Table 3: Mixed Model* Predicting Error in the Assessment of Professor's Partisanship [Late Semester]
Independent Variables
(Constant)
1.114 ***
(0.267)
Student Sex
0.045
(0.047)
Class Standing
0.001
(0.027)
Expect Grade
-0.080
(0.046)
Follows government and public affairs
- [E]
-0.064 *
(0.031)
Avoids political discussions
- [E]
-0.004
(0.025)
First course with instructor
0.052
(0.083)
Confidence in professor's partisan affiliation
- [L]
-0.181 ***
(0.029)
Recall instructor telling class his/her partisan preference
- [L]
-0.177 *
(0.084)
Paid attention in class
-0.003
(0.028)
Strength of Professor's Partisanship (Actual)
0.338 ***
0.064
n
1,274
Coefficient
Standard errors listed in (parenthesizes) were * p < .05, ** p < .01, *** p < .001
*Classroom variations set as the basis for the "random effects" in the mixed model.
19
Table 4: Shifts in Student Party Identification from Early to Late Semester
Partisan Movement
Strong
Democrat
Moderate
Democrat
Independent
Moderate
Republican
Strong
Republican
Democrat
Independent
Republican
Total
Shift > 1 Unit Dem
6%
7%
6%
3%
4%
2%
3%
Shift 1 Unit Dem
11%
15%
13%
31%
13%
11%
12%
13%
No Change
75%
73%
68%
74%
63%
72%
70%
77%
72%
Shift 1 Unit GOP
21%
12%
10%
7%
11%
13%
6%
10%
Shift > 1 Unit GOP
4%
3%
1%
1%
2%
2%
2%
Lower 95% Confidence
-0.21
-0.02
0.22
0.26
0.56
0.02
-0.06
-0.03
0.02
Mean Units Shift Dem
-0.31***
-0.08**
0.15***
0.19***
0.44***
0.06**
0.03
0.07
0.06***
Upper 95% Confidence
-0.40
-0.13
0.08
0.13
0.32
0.10
0.11
0.17
0.09
n
183
484
354
415
105
1117
261
163
1541
Percent
12%
31%
23%
27%
7%
72%
17%
11%
100%
Student's Party Identification (Early Semester)
Professor's Party Identification
* p < .05, ** p < .01, *** p < .001
Convention