A Sigmoid Model of Correlation to Identify Bias in Students' Evaluation of Teaching
A statistical analysis in determining if students in science and engineering classes shows bias when evaluating the effectiveness of their professor as a result of their grade in the course.The bias is often thought to be a result of the difficulty or rigor of the course. An alternative theory is proposed; the statistical relationship between a professor's student evaluation is not determined by rigor but is instead a reaction of grade inflation. This is shown in that overall grades are not statistically significant but the percentage of what has been traditionally named the average grade - or C. Whether it is rigor or grade inflation this indicates that students are exhibiting bias in their evaluations by altering their opinion from the professor's effectiveness to their expectation of the grade that they would receive.The analysis investigates the correlation between student evaluations and grades in two ways; the common Pearson linear correlation, and since the range of evaluative scores and grade percentages are both constrained, a nonlinear sigmoid regression model.
Files
Metadata
Work Title | A Sigmoid Model of Correlation to Identify Bias in Students' Evaluation of Teaching |
---|---|
Subtitle | Proceedings of 2021 Northeast Decision Sciences Institute Conference |
Access | |
Creators |
|
Keyword |
|
License | In Copyright (Rights Reserved) |
Work Type | Article |
Publisher |
|
Publication Date | March 2021 |
Related URLs | |
Deposited | November 10, 2024 |
Versions
Analytics
Collections
This resource is currently not in any collection.