Analyzing Student Perceptions of Teaching with Quantile Regression

• Activity Number: 444
• Type: Contributed
• Date/Time: Tuesday, August 6, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Survey Research Methods Section
• Author(s): Kellie Keeling*+ and Robert Pavur
• Companies: University of Denver and Univ of North Texas
• Keywords: Quantile Regression ; Student Evaluations ; Regression

Abstract

In this study, we use quantile regression to answer the question: Is there a difference in the importance of questions for the students who score the course or the instructor as low, medium, or high? Two advantages of using quantile regression are that in a series of observations, there may be a small portion of "outliers" which are represented more accurately by the quantile regression method and that the distribution of the sample can be accurately fitted. We have collected data on over 5,000 student evaluations across multiple disciplines (Arts and Humanities, Business, Engineering, and Natural Science and Mathematics) and across undergraduate and graduate sections of courses. We collected data in the Fall quarter for the years 2005, 2008 and 2011. We report our findings and note that using Quantile Regression can provide insights into which components of the instructor or course weigh against a student rating a professor higher or lower. And we note differences that we found across disciplines and course levels.