PERC 2011 Abstract Detail Page
Previous Page | New Search | Browse All
| Abstract Title: | A 12,000-student analysis of normalized gain and alternative metrics using the FCI |
|---|---|
| Abstract: | The normalized gain, g, is among the most widely- and easily-used metrics in physics education research, particularly as it is applied to the Force Concept Inventory (FCI). However, its ease and ubiquity mask several pitfalls associated with the metric, particularly when it is used to directly compare a small number of courses. Specifically, these pitfalls are: the ambiguities between "class" and "student average" gain, the tendency not to report error margins, the effect of losses in initially high-scoring populations, and the extent to which details about classes are washed out by reporting a single number. Here, we extend the work of others by highlighting these pitfalls using data collected from over 12,000 students in several universities. Ultimately, we propose an alternative metric that assesses the rate of Newtonian thinking in a class, which is more closely aligned with the description of the FCI suggested by the survey authors. |
| Abstract Type: | Contributed Poster |
Author/Organizer Information | |
| Primary Contact: |
Jason Dowd Harvard University 17 Oxford Street Physics Dept. - Harvard U. Cambridge, MA 02138 Phone: 7733830088 |
| Co-Author(s) and Co-Presenter(s) |
Kelly Miller, Harvard University Brian Lukoff, Harvard University Nathaniel Lasry, Harvard University and John Abbott College Eric Mazur, Harvard University |
