Perceptions surrounding the political nature of the academy (i.e., liberalism) remain a salient topic for many academics and political commentators (Horowitz 2004, 2007; Mariani and Hewitt 2008; Tierney 2011).  The ubiquitous liberalism of professors is virtually uncontested (Mariani and Hewitt 2008; Zipp and Fenwick 2006), and this liberalism stands in contrast to the political ideologies of the American population in general. Fosse and Gross (2010), using pooled 1974-2008 General Social Survey data, found that on a seven-point political ideology scale, ranging from extremely conservative to extremely liberal, professors averaged 4.45, while non-professors averaged 3.89.  Although this statistically significant mismatch between academics and non-academics might seem small, consider that this gap is actually greater than the ideological distances observed by race, income, and gender. Not surprisingly, therefore, the causes and consequences of such incongruent political ideologies between professors and the general public is the focus of a considerable amount of discourse and scholarship. 

    Some social scientists have especially focused their efforts on exploring professors' liberalism and the potential marginalization of conservative faculty (Fosse and Gross 2010; Rothman, Lichter, and Nevitte 2005).  A recent article published in Sociation Today exemplifies this approach.  In this study, Bullers et al. (2010) analyzed data derived from a survey of 226 faculty members at a comprehensive public university and found that conservative faculty were not only more likely than their moderate or liberal counterparts to perceive a need to conceal their political beliefs but also more likely to report that their political ideologies had negatively affected their careers.  More recently, Inbar and Lammers (2012) found that conservative social psychologists were more likely than their liberal colleagues to experience a hostile climate within their field because of their political beliefs. Furthermore, a substantial percentage of social psychologists reported that they were willing to discriminate against conservatives when making hiring decisions and reviewing grant applications or manuscripts. 

    Even though concern over the contagiousness of faculty liberalism is the focus of most political commentary on the topic, considerably less research has focused on assessing the effects of professors' political ideologies on student outcomes.  The perception is that faculty liberalism leads to indoctrination on college campuses, and conservative ideas are marginalized as a liberal agenda is "pushed" onto students (see Horowitz 2004).  Although anecdotal evidence of such experiences abounds (e.g., see www.thefire.org and www.studentsforacademicfreedom.org), empirical evidence supporting the liberal indoctrination hypothesis is limited (see Mariani and Hewitt 2008; Woessner and Kelly-Woessner 2009; Zipp and Fenwick 2006).  In fact, with a few notable exceptions (see Kelly-Woessner and Woessner 2006, 2009; Woessner and Kelly-Woessner 2009), little research has been done on how students' perceptions of their instructors' political ideologies influence their college experiences, especially in the classroom.

    Furthermore, although we have good estimates of the political ideologies of college professors, we know next to nothing about the ideologies of graduate students who are being increasingly relied upon to teach their own courses at PhD-granting institutions (Benjamin 2002; Burmila 2010).  This constitutes a salient area of empirical neglect given that significant numbers of undergraduates take courses taught by graduate students.  Moreover, it is in these teaching experiences that graduate student instructors—many of whom will be future professors—develop their philosophies about teaching. Therefore, in this research note we attempt to contribute to (a) the limited scholarship on the classroom effects of students' perceptions of their instructors' political ideologies and (b) the significant lack of research on graduate student instructors. We do so by asking the following research question: Do sociology students' perceived ideological distances from their graduate student instructors affect students' experiences in and evaluations of their courses? 

    Professors are liberal, and social science professors are especially liberal—on average (Bullers et al. 2010; Klein and Stern 2006). In a survey of over 1200 faculty throughout the United States, Smith, Mayer, and Fritschler (2008) found that 58% of faculty members self-reported as liberal, while 80% of sociologists did so. According to Gallup, only 21% of the US population identifies as liberal (Saad 2012). Probably since Plato's Academy, academics have been distinguishable from non-academics in terms of their worldviews, so a persisting political mismatch is unsurprising. However, given the evolution of the modern university from an independent place of knowledge production to a site of commercialization of knowledge and professional career training (Goldstein 2010), the size of the ideological gulf between professors and the rest of Americans is noteworthy. Furthermore, in pop-culture debates, this gap is used as evidence of the marginalization of conservative thinkers and students in higher education (Horowitz 2004, 2007).

