politicaltypes.com

One of the more significant issues in the "culture wars" debate centers on "politicizing the classroom." On the one hand, some political conservatives have proposed an Academic Bill of Rights that seeks to ensure that students of all political orientations are treated fairly in the classroom.  To some extent, this approach seems to assume that some university liberal arts curriculums across the United States constitute a form of ideological indoctrination into Marxism, Radical-Feminism and other forms of radical "Leftism." On the other hand, various academic organizations view these conservative efforts at instituting a "student bill of rights" as attempts to limit academic freedom, and thus, as actions that serve to politicize the classroom anyway--albeit with a conservative slant.

What this debate seems to ignore, however, is the sociological perspective that classrooms inherently are political. The classroom is a political arena in the sense that people (faculty and students) pursue personal interests and in doing so often come into conflict with others who are pursuing their own interests. The concern with politics as "liberal" vs. "conservative" or Republican vs. Democrat and so on reflects only one possible dimension of the political dynamics of the classroom. Moreover, the extent to which political party affiliation and ideological orientation may or may not be significant influences on classroom dynamics remains an empirical question that has received scant attention.

The general purpose of the two research studies summarized here was to examine the concept of classroom politics from a more social-psychological perspective, similar to what has been done in research on organizational politics. This approach addresses two basic questions:

• First, do students perceive a political dimension to the classroom experience?
• Second, what kinds of factors, if any, are associated with this experience?

A Model of Classroom Politics Perceptions

To address these questions we employed (see figure) a well-accepted model of organizational politics perceptions used in studying management and organizational psychology. Briefly, this model proposes work/class environments that involve ambiguity and uncertainty (and thus are possibly ego-threatening) increase the potential for employees/students to perceive their environment as political. Furthermore, the model proposes that these perceptions have significant organizational/class outcomes related to satisfaction, stress, commitment and so on.

A Model of Classroom Politics Perceptions

 

Antecendents. This model suggests four key antecendent factors may influence the extent to which students rate their class as political. These factors are defined as follows:

• Instructor Supportiveness: The extent to which students perceive their instructor listens to, and cares about, student interests.
• Shared Governance: The extent to which students perceive they have a say in how the class is managed.
• Community: The extent to which students perceive their entire class has "bonded".
• Impact: The extent to which students perceive they individually have "control" over what happens in the classroom.

Generally, the more that students perceive the instructor is supportive, that there is shared governance and so on, the more likely it is that students will experience less uncertainty and ambiguity in the classroom. A general hypothesis, therefore, is the relationship between perceptions of "antecedent" factors and classroom politics perceptions should be a negative one. For example, to the extent that students perceive they have a say in what happens in class (class governance), then students should perceive the class as less political.

Outcomes. The model also suggests that students who perceive the classroom as politicized will experience more stress as well as less satisfaction with their learning and commitment to their college or university. These expectations are based on similar findings in the research on organizational politics perceptions.

Methods

Sample. The sample included 217 college and university students who provided complete responses to a survey administered during the spring of 2006. Demographics for those who provided complete information for all research materials (n = 204) indicated the sample comprised predominantly females (63%) enrolled in courses required for their major (81%) with male instructors (67%). The participants’ mean grade point average was 3.2 (s.d. = .51).

Procedures. Participants were asked to volunteer to complete a questionnaire pertaining to the research topic. Questionnaires were administered in classes selected to represent a spectrum of teaching styles, ranging from traditional to highly experiential. Participants also were solicited via social networking venues (facebook.com and myspace.com) to further ensure variance in responses. These participants completed an on-line version of the survey. Approximately 80% of responses were obtained from a private university in the southeastern United States with the remaining 20% of responses coming primarily from schools in the northeastern United States.

Perceptions of Classroom Politics. Students’ perceptions of classroom politics were measured using a revised version of the Perceptions of Organizational Politics Scale. The scale consisted of 15 items hypothesized to measure 3 dimensions of perceptions of politics: General, Go Along to Get Along, and Pay and Promotion/Rewards. Examples of a revised item for each of these factors, respectively, are: “People in the class attempt to build themselves up by tearing others down,” “Sometimes it is easier to remain quiet than fight the system in this class,” and “When it comes to grade decisions, policies from the syllabus are irrelevant.”

Outcomes of Perceptions of Politics. Three scales of 3 items each were used to measure outcomes: stress (e.g., “This class causes me a lot of stress”), satisfaction (e.g., “I am satisfied with my instructor’s performance in this class”), and organizational commitment (e.g., “This class makes my overall university experience more personally meaningful to me”). The satisfaction and stress scales were created for this study. The organizational commitment items were adapted from the Organizational Commitment Questionnaire .

Demographics. For exploratory purposes, sample participants were asked to indicate their sex and grade point average as well as the sex of their instructor and whether or not the class was a class required for their major. Finally, sample members completed a 4-item scale indicating their political orientation (e.g., “When it comes to politics, I generally support Republicans”).

