to compare placement policies (and point out substantial drawbacks in them).
- Kolesnikov and Wang; Bonaduce, Cunningham, Fontinell, Halliwell, and Twillman. (2020) Imagine No Remediation: Evaluation of a Placement Policy Change. Research in Higher Education. 61(7), 849-870.
- Frank, Kolesnikov, and Wang. (2022) Comparing Placement Policies at a Four-Year Institution. PRIMUS. 32(10), 1076-1091.
- Frank, Kolesnikov, and Wang. (preprint) A New Metric for Evaluating and Comparing Placement Policies.