     According to Surber (2010), professors are more liberal because of their "practical deliberation, factual investigation, and rational and moral conviction." Although a complete explanation for the mismatch is likely more complex than this (see Fosse and Gross 2010), Surber's assumptions contribute to the cultural picture of professors as superior thinkers who have carefully found the answers and see it as their duty to share their "moral convictions" with students.  To the degree that liberal professors are exemplified by such a portrayal, it is not surprising that, relative to their conservative counterparts, they are more likely to confess that their ideology becomes manifested in their teaching (Bullers et al. 2010).  In such a scenario, it might be expected that the classroom experiences of students whose political ideologies are discordant with their professors may differ from their counterparts whose political ideologies are congruent with their professors. However, as others have pointed out, scant systematic evaluation of such marginalization has been conducted (Kelly-Woessner and Woessner 2006, 2009; Mariani and Hewitt 2008; Tollini 2009).

     Students form perceptions about their instructors (and vice versa) —likely before a course ever begins—and these perceptions are used to define the classroom experience throughout the semester and beyond.  Certainly one perception, among many, that appears to matter within the classroom concerns the political leanings of instructors (Dixon and McCabe 2006; Kelly-Woessner and Woessner 2006, 2009).  Although many students are likely unversed in the cultural debate about politics in the academy, and might not even notice biased teaching when they are exposed to it (Tollini 2009), they likely do enter the classroom with a cultural script in mind for the occupation we call professor. As Fosse and Gross (2010) suggest, professors have a certain occupational reputation that is undeniably liberal.  Thus, conservative students might come to their college classes with messages they have received from parents and others warning of the liberal nature of their instructors; liberal students, on the other hand, might be more likely to assume a shared political philosophy exits between them and their professors.  Furthermore, given that sociologists are among the most liberal within the academy and that sociology as a discipline tends to involve politically-charged issues, we thought it important to evaluate how instructors' and students' political ideologies influence the sociology classroom. 

     To assess sociology students' perceptions of their instructors' political ideologies, we largely follow the work of Kelly-Woessner and Woessner (2006, 2009; see also Woessner and Kelly-Woessner 2008). Using a sample of 1,385 political science students, they found that greater perceived partisan and ideological distance between students and professors resulted in negative evaluations of professors and their courses. For instance, the authors found that the greater the perceived distance, the more likely students were to report feeling that their professors were not open to discussing diverse viewpoints and that their instruction lacked objectivity.  Ideological distance was further found to be predictive of students' perceptions of their professors' commitment to student learning and intellectual growth such that students whose partisan affiliation was perceived to be discordant with their professors were more likely to report that their professors do not care about students. As might be expected, the authors also found that students who perceived their partisan affiliations to be discordant with their professors expressed less interest in the subject, exerted less effort, and were less likely to recommend the course to fellow students.

     The Kelly-Woessner and Woessner (2006) study makes clear that students' perceptions matter, and that a perceived ideological mismatch potentially creates an unfavorable climate for student learning as professors who are seen as different from their students are likely discredited and students simultaneously disengage from the course content.  There is evidence for how perceived political distance affects political science students, so how do such perceptions influence students in sociology courses taught by graduate student instructors?

    We administered surveys toward the end of the semester to a convenience sample of 277 sociology students within the classrooms of five different graduate student instructors at a large research university.  Completed surveys were obtained from 205 of those students sampled, yielding a response rate of 74%.  Survey questions were mostly replications of those used by Kelly-Woessner and Woessner (2006).

     Our key independent variable was the perceived ideological distance between students and their instructors.  Students were asked their political ideology on a 5-point scale ranging from extremely liberal (1) to extremely conservative (5), and their perception of their instructor's ideology on the same scale.  Our ideological distance variable is the absolute value of the difference between student's self-reported political ideology and the perceived ideology of their instructor. 

     We examined the effect of perceived ideological distance across eight dependent variables that were constructed to capture different dimensions of students' classroom experiences and evaluations. The Instructor Evaluation Scale (α = 0.77) consists of five indicators asking students to rate (on a 5-point scale) the degree to which their instructor objectively presented material, graded fairly, encouraged students to express their own views, provided a comfortable learning environment, and cared about students and their success. The Student-Instructor Relationship Scale (α = 0.78) captures the likelihood students were to talk to their instructor after class, ask the instructor for a letter of recommendation, recommend the instructor to other students, and stay in contact with the instructor after the semester terminates.  Another dependent variable we used was students' responses to a statement about whether they would consider sociology as a major after taking their current course, where 5= Strongly Agree and 1= Strongly Disagree. 

