Abstract
Political scientists have noted that, in a variety of communication settings, people are less receptive to information that comes from a politically disagreeable source. Yet, there is little research on patterns of communication across lines of political difference in an educational setting, which we argue is unique in a number of ways. Using a large-scale national survey of college students enrolled in political science courses, we examine how perceptions of professors' political orientations contribute to student learning, interest in politics, and effort in the course. We consider both direct effects of partisan difference on education and indirect effects, which work through a number of source credibility measures. Our analysis indicates that students who believe their professor to be a political ally report more learning, higher levels of effort, and greater interest in the subject than those who believe their professor to be a political foe.
Notes
Significant correlations shown with ∗ < .05, ∗∗ < .01, ∗∗∗ < .001 using a two-tailed test.
∗Classroom variation set as the basis for the “random effects” in the linear mix models.Standard errors listed in (parenthesis) where ∗ < .05, ∗∗ < .01, ∗∗∗ < .001 with a two tailed-test.
∗Classroom variation set as the basis for the “random effects” in the linear mix models.Standard errors listed in (parenthesis) where ∗ < .05, ∗∗ < .01, ∗∗∗ < .001 with a two tailed-test.
∗Classroom variation set as the basis for the “random effects” in the linear mix models.
“ ⇐ ” Denotes the variable with the highest standardized coefficient.
The IDEA (Individual Development and Education Assessment) center is located at the University of Kansas. Further information about IDEA evaluations is available online at www.idea.ksu.edu/.
Owing to the very few instances in which a student placed oneself on one end of the partisan spectrum, and the faculty member the other, we combined distances of three and four units into one collapsed measure, with a value of “3” indicating partisan distance of three or more points.
Pearson's R correlations between the partisan distance and the six dependent variables are as follows: interest, R = − 0.128; learning, R = − 0.099; effort, R = − 0.066; course evaluations, R = − 0.136; professor evaluations, R = − 0.146; recommend the course to others, R = − 0.116. Each of the correlations is statistically significant at the .05 level.
We use the term “relatively conservative,” as the sample of perceptions is largely skewed toward the left. Only 20% of the students identified their professor as conservative. The bulk of the observable variation (72%) occurs between “moderates” and “fairly liberal.”