PERC 2013 Abstract Detail Page
Previous Page | New Search | Browse All
| Abstract Title: | Applying Latent Class Analysis to Explore Patterns of Student Responses in Physics Education Research |
|---|---|
| Abstract: | We use the statistical method of Latent Class Analysis (LCA) to explore patterns in student ideas and attitudes. LCA is an exploratory multivariate analysis similar to cluster analysis that can use student responses to a survey or qualitative binary codes to group students into latent classes. LCA differs from the more commonly used factor analysis because it groups individuals rather than items. Here, we present multiple examples of how LCA can be used to explore questions of interest to Physics Education Researchers. Our first example uses LCA to understand student responses to the Colorado Learning Attitudes towards Science Survey (CLASS), illustrating how to use LCA for quantitative survey responses. Our second example uses LCA to identify patterns in students' drawings about day and night cycles, illustrating how LCA can be used to understand patterns in qualitative artifacts. |
| Abstract Type: | Contributed Poster Presentation |
Author/Organizer Information | |
| Primary Contact: |
Danielle B. Harlow UC-Santa Barbara Gevirtz Graduate School of Education University of California, Santa Barbara Santa Barbara, CA 93106-9490 |
| Co-Author(s) and Co-Presenter(s) |
Karen Nylund-Gibson, University of California Santa Barbara |
