Using Advising and Enrollment Data to Inform a First-Year Math Placement Program

We report on a three-year project to make data-driven improvements in the mathematics placement process at the University of Northern Colorado. We began by analyzing fall 2007 placement recommendations for a sample of N=1466 first-year students to the university. These recommendations came from brief faculty-student interviews during summer orientation sessions in which math instructors suggested one or more courses for students based on their most recent mathematics course and grade, high school grade point average, ACT math score, college major, and other information. We compared these recommendations to advising and enrollment data over the subsequent year, and, using logistic regression modeling, identified the background variables that best modeled success in students’ first mathematics courses. This led us to make changes in the math placement process for summer 2009. We describe the new placement guidelines and summarize preliminary findings from a follow-up study on the impact of the changes.