Proportional Hazards Model for Risk Factors
A challenge in assessing the effectiveness of a first semester Calculus course is the relatively large number of students who withdraw during the semester. Standard techniques (e.g., correlation, linear regression) are hampered by the loss of data due to student withdrawals. This study applies survival analysis methodology to examine risk factors contributing to student failure to successfully complete a first semester undergraduate course in Calculus. The concept of censored data is discussed, and the proportional hazards model is introduced. Some of the underlying theory is discussed, including Cox regression to estimate parameters of the model. Finally, Cox regression is applied to data from the Fall 2003 semester to investigate risk factors in Calculus 1.
All Mathematics students and faculty, as well as other interested persons, are invited.
- Led by: Dr. V. J. Durapau, Jr., Math Department
- Date: Thursday, March 11, 2004
- Time: 10:50 AM - 12:05 PM
- Location: Administration 320
- Sponsor: CAT
Tags: STEM, statistics
Event ID: 00238