     For the Silenced Beliefs Scale (α = 0.63) students were given Likert-style responses for questions about whether they wrote what the instructor wanted to hear rather than their own opinions on assignments, did not feel comfortable expressing their own opinions in the classroom, and felt the need to hide their own beliefs when participating or doing coursework.  Students also were asked how often during the class they felt disinterested, angry, isolated, and unhappy, and these responses constitute the Negative Emotions Scale (α = 0.42).  Although each indicator reflects a separate emotion that may stem from distinct etiological pathways, all four indicators of classroom affect are combined into a single index because each has been associated with reduced motivation, self-regulation, and achievement in academic settings (Pekrun et al. 2002). Relatedly, the low alpha coefficient for this index is unsurprising because internal consistency should not always be expected when measuring such multidimensional constructs (Boyle 1991; Cattell 1978; Hayes, Nelson, and Jarrett 1987).  The Discrimination Scale (α = 0.87) asked students how often they felt discriminated against or singled out in class because of race, sex, political beliefs, religious beliefs, physical appearance, and involvement in extracurricular activities (athletics, clubs, etc.).  We also constructed a Disagreement Scale, which consists of students' responses to two statements about how often they found themselves disagreeing with their professor's viewpoints and the readings or other course content (r = 0.58 for the two items used in the scale).  Finally, we used as an outcome variable responses to a statement asking students if the course provided useful skills and/or knowledge, where 5= Strongly Agree and 1= Strongly Disagree.  

     In our OLS regression analyses, we controlled for the effects of students' age, gender, and expected grade in the course. We also included dummy variables to control for instructor effects.  Unfortunately, we did not have enough variability by instructors or classrooms in our sample to allow for adequate examination of classroom-level or instructor-level contextual effects.  However, we were able to check the robustness of our OLS regression findings by replicating all models using a hierarchical linear modeling approach that accounted for our clustered data structure by nesting students within classrooms. Results of these multilevel models confirmed conclusions derived from our OLS analyses.

     Students rated their political ideology on average as neither liberal nor conservative (M= 3.02; SD= .87) on a 5-point scale, and they rated their professor's political ideology as liberal (M= 2.25; SD= .73).  This yielded an average political distance score of about 1 (see Table 1).

Variable	Mean or %	SD	Min.-Max
Political Ideological Distance	1.04	.97	0-4
Age	19.19	2.18	17-39
Male Sex	47.7%
Expected Course Grade	4.24	.73	2-5
Instructor Evaluation Scale	20.84	2.65	11-25
Student-Instructor Relationship Scale	11.87	3.38	4-20
Consider Majoring in Sociology	2.63	1.14	1-5
Silenced Beliefs Scale	6.43	1.94	3-12
Negative Emotions Scale	8.91	2.12	4-17
Discrimination Scale	7.87	3.35	6-30
Disagreement Scale	4.92	1.37	2-9
Useful Course	3.85	.92	1-5

*This sample was generated via. listwise deletion of missing data.

     Table 2 contains the results of our multivariate OLS regression analyses where we model the effects of political ideological distance on eight outcomes related to students' classroom experiences. Ideological distance has a statistically significant negative effect on instructor evaluations and on the student-instructor relationship. In addition, students are less likely to consider majoring in sociology as perceived ideological distance increases.  Further, our models predict that as ideological distance increases students are more likely to silence their beliefs and experience negative emotions.  Moreover, with greater political ideological distance, students report increased perceived discrimination and greater frequencies of disagreement with their professor and course content.  Finally, ideological distance appears unrelated to whether students think their current course is useful.  In sum, we find deleterious effects for political ideological distance on seven of the eight outcomes related to students' experiences in their sociology class.

Variable	Instructor Evaluation Scale	Student- Instructor Relation -ship Scale	Consider Majoring in Socio- logy	Silenced Beliefs Scale	Negative Emotions Scale	Discrim. Scale	Disagree Scale	Useful Course
Intercept	16.38*	9.86*	1.79*	8.16*	9.21*	10.84*	5.08*	3.04*
Ideological Distance	1.10*	-.504*	-.326*	-.442	-.523	-1.76*	.503*	.02
Ideological Distance Squared	-.684*			.447*	.357*	1.04*
Age	.138	-.024	.029	-.001	.014	.015	.001	-.006
Male	.238	-.874	-.098	-.263	.108	-.604	.119	-.115
Expected Grade	.652	1.01*	.213	-.525*	-.221	-.800*	-.192	.308*
Adj. R square	.246	.196	.114	.153	.115	.152	.124	.217
n	205	196	206	206	205	203	206	206

*p<.05
Table entries are unstandardized coefficients.
Models also include dummy variables to control for instructor effects.
A squared term for ideological distance was explored for each outcome; however, we report  only the trimmed models here.