Analyses

Confirmatory factor analysis (CFA) with AMOS 7.0 was used to assess the fit of two measurement models. The first analysis (n = 217) assessed the fit of the 3-factor measurement model for the 15 revised POPS items. The second analysis (n = 217) assessed the fit of the 7-factor measurement model for the 21 antecedent and outcome items. Both analyses specified congeneric measurement models with each item constrained to load only on its hypothesized factor. All possible correlations among latent variables also were specified in each analysis. Although a single analysis would have been preferable, a compromise was reached by conducting two analyses to maintain a sufficient sample size for each analysis.

Hierarchical multiple regression analysis was used to assess the relationship of POPS factors to antecedent and outcome classroom climate factors as well as demographic factors. Two analyses were conducted, one each with the reward factor and the “go along” factor as dependent variables. (No analysis was conducted for the 2-item general factor because of its low reliability.) In each analysis one POPS factor was first regressed against the other POPS factor. The antecedent, outcome, and demographic factors were entered on the second step and the change in R² assessed. Given a significant change in R², individual variables then were examined to determine which specific factors related to politics perceptions.

Results

Measurement Models. Results from the CFA of the revised POPS scale indicated a good fit for the 3-factor model (Chi-square = 157.70, d.f. = 87, p = .000, GFI = .91, RMR = .05, CFI = .93, RMSEA = .06). An examination of the item loadings, however, showed two items with non-significant loadings. All other loadings were statistically significant and substantial with standardized loadings ranging from .54 to .90. Given that the two non-significant items also were listed (at random) as the first two items on the questionnaire, it is possible that the non-significant results may have been due to a “priming” effect. Thus, the measurement model was again estimated without these two items. Results indicated a slightly better fit (Chi-square = 102.94, df = 62, p = .001, GFI = .93, RMR = .04, CFI = .96, RMSEA = .06). Items from this analysis thus were used to create scales for the regression analyses.

Results from the CFA of the measurement model for the antecedent and outcome scales also indicated a good fit. However, examination of the estimated correlations among the latent variables suggested the satisfaction and instructor support scales (r = .91) may not have measured distinct constructs. The satisfaction items thus were omitted and a 6-factor measurement model estimated. Results from this model also indicated an acceptable fit (Chi-square = 214.33, df = 120, p = .000, GFI = .91, CFI = .95, RMR = .05, RMSEA = .06).

Regression Analyses. Results for both hierarchical regression analyses indicated a significant change in R² when antecedent, outcome, and demographic variables were entered on the second step for each analysis. And, as has been found with previous research concerning perceptions of politics, different variables were related to the two politics perceptions variables.

After controlling for “go along” perceptions, five variables were significantly related to reward perceptions:

• Instructor Supportiveness
• Commitment
• Classroom Community
• Instructor Sex
• Required Course

In general, increasing perceptions that grading was political were associated with lower perceptions of instructor supportiveness and organizational commitment, higher perceptions of classroom community, female instructors, and required course status.

After controlling for reward perceptions, four classroom climate variables were found to be significant predictors of “go along” perceptions:

• Instructor Supportiveness
• Impact
• Shared Governance
• Stress

In general, higher perceptions that one needed to “go along to get along” were associated with lower ratings of instructor supportiveness, impact, shared governance, and higher ratings of stress. One demographic variable—instructor sex—showed a near-significant relationship with “go along” perceptions, indicating female instructors engendered such perceptions more so than did male instructors.

A more detailed discussion of Study 1 may be found here.

Study 2

Methods

Sample. The sample included 341 college and university students who provided complete responses to a survey administered during the fall of 2006. Demographics indicated the sample comprised predominantly females (56%) enrolled in courses required for their major (85%) with male instructors (67%). The participants’ mean grade point average was 3.3 (s.d. = .44). The majority of students were junior and senior undergraduates (69%) and 52% of sample members were enrolled as business majors. Responses were received from at least 18 different colleges and universities.

Procedures. Student participation was solicited by e-mailing 10 professors at 4 colleges and universities in each of the 50 states in the United States. Professors were requested to encourage their students to complete the anonymous, on-line survey.

Measures. Forty items that were hypothesized to measure 7 dimensions of classroom climate and 3 dimensions of perceptions of politics were administered as one survey. A five-point Likert format was used for all items: Strongly Disagree, Disagree, Neutral, Agree, Strongly Agree. Scale scores were created by calculating the arithmetic mean of each set of items.

Perceptions of Classroom Politics. Students’ perceptions of classroom politics were measured using the revised version of the Perceptions of Organizational Politics Scale used in Study 1. As a check for the potential priming effect found in Study 1, all items again were numbered randomly.

Antecedents to Perceptions of Politics. The four scales used in Study 1 also were used in Study 2.

