«UNDERSTANDING EMPLOYEES’ BEHAVIORAL REACTIONS TO AGGRESSION IN ORGANIZATIONS by MARIE S. MITCHELL B.S., George Mason University, 1993 M.A.H.R., ...»
Figure 1 shows the interaction exhibited a pattern consistent with Hypothesis 1(b). In the low fear of retaliation condition, the relationship between perceived aggression and retaliatory aggression was positive; such that subjects in the no aggression condition engaged in significantly lower levels of retaliation than subjects in the aggression condition. Although the pattern of the slope suggests a negative trend for high fear of retaliation, the means between no aggression/high fear of retaliation and aggression/high fear of retaliation were not significantly different (see Table 4). Overall, the results provide support for Hypothesis 1(b).
perceived aggression and (a) retaliatory aggression and (b) displaced aggression such that the relationship will be stronger when trait anger is high than low, but that the positive relationship between perceived aggression and (c) constructive problem-solving will be stronger when trait anger is low than high. Contrary to my predictions, the trait anger x perceived aggression interaction was not significantly related to displaced aggression or constructive problem-solving.
Therefore, Hypothesis 2(b) and 2(c) were not supported.
The results show the trait anger x perceived aggression interaction was significantly related to retaliatory aggression. Figure 2 shows the interaction does not exhibit a pattern consistent with Hypothesis 2(a). The relationship between perceived aggression and retaliation was positive when trait anger was low, such that as perceived aggression increased, retaliatory aggression increased. However, the relationship was not significant when trait anger was high.
Therefore, Hypothesis 2(a) was not supported.
Hypothesis 3(a-c) predicts locus of control will moderate the positive relationship between perceived aggression and (a) retaliatory aggression and (b) constructive problemsolving such that the relationship will be stronger when locus of control is high than low, but the positive relationship between perceived aggression and (c) displaced aggression will be stronger when locus of control is low than high. Contrary to the predictions, the perceived aggression x locus of control interaction was not significantly related to any of the dependent variables.
Therefore, Hypothesis 3(a), 3(b), and 3(c) were not supported.
Hypothesis 4(a-c) suggests need for social approval will moderate the positive relationship between perceived aggression and (a) constructive problem-solving such that the relationship will be stronger when need for social approval is high than low, but the positive relationship between perceived aggression and (b) retaliatory and (c) displaced aggression will be stronger when low need for social approval is low than high. Contrary to the predictions, the results show the perceived aggression x need for social approval interaction was not significantly related to any of the dependent variables. Therefore, Hypotheses 4(a), 4(b) and 4(c) were not supported.
This study sought to understand individuals’ reactions to perceived aggression. Although some theorists contend aggression is a primary reaction to perceived aggression (e.g., Dollard et al., 1939; Lorenz, 1962), we also know individuals may not always react aggressively (e.g., Keashly et al., 1994; Tripp & Bies, 1997). Aggression research suggests what influences aggressive versus non-aggressive reactions are restraining factors and characteristics of the victim (see Anderson & Bushman, 2002 and Baron, 2004, for reviews). The results of this study suggest that fear of retaliation and trait anger moderate the relationship between perceived aggression and retaliatory reactions. However, only fear of retaliation influenced the relationship as predicted. Further, the results did not provide support for the moderating effects of locus of control or the need for social approval. I discuss the findings in more detail below.
Fear of Retaliation as a Situational Moderator of Reactions to Perceived Aggression The theory of frustration—aggression (Dollard et al., 1939) suggests fear of retaliation can influence aggressive reactions to aggression. Accordingly, victims of aggression who fear retaliation from the harmdoer refrain from retaliating against the source of harm. Doing so, however, causes individuals to become highly frustrated and, consequently, displace their aggression on other targets. The results of this study provide some support for these contentions, but not fully. Specifically, fear of retaliation influenced retaliatory reactions to perceived aggression. As predicted, the relationship between perceived aggression and retaliatory aggression was stronger when fear of retaliation was low rather than high. However, high fear of retaliation did not significantly weaken retaliatory reactions to perceived aggression, as suggested by Dollard et al. Rather, the differences in the means between the no aggression and aggression conditions for high fear of retaliation were not significantly different. Nevertheless, the results do suggest that when fear of retaliation was low, students were more likely to respond with retaliation in the aggression condition compared to the non-aggression condition.
The difference in the means in the no aggression condition suggests the high fear of retaliation manipulation might not have instigated high fear, as intended. Specifically, individuals in the no aggression/high fear of retaliation condition responded similarly to individuals in the high aggression condition. It appears the fear of retaliation manipulation elicited responses similar to the aggression manipulation. Rather than elicit fear of retaliation, the manipulation appears to be interpreted as another form of aggression.
High fear of retaliation was manipulated by telling the students that they were required to write their names on the Instructor Evaluation form and telling them that their instructor would see the evaluations within the next day or so. In general, students are used to filling out Instructor Evaluations at the end of the semester. The end of the semester evaluations are anonymous and students know that instructors will not see responses until after semester grades are posted. Thus, the sharp contrast in the instructions may have angered (rather than incited fear in) the students because the Instructor Evaluation instructions were particularly different in the experiment versus at the end of the year. In short, perhaps the high fear of retaliation manipulation did not adequately inhibit retaliatory reactions because the instructions were interpreted as another form of aggression.