    Notice from Table 2 that we fit a squared term for ideological distance in our models. A few of the non-linear effects are worth noting. For the silenced beliefs and negative emotions outcomes, the models suggest that some perceived ideological distance is inconsequential for students (see the statistically non-significant first-order effects). However, the second-order terms show that, as perceived ideological distance becomes greater than one, reports of silenced beliefs and negative emotions increase. We find statistically significant first- and second-order terms for the instructor evaluation and discrimination scales. With little perceived ideological distance, instructors are evaluated more positively and perceived discrimination is less likely (see the statistically significant first-order terms). When perceived ideological distance becomes greater than one, however, instructor evaluations decline and perceived discrimination increases. These non-linear effects suggest that a small amount of perceived ideological distance is inconsequential for some outcomes (i.e., silenced beliefs and negative emotions) and interpreted favorably by students according to other outcomes (i.e., instructor evaluations and perceived discrimination). Perhaps some ideological difference creates interest and promotes engagement in the classroom. Alternatively, these findings might reflect more moderate instructors' preference to convey course content objectively without making his/her political beliefs a salient issue in the classroom. 

     A considerable amount of attention has been given to the nature of professors' liberal majority on college campuses, and there is a growing amount of empirical evidence that reveals negative consequences of this liberal climate for conservative professors (Bullers et al. 2010; Inbar and Lammers 2012). Scholarship on students' experiences with their professors' political ideologies, however, is limited, and this is especially the case with regard to professors-in-training (i.e., graduate student instructors). The most relevant exception is the work of Kelly-Woessner and Woessner (2006), whose work we extended here using a sample of sociology undergraduate students who were instructed by graduate students at a large research university.

     We found that when students' perceive themselves to be at ideological odds with their graduate student instructors their experience in the classroom is compromised. Even though students are more likely to perceive these instructors as liberal than conservative, our models suggest that students' perceived distance is important regardless of the direction of this distance. As the mismatch between students' political beliefs and those of their instructors grows, students report increasing levels of negative classroom experiences. We found that students silence their own beliefs, experience negative emotions, and perceive that they are discriminated against, among other outcomes. It is logical to suspect that these types of negative classroom experiences have the potential to compromise student learning and university engagement.  These experiences, based on students' perceptions, have real consequences for instructors, too, in the form of lower instructor evaluations, diminished relationships with students, and a reduced willingness for students to consider sociology as a major.

     The perceived political ideologies of the graduate student instructors in our sample are comparable with other estimates of student perceptions of professors' ideologies (Kelly-Woessner and Woessner 2006) and faculty self-report data (Zipp and Fenwick 2006).  Therefore, it is unlikely we found such consistent negative effects for ideological distance because graduate student instructors are perceived as exceptionally liberal.  More research needs to be done, however, on graduate student instructors' political ideologies and how they handle political and controversial issues in the classroom.  Researchers might even consider whether perceived ideological distance from students shapes graduate students' teaching philosophies and pedagogies.

     Given that most college and university instructors, irrespective of rank and discipline, wish to optimize, not compromise, students' learning experiences in the classroom, our findings will understandably invoke concern.  Although we are uncomfortable suggesting solutions based on our limited study, we do want to make two observations. First, we do not think our findings necessitate a call for value-free teaching. In sociology classes, political ideology permeates course content, discussions, and pedagogy, and this can be useful as we help students grapple with and explore important sociological topics—most all of which are political in nature. However, our findings do suggest that instructors need to be mindful of students' perceptions and experiences to avoid marginalizing students if their goal is to construct a classroom context that maximizes student learning.

     Second, we want to dissuade professors from thinking it is their job to close the perceived ideological distance between them and their students. Although our data suggest that ideological homogeneity between professors and students might improve course evaluations and some classroom experiences, we do not believe political uniformity is a desirable goal. In fact, ideological homogeneity is antithetical to higher education, and it would likely result in undesirable outcomes (e.g., weakened critical thinking and dampened innovation). Rather, we think that the challenge before us is to develop approaches for strategically using political ideological difference to promote curiosity, engagement, and even constructive disagreement.  Instructors, regardless of their political ideologies, need to consider how they can connect with students with varied ideological backgrounds and life experiences.   

     We have presented findings derived from a small convenience sample of undergraduate students at one large university who were enrolled in one of five sociology classes, all of which were taught by graduate students.  Despite the obvious limitations, our study points to considerable consequences associated with the issue of  student-instructor political mismatch for the sociology classroom.  Our project is an attempt to begin filling what is a sizable gap in knowledge on this issue, and we urge our colleagues to do more empirical investigations on students' and instructors' political ideologies. As a discipline, we are already far behind the conclusions on this topic that are held as common knowledge in popular culture. 

