Leader Communication: Faculty Perceptions of the Department Chair

Abstract

This project asked 202 randomly selected faculty members to evaluate the supportive and defensive communication and leadership behaviors of their department chair. The supportive behavior of problem orientation alone explained 43% of the variance in faculty ratings of chair job effectiveness. On the other hand, empathy explained 68% of the variance in faculty ratings of relational satisfaction with their department chair. In addition, a t test revealed that more effective chairs utilized all 6 of Gibb's supportive communication behaviors more, and 5 of 6 defensive behaviors less, than their more negatively evaluated peers. Second, a series of stepwise regression procedures explained 16% of the variance in bureaucracy scores, 69% of the variance in Machiavellianism, and 62% of the variance in transformational leadership. Finally, the communication behaviors of strategy, neutrality, and problem orientation explained 56% of the variance in faculty job satisfaction and strategy, neutrality, evaluation, gender, and age explained 41% of the variance in organizational commitment scores.

Keywords:

• Defensive And Supportive Communication
• Department Chair
• Job Satisfaction
• Leadership Communication
• Organizational Commitmen

This article was presented at the National Communication Association, Chicago, IL, 2007. This article is based on Kathleen Czech's doctoral dissertation under the direction of Fred Galloway.

Notes

Note. N = 202. DE = description; PO = problem orientation; PR = provisionalism; EM = empathy; EQ = equality; SP = spontaneity; SU = superiority; EV = evaluation; CE = certainty; NE = neutrality; CO = control; ST = strategy; EF = chair effectiveness; CS = chair satisfaction; MA = Machiavellianism; BU = bureaucracy; TR = transformational; JS = job satisfaction; OC = organizational commitment.

*p = .05. **p = .01.

Note. N = 202; SE = 1.53; adjusted R² = .53; constant = −.51.

*p < .05. **p < .01. ***p < .001.

Note. N = 202; SE = 1.14; adjusted R ² = .77; constant = −2.21.

*p < .05. **p < .01. ***p < .001.

Note. All differences between means are statistically significant at p < .001.

^a n = 74.

^b n = 53.

Note. All differences between means are statistically significant at p < .001.

^a n = 104.

^b n = 42.

Note. N = 202; SE = .48; adjusted R ² = .69; constant = 1.34.

*p < .01. **p < .001.

Note. N = 202; SE = .69; adjusted R ² = .16; constant = .24.

*p < .001.

Note. SE = 0.54; adjusted R ² = .62; constant = −.24.

*p < .05. **p < .01. ***p < .001.

Note. SE = 0.45; adjusted R ² = .56; constant = 2.87.

*p = .01. **p < .001.

Note. SE = 0.68; adjusted R ² = .41; constant = 3.25.

*p < .05. **p < .01. ***p < .001.

Note. Variables are listed in order of importance in a stepwise regression model. + = positive relationship; − = inverse relationship.

^a R ² = .56.

^b R ² = .41.

^c R ² = .77.

^d R ² = .53.

Note. Variables are listed in order of importance in a stepwise regression model. + = positive relationship; − = inverse relationship.

^a R ² = .69.

^b R ² = .16.

^c R ² = .62.

The focus of this article is specifically on the relationship between faculty and the department chair. We recognize that the relationship between the chair and other administrators is just as important to success and continued tenure in that office, but involves additional factors that are somewhat different from those explored in this research.

Due to the number of predictor variables involved and the statistical tests employed in this research, we followed the advice of Hair, Anderson, Tatham, and Black (1992, p. 46) and established N ≥ 200 as the goal for our sample. We believed an initial mailing list of 420 faculty members would enable us to reach that size sample.

Our decision to use single-item measures of these two variables was based on several factors. First, because we were interested in global opinions about chair effectiveness and relational satisfaction, rather than the specific components that may contribute to those opinions, we used the simplest overall measure of the construct (Bernard, 2000; Wanous, Reichers, & Hudy, 1997). Second, previous research that successfully utilized this method for summary purposes enabeling us to create a parsimonious instrument in an effort to increase our overall response rate (Dillman, 2000).

We compared the four academic domains using one-way analyses of variance. There were no statistically significant differences between any of the groups relative to relational satisfaction, job satisfaction, or organizational commitment.

Warner (2008) noted that in non-experimental group comparisons, “subjects need to be chosen so that they are different enough to show detectable differences in outcome” (p. 198). This method enabled us to compare the top third with the bottom third on this global assessment, and is consistent with other published research (White, Tynan, Galinsky, & Thompson, 2004). Use of this methodology for creating discrete groups required the use of a Bonferroni correction for a more conservative alpha of p < .01. The potential problem of attenuation was clearly not an issue here because all the t test findings were statistically significant below that level.

Cell sizes that are very uneven represent a potential threat to the validity of t test findings. Warner (2008, p. 161) suggested that this concern can be addressed by examining the results of Levene's test and by setting a more conservative alpha level. In this case, Levene's test indicates equal population variances, and statistical significance was established at p < .001.

The neutrality subscale does not seem to function as Gibb (1961) hypothesized. It had the lowest Cronbach's alpha (.65) and was the only defensive communication behavior not negatively valenced, as predicted. One possible explanation is that because academics are highly educated, competent individuals, they prefer and are used to high levels of autonomy. Therefore, some professional and interpersonal distance may be both desirable and expected. Second, this may be an artifact of changing times. The patriarchal model of a beneficent employer is neither desired nor expected as much as it was when Gibb first articulated his theory (Fairtlough, 2007).

Additional information

Notes on contributors

Kathleen Czech

Kathleen Czech (Ed.D., University of San Diego, 2007) is a professor in the Department of Communication & Theatre at Point Loma Nazarene University.

G. L. Forward

G. L. Forward (Ph.D., The Ohio State University, 1994) is a professor in the Department of Communication & Theatre at Point Loma Nazarene University.