The experiment was not successful in demonstrating the moderating effects of fear of retaliation on the relationship between perceived aggression and displaced aggression. The lack of findings for fear of retaliation of displaced aggression is perplexing, particularly given its support in social psychology experiments (cf., Marcus-Newhall et al., 2002). One explanation for the non-finding may be that the relationships students hold with fellow team members was not worthy of ruin. Displaced aggression in this study would provide for catharsis, meaning students could release their negative and aggressive energy on their fellow team members.
However, both Miller (1941) and Sears (1941) argued that individuals may not choose to engage in aggression if learned experiences suggest it is not an appropriate response. Similarly, in a review of the frustration—aggression theory, Tedeschi and Felson (1994) argued that research has not consistently demonstrated cathartic effects of displaced aggression; rather some forms of displaced aggression are better explained through learning effects. In particular, they argued that experience and learned history allow individuals to discriminate when to be or not to be aggressive, suggesting displaced aggression can be a cognitive reaction as well.
In this study, the experiment took place relatively early in the semester (after the first exam). This was purposeful in order to ensure students were not previously exposed to the instructor’s exam grading style. However, team projects were only just starting to gear up for the semester. Perhaps students refrained from aggressing against fellow team members because they do not want to jeopardize the emerging relationships with their team members, particularly in a class that allows students to decide on the overall distribution of team grades to individual team members. If students understand from past experience that aggressing against team members would only hurt them in the long run (i.e., they feared retaliation from fellow team members), then they may refrain from displacing aggression. Therefore, students may not have displaced aggression because they did not believe they have a safe target for their displaced aggression.
With this in mind, I conducted post-hoc analysis to explore whether or not individuals might displace aggression on other targets—safer targets. Specifically, in the study’s postexperimental questionnaire, students evaluated some of the Management Department’s services (i.e., quality of books and classes). If students wanted to vent their hostilities, they may have safely displaced their aggression toward the Management Department by rating their services poorly. Therefore, I tested the moderating effects of fear of retaliation on the relationship between perceived aggression and other forms of organization-targeted displaced aggression (evaluations of the quality of books, quality of the classes offered, and overall fairness of the department). The regression results show the statistical models were not significant for evaluations of the quality of the classes (F=.70, n. s.) or books (F=1.37, n. s.). However, the statistical model for the analysis of overall fairness of the Management Department was significant (R2=.31, Adj. R2=.19, F=2.59, p.01), and the results show the perceived aggression x fear of retaliation interaction was significant at the p.10 level (b=.47, se=.27, β=.35, p.10).
Pedhazur and Schmelkin (1991) argue that interactions at the p.10 level are worthy of further exploration. Therefore, the plotted interaction for overall fairness evaluations is shown in Figure
3. The pattern provides support for the prediction that the positive relationship between perceived aggression and displaced aggression will be stronger when fear of retaliation is high than low. As perceived aggression increased, subjects in the high fear of retaliation condition evaluated the overall fairness of the Department more poorly (i.e., displacing aggression) than subjects in the low fear of retaliation condition. Although overall fairness of the Management Department is not an exact measure of overall assessments of the Management Department, they do represent a “safer” target of displaced aggression. The results, therefore, provide some support for Dollard et al.’s contention that fear of retaliation from the source of the harm may enhance reactions of displaced aggression to perceived aggression.
The lack of findings for the other, “safer,” forms of displaced aggression (i.e., evaluations of the quality of classes and books) compared to the findings with overall fairness displaced aggression is also worthy of further comment. The differences in the findings suggest it might also be similarity of the target of displaced aggression to the source of the perceived aggression that affects individuals’ displaced aggression. Research on displaced aggression in social psychology has demonstrated that individuals are more likely to displace aggression against targets that are more similar to the harmdoer (see Marcus-Newhall et al., 2002, for a review).
This may be the case here; the overall fairness measure asked students to evaluate the fairness of treatment provided by Management Department instructors (i.e., “Overall, I’m treated fairly in courses” and “For the most part, instructors treat students fairly”). Therefore, subjects may have rated the overall fairness of the Management Department more poorly because the treatment described was treatment by instructors, in general, in the Management Department—a similar target to the source of the perceived aggression in this study (the instructor who wrote the comments on their exam).
perceived aggression. Rather, fear of retaliation was positively and directly related to constructive problem-solving, suggesting fear of retaliation enhanced these reactions. These findings are in contrast to the sexual harassment and whistleblowing literatures, which suggest that individuals who fear retaliation from the harmdoer refrain from trying to reconcile or resolve the problem with that individual. However, the nature of this experiment was slightly different than that of sexual harassment or whistleblowing (where an individual is dealing with illegal and potentially immoral acts against them). Students may have engaged in constructive problemsolving in this setting because the setting of the experiment itself was a learning environment (i.e., a class setting), where the role of the instructor is to provide feedback on performance.
Therefore, if students wanted to understand more about how they performed on their exam, regardless if they knew the instructor was going to see the completed instructor evaluation, they might ask the instructor about the exam.
One last comment on fear of retaliation is worth noting. The difference between the means of the high and low fear of retaliation conditions was relatively small, although statistically significant. Therefore, the manipulation may not have evoked considerable fear of retaliation from those in the high fear of retaliation condition to influence reactions to perceived aggression. Specifically, had the differences in means been larger, fear of retaliation may have more strongly influenced the relationship between perceived aggression and retaliatory aggression. Similarly, there may have been effects for displaced aggression. Lastly, had individuals more strongly felt fear of retaliation from the instructor, they may have been less inclined to engage in constructive problem-solving behaviors.