Benjamin, Ernst. Fall 2002. "How Over-Reliance on Contingent Appointments Diminishes   Faculty Involvement in Student Learning." Peer Review 5(1):1-10.

Boyle, Gregory J. 1991. "Does Item Homogeneity Indicate Internal Consistency or Item Redundancy in Psychometric Scales?" Personality and Individual Differences 12:291-294.

Bullers, Susan, Melissa Reece, and Christy Skinner. 2010. "Political Ideology and Perceptions of     Bias Among University Faculty." Sociation Today 8(2).

Burmila, Edward M. 2010. "Graduate Students as Independent Instructors: Seven Things to Know about Teaching Your Own Course while in Graduate School." PS: Political  Science & Politics 44:557-560.

Cattell, Raymond B. 1978. Personality and Mood by Questionnaire. San Francisco: Jossey-Bass.

Dixon, Jeffrey C. and Janice McCabe. 2006. "Competing Perspectives in the Classroom: The  Effect of Sociology Students' Perceptions of "Balance" on Evaluations." Teaching  Sociology 34: 111-125.

Fosse, Ethan and Neil Gross. 2010. "Why are Professors Liberal?" Unpublished Working Paper.

Goldstein, Harvey. 2010. "The 'Entrepreneurial Turn' and Regional Economic Development Mission of Universities." The Annals of Regional Science 44:83-109.

Hayes, Steven C., Rosemary O. Nelson, and Robin B. Jarrett. 1987. "The Treatment Utility of Assessment: A Functional Approach to Evaluating Assessment Quality." American Psychologist 42:963-974.

Horowitz, David. 2004. "In Defense of Intellectual Diversity." The Chronicles of Higher Education 50(23): B12.

Horowitz, David. 2007. Indoctrination U: The Left's War Against Academic Freedom. New  York: Encounter Books.

Inbar, Yoel and Joris Lammers. 2012. "Political Diversity in Social and Personality Psychology."    Perspectives on Psychological Science (forthcoming).

Kelly-Woessner, April and Matthew Woessner. 2006. "My Professor is a Partisan Hack: How  Perceptions of a Professor's Political Views Affect Student Course Evaluations." PS:  Political Science & Politics 39:495-501.

Kelly-Woessner, April and Matthew Woessner. 2008. "Conflict in the Classroom: Considering the Effects of Partisan Differences on Political Education." Journal of Political Science Education 4:265-285.

Klein, Daniel B. and Charlotta Stern. 2006. "Sociology and Classical Liberalism." The Independent Review XI:37-52.

Mariani, Mack D. and Gordon J. Hewitt. 2008. " Indoctrination U.? Faculty Ideology and  Changes in Student Political Orientation." PS: Political Science and Politics 41:773-783.

Pekrun, Reinhard, Thomas Goetz, Wolfram Titz, and Raymond P. Perry. 2002. "Academic Emotions in Students' Self-Regulated Learning and Achievement: A Program of Qualitative and Quantitative Research. Educational Psychologist 37:91-105.

Rothman, Stanley, S. Robert Lichter, and Neil Nevitte. 2005. "Politics and Professional  Advancement Among College Faculty." The Forum 3(1): Article 2. Retrieved January  12, 2010 (http://www.bepress.com/forum/vol3/iss1/art2).

Saad, Lydia. 2012. "Conservatives Remain the Largest Ideological Group in U.S." Retrieved  April 29, 2013 (http://www.gallup.com/poll/152021/conservatives-remain-largest-ideological-group.aspx).

Smith, Bruce L. R., Jeremy D. Mayer, and A. Lee Fritschler. 2008. Closed Minds: Politics and Ideology in American Universities. Washington, DC: Brookings Institution Press.

Surber, Jere P. 2010. "Well, Naturally We're Liberal." The Chronicle of Higher Education, February 7. Retrieved February 10, 2010 (http://chronicle.com/article/Well-Naturally-Were-Liberal/63870).

Tierney, John. February 7, 2011. "Social Scientist Sees Bias Within." The New York Times.  Retrieved from http://www.nytimes.com/2011/02/08/science/08tier.html?_r=0.

Tollini, Craig. 2009. "The Behaviors that College Students Classify as Political Bias:  Preliminary Findings and Implications." Teaching Sociology 37: 379-389.

Woessner, Matthew and April Kelly-Woessner. 2009. "I Think My Professor is a Democrat: Considering Whether Students Recognize and React to Faculty Politics." PS: Political  Science & Politics 42:343-352.
 
Zipp, John F and Rudy Fenwick. 2006. "Is the Academy a Liberal Hegemony? The Political  Orientations and Educational Values of Professors." Public Opinion Quarterly 70:304-326.