The Effect of Negative Workplace Gossip about Supervisor on Workplace Deviance and Impression Management: The Mediating Roles of Anxiety and Guilt

Abstract

Drawing upon cognitive-motivational-relational theory, we offer a framework for understanding the mechanism underlying the effect of negative workplace gossip about supervisors (NWGS) on employees as gossip senders. We propose that NWGS can elicit two negative emotions, namely, anxiety and guilt, which further lead to employee behaviors that reflect avoidance and approach tendencies, respectively. We first tested these proposals in a survey study of graduate students (Study 1) and then replicated our findings among employees in the workplace (Study 2). In both studies, we found that NWGS is positively related to both anxiety and guilt, which in turn are linked to further avoidance-related workplace deviance and impression management directed toward the supervisor, respectively. Our results also show that the effects of NWGS on employees’ impression management through guilt are significant only when the employees score high on supervisors’ approval contingent self-worth. We discuss the implications of these results for managing workplace gossip in organizations.

Appendix. Additional Experimental Study

Description of the Additional Experimental Study

We conducted an additional experimental study to test the hypothesized directionality of the relationships between NWGS and our target emotions (i.e., anxiety and guilt). As noted in our literature review, employees might engage in NWGS as a consequence of their negative emotions (Brown et al., 2005), which suggests a reverse causality between NWGS and emotions. However, the cross-sectional nature of our Study 1 and Study 2 prohibits ruling out this possibility. Therefore, we used an experimental design in this additional study to test the causal effects of NWGS on emotions. It should be noted, however, that we did not test the causal effect of the two emotions on the two behaviors for the following reasons. First, CMRT posits that individuals’ emotions, with their innate action tendencies, lead to subsequent behaviors, not vice versa. Second, previous experimental studies have supported the causal effect of anxiety on avoidant behaviors (Cryder et al., 2012), as well as that of guilt on reparative behaviors (Choe et al., 2019).

To examine the proposed causal effect of NWGS on gossip senders’ anxiety and guilt, we manipulated NWGS to examine whether engaging in NWGS would enhance participants’ levels of anxiety and guilt.

Participants, Materials, and Procedures

We recruited graduate students with science and engineering majors from two universities in Beijing, China. Since the relationships between graduate students and their advisors are similar to those between employees and their supervisors in the workplace, we believe that this affords an appropriate context with which to test our hypotheses using a sample of graduate students. Each participant was asked to bring in a partner to the experiment. Each partner was in the same major at the same school as their related participant. We manipulated NWGS by asking each participant to gossip about his or her advisor with his or her partner during the experiment. The total sample size was 146. The ages of participants ranged from 21 to 29 (M = 23.38, SD = 1.55), and 13.70% of them were male. While 83.56% of them were master students, 16.43% of them were doctoral students. Participants majored in a variety of subjects, with the most prevalent being Environmentology (21.92%), Education (13.70%), Geography (9.59%), Biology (7.53%), and Chemistry (6.85%).

We adopted a one-factor between-subjects design (gossip vs. leisure talk). The participants and their partners were randomly assigned to either the gossip group (N = 78) or the leisure talk group (N = 68). On arrival at the laboratory, the participant and his or her partner in the gossip group received the following instruction:

“For a better understanding of graduate students’ daily life, our study invites you to talk to each other about your relationship with your advisor, specifically, the negative aspects of your relationship with your advisor. You can begin by asking each other the following questions. You are also welcome to talk about other things regarding this topic. Your talk will last for 12 minutes.

• Q1 Are you dissatisfied with your advisor?

• Q2 Have you ever had any doubts about your advisor’s competence?

• Q3 Have you ever complained about your advisor in private?

• Q4 Have you ever criticized your advisor in private?

• Q5 Has your advisor ever done anything that you thought was unfair or inappropriate?

• Q6 Have you ever heard other students complain about your advisor?”

For the leisure activities group, the instructions were as follows:

“For a better understanding of graduate students’ daily life, our study invites you to talk to each other about your leisure time activities. You can begin with asking each other the following questions. You are also welcome to talk about other things regarding this topic. Your talk will last for 12 minutes.

• Q1 Are you dissatisfied with your leisure time activities?

• Q2 What do you usually do in your leisure time?

• Q3 Have you ever been to any great restaurants in your leisure time?

• Q4 Have you ever exercised in your free time?

• Q5 To you, what would be an ideal plan for spending your free time?

• Q6 Are you aware of other students' leisure time plans?"

Immediately after the 12-min talk, the participants were asked to briefly report their discussion to the experimenter. Then, they were asked to complete questionnaires that included a manipulation check for NWGS (using the NWGS scale from the main studies) and measures for anxiety and guilt (using the same scales as used in the main studies). Then, they were thanked and debriefed.

Results

An independent sample t test was first conducted to examine the effectiveness of the NWGS manipulation. The results showed that the gossip group (M = 22.88, SD = 8.09) reported more NWGS than the leisure-activity group (M = 10.51, SD = 4.44; mean difference = 12.37, 95% CI [10.266, 14.473], t (122.71) = 11.64, p < 0.01, Cohen’s d = 2.10), indicating successful manipulation. Then, we conducted another independent sample t test to test the effects of NWGS on emotions. In line with our H1 and H2, participants in the gossip group (M_anxiety = 7.15, SD_anxiety = 2.95; M_guilt = 4.19, SD_guilt = 1.51) experienced significantly higher levels of anxiety (mean difference = 1.01, 95% CI [0.067, 1.947]) and guilt (mean difference = 0.59, 95% CI [0.111, 1.068]) than those participants in the control group (M_anxiety = 6.15, SD_anxiety = 2.77; M_guilt = 3.60, SD_guilt = 1.39; t_anxiety (144) = 2.11, p < 0.05, Cohen’s d = 0.35; t_guilt (144) = 2.44, p < 0.05, Cohen’s d = 0.41). Therefore, these results further supported the causal effects of NWGS on gossip senders’ guilt and anxiety.