|
Activity Number:
|
254
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistical Education
|
| Abstract - #308759 |
|
Title:
|
Calibrated Peer Review (CPR) as a Tool To Teach Interpretation of Linear Regression Coefficients: Development of the CPR Module and Examination of Students' Results
|
|
Author(s):
|
Felicity B. Enders*+ and Sarah Stoner and Verna Hoverman
|
|
Companies:
|
Mayo Clinic and Mayo Clinic and Mayo Clinic
|
|
Address:
|
200 First Street SW, Rochester, MN, 55905,
|
|
Keywords:
|
Statistics education ; Calibrated peer review ; Regression
|
|
Abstract:
|
Linear regression can be an extremely challenging topic. In this class of 27 masters of clinical research students, prime importance was placed on learning to correctly interpret regression parameters. In order to facilitate learning, we developed a calibrated peer review (CPR) module (http://cpr.molsci.ucla.edu) to help students interpret coefficients for a single binary predictor, a single continuous predictor, or a set of dummy variables. Students' interpretation scores improved after completing the module (p=0.06). Evaluation data show that while some students did not enjoy using the CPR system, they did feel it helped them to learn (p=0.04). In particular, students reported that grading their own initial assignment was a helpful learning exercise (p=0.002). Calibrated Peer Review can be a useful learning tool for text-based statistical exercises.
|
