PERC 2021 Abstract Detail Page
Previous Page | New Search | Browse All
| Abstract Title: | A Critique of the use of student pairs as the unit of analysis when examining self-sorting for group learning activities |
|---|---|
| Abstract: | This study provides a critique of the 2017 paper by Freeman et al., "Likes attract: Students self-sort in a classroom by gender, demography, and academic characteristics," which used demographic and performance categories to examine how students self-selected into groups of 3 for in-class group learning activities. Their study used logistic regression with the unit of analysis being the student pair, where every possible student pair within the course is considered, independent of if they were in the same group together. For each of their demographic and performance categories, their analysis approach allowed them to extract the Odds Ratio that people sharing the same identity within a category would form a group together as compared to people not sharing that same identity. This critique demonstrates that their analysis approach is highly sensitive to the proportion of people with each identity within a binary category. Specifically, simulated data sets are used to show that effect sizes and statistical significance for their grouping-preference results are much higher for low-proportion binary identities than for even-proportion ones, given the same underlying levels of affinity when generating the simulated data sets. Toy models and data from self-sorting during group exams provide further insight. |
| Abstract Type: | Contributed Poster Presentation |
| Session Time: | Poster Session 2 Room B |
| Poster Number: | 2B-16 |
Author/Organizer Information | |
| Primary Contact: |
Joss Ives University of British Columbia - Vancouver Non U.S. |