Outcomes of Perceptions of Politics. The scales used in Study 1 to measure stress and organizational commitment also were used in Study 2. One item of the satisfaction scale was changed to more clearly indicate satisfaction with learning. The changed item had focused on satisfaction with the instructor.

Demographics. As was done in Study 1, sample participants were asked to indicate their sex and grade point average as well as the sex of their instructor and whether or not the class was a class required for their major. Sample members again completed a 4-item scale indicating their political orientation (e.g., “When it comes to politics, I generally support Republicans”).

New demographic and contextual variables also were included in Study 2. Participants were asked to indicate their class status (Freshman, Sophomore, etc.) and their major. Concerning contextual factors, participants were asked to indicate class size (< 25, 26 to 50, > 50), whether or not class members worked in teams (yes or no), how frequently the instructors talked in class about his or her political opinions  and the participant's perception of instructor political orientation .

Analyses

Confirmatory factor analysis (CFA) with Mplus 4.1 was used to assess the fit of two measurement models. The first analysis assessed the fit of the 3-factor measurement model for the 15 revised POPS items. The second analysis assessed the fit of the 7-factor measurement model for the 21 antecedent and outcome items. Both analyses specified congeneric measurement models with each item constrained to load only on its hypothesized factor. All possible correlations among latent variables also were specified in each analysis. Although a single analysis would have been preferable, a compromise was reached by conducting two analyses to maintain a sufficient sample size for each analysis. The Mplus WLSMV procedure with delta parameterization was used for estimation.

Hierarchical multiple regression analysis again was used to assess the relationship of POPS factors to antecedent and outcome classroom climate factors as well as demographic factors. The same procedures used in Study 1 were used in Study 2.

Results

Measurement Models. Results from the CFA of the revised POPS scale (Chi-square = 237.53, d.f. = 43, p = .000) indicated a good fit for the 3-factor model on two fit indices (CFI = .92, TLI = .96). Fit indices focusing on residuals, however, suggested a somewhat worse fit (RMSEA = .12, SRMR = .06). An examination of the item loadings indicated that the four reversed items on the adapted POPS had significant, but lower loadings. The results suggested the lower loadings might be artifactual; we thus again fit a 3-factor model with a fourth artifactual factor added. The fit of this model showed a modest improvement in overall fit (Chi-square = 133.73, d.f. = 45, p = .000, RMSEA = .08, SRMR = .05, CFI = .96, and TLI = .98) and the four negatively-worded items all loaded significantly on the artifact factor.

Results from the CFA of the measurement model for the antecedent and outcome scales suggested a high degree of multicollinearity in that some loadings produced inflated estimates. Thus, each hypothesized scale was assessed separately by examining item-total correlations as well as a scale's coefficient alpha with and without the item. Most scales demonstrated good internal consistency as shown by the calculated coeffient alpha for each scale (see Table 1).

Regression Analyses. Results for both hierarchical regression analyses indicated a significant change in R² when antecedent, outcome, and demographic variables were entered on the second step for each analysis. The change in R² for the 3rd step was close to significant (p = .08). Thus, Table 1 shows results for the full model.

After controlling for grading perceptions, four classroom climate variables were found to be significant predictors of “go along” perceptions. Instructor supportiveness showed the strongest relationship (B = -.26, p <.001) followed by stress (B = .23, p < .001), impact (B = -.15, p < .001), satisfaction with learning (B = -.13, p < .05), and . In general, higher perceptions that one needed to “go along to get along” were associated with lower ratings of instructor supportiveness, impact, and satisfaction, as well as higher ratings of stress. Three demographic variables—instructor sex, student sex, and required course—also showed significant relationships with “go along” perceptions. On average, males students in a required course with a female instructor reported higher perceptions of needing to "go along to get along."

Discussion

Do students perceive a political dimension to the classroom experience? The results from these two studies suggest the answer is "yes." Further, the modified version of the POPS used in the studies appears to measure validly two separate dimensions of political climate, one which focuses on issues related to conforming to classroom norms ("Go Along to Get Along") and another which focuses on issues related to perceived fairness and consistency in grading practices. Not surprisingly, these two dimensions are correlated.

What factors are related to perceptions of classroom politics? Results from these two studies suggest the answer to this question depends upon the political dimension being considered. As has been found in research on organizational politics perceptions, different factors related to different dimensions of classroom politics (although there is overlap).

As can be seen in Table 2, some generalizations appear reasonable for the following factors:

• Instructor Supportiveness. Clearly the more that students perceive their instructor as supportive, the less likely they are to perceive the classroom as politicized in terms of either dimension.
• Stress. The more that students perceive they have to "go along," the more stress they report. Interestingly, reported stress is not associated with the grading dimension in either study.
• Impact. The more that students perceive they individually have control over what happens in class, the less they perceive the class as political in terms of "going along".

Comments?

Please feel free to leave your comments about the studies. 

Also, you may complete a copy of the survey here. 

---