«UNDERSTANDING EMPLOYEES’ BEHAVIORAL REACTIONS TO AGGRESSION IN ORGANIZATIONS by MARIE S. MITCHELL B.S., George Mason University, 1993 M.A.H.R., ...»
The different findings for displaced aggression from the survey and experimental study suggest the context in which the perceived aggression takes place may influence to the target of displaced aggression. The experiment involved relationships that were relatively short-term and students’ grades were based on assessments made by both the instructor and team members. In this class, students were told they were to assess the overall distribution of team grades at the end of this semester. Because of this, I argued that students might not consider displacing aggression against team members for fear they might retaliate in the future (specifically, with the amount of points they received for the team grade). Further, the instructor did not consider how the students treated one another as a basis of the grades they would receive. This setting differs from a usual work context. In general, supervisors are the only individuals responsible for assessing employees’ work performance. Thus, if a supervisor saw or learned that an employee displaced aggression on others (the organization, coworkers and customers) it would most likely reflect the supervisor’s assessment of the employee’s work performance. Hence, employees would be less likely to displace aggression within an organizational context because they would not want to provide the aggressive supervisor further reasons to be punitive and aggressive, particularly when they feared retaliation from that supervisor.
Indeed, when fear of retaliation was low, victims of perceived aggression were more likely to displace aggression. In the classroom setting, however, students had to worry about maintaining relationships with both the instructor and their team members because both had decision authority over their grades. Because team members were not the source of perceived aggression, they did not displace aggression (or may not have even considered displacing aggression, whether fear of retaliation from the supervisor was high or low) against team members for fear of future recourse. The different context, therefore, may explain why fear of retaliation negatively influenced the relationship between perceived aggression and displaced aggression in the work setting but not the classroom setting.
Altogether, the results for both studies support Tedeschi and Felson’s (1994) observations that some forms of displaced aggression are better explained by learning effects rather than cathartic reactions. In both studies, individuals discriminated when to or not to displace aggression. Also, they discriminated to whom to displace or not to displace aggression. The results of both studies suggest individuals might be less likely to displace aggression against targets when doing so will lead to future recourse—whether from the source of the initial harm or by the target of the displaced aggression. Further, individuals seemed to displace aggression in response to perceived aggression when doing so was safe (meaning the target would be less likely to seek retribution) and the target was similar to the source of the harm. Therefore, future research might consider a variety of different forms of displaced aggression (e.g., based on the safeness or similarity to the aggressor), as well as different work context (e.g., 90 degree performance appraisal systems, 360 degree performance appraisal systems, versus only supervisor to employee performance appraisal systems) to further explore these ideas.
Although fear of retaliation did not moderate the relationship between perceived aggression and constructive problem-solving in the experimental study, it did moderate the effects in the survey study. The results, however, did not support my predictions that fear of retaliation would weaken constructive problem-solving reactions. Rather, the results shows that fear of retaliation enhanced rather than weakened constructive problem-solving behavior. In the experimental study, fear of retaliation did not moderate the relationship between perceived aggression and constructive problem-solving, but it did significantly and directly influence constructive problem-solving.
The differences in the survey and the experimental studies may be due to the lack of variety of constructive problem-solving activities assessed in the experimental study compared to those in the survey study, as well as the time frame involved in which individuals had to consider engaging in constructive problem-solving. In the experimental study, subjects were limited to only one form of constructive problem-solving (writing comments on the back of the exam) within a very limited time frame (an hour at most). If they had more time, they might have engaged in other types of constructive problem-solving (e.g., e-mailing the instructor, going to see the instructor). The survey study, in contrast, assessed different types of constructive problem-solving behaviors that individuals may have engaged in over the course of a year (e.g., asked the supervisor to clarify the problem, tried to change the situation to benefit all parties involved). Therefore, whether an individual engages in constructive problem-solving activities in reaction to a feared and aggressive individual might depend on the variety of choices they have in terms of constructive problem-solving, as well as how long they have to consider their options.
In general, the pattern of results may support research in the stress literature. Since fear of retaliation directly influenced constructive problem-solving in the experimental study and moderated the effects in the survey study may suggests high fear of retaliation enhanced stress reactions. The stress literature suggests individuals engage in coping behaviors to help buffer the negative effects of the stressful situation (such as dealing with an aggressive supervisor who you also fear). Constructive coping activities (like constructive problem-solving) allow the individual to try to resolve the problem in an effort to reduce the level of the stress (e.g., Folkman & Lazarus, 1985; Thoits, 1994). Therefore, the more stressful the situation is perceived, the more likely an individual would engage in constructive problem-solving.
The Role of Personality: The Influence of Trait Anger and Need for Social Approval Based on previous aggression research (see Perrowé & Spector, 2002, for a review), trait anger was predicted to heighten aggressive reactions to perceived aggression. I argued that because trait-angry individuals view a variety of situations as anger-provoking and have intense reactions to personal attacks, high trait-angry individuals would be more likely to react to perceived aggression with aggression. Consistent with these arguments, trait anger strengthened retaliatory reactions to perceived aggression in the survey study. However, trait anger did not strengthen retaliatory reactions in the experimental study. Rather, individuals who were high in trait anger seemed to maintain high levels of retaliation whether they received aggressive or nonaggressive comments on their exam. In contrast, and consistent with the survey results, individuals with low levels of trait anger found the perceived aggression unnecessarily hostile and retaliated, accordingly.
The differences in results from the survey and experimental studies suggest that perhaps different forms of aggression may heighten retaliatory motives more than other forms of perceived aggression. Many of the items in the perceived supervisor aggression measure asked about behaviors that might be experienced with others seeing them, suggesting perceived supervisor aggression is a particularly humiliating and public form of aggression. The aggressive comments on the students’ exams were only viewed by the student and not others in the class. Consequently, high trait-angry individuals might be more motivated to retaliate against an aggressive supervisor because they may view the behavior as a public and personal attack. Future research could investigate this idea by comparing how high (and low) trait anger individuals respond to aggression that varies based on its “publicness.” Need for social approval was also investigated in both studies. Although the experimental design did not show need for social approval moderated the relationship between perceived aggression and reactions, the survey study found need for social approval moderated the relationship with two forms of displaced aggression (coworker and customer). Consistent with my expectations, individuals with high need for social approval were less likely to displace aggression on coworkers and customers than those with low need for social approval in response to perceived supervisor aggression. Over both studies, however, need for social approval did not influence constructive problem-solving reactions. This was surprising because research on need for social approval suggests individuals high in need for social approval engage in more helpseeking and approval-seeking behaviors (e.g., Hill, 1991; Nadler, 1983). Thus, I argued that given these tendencies, they would be more likely to engage in constructive problem-solving because these types of activities generally involve seeking guidance about the perceived aggression in an effort to resolve the issue. However, in hindsight, the results make much sense.
Research also shows that individuals high in need for social approval also have the tendency to retreat from threatening and harmful situations (e.g., Exline, 1963; Mehrabian & Ksionzky, 1974; Terhune, 1968). Therefore, individuals with high need for social approval would avoid behaviors that would be threatening (such as seeking help from the source of aggression).
Other Patterns and Trends across Study 1 and Study 2 One pattern that emerged across the survey and experimental studies was that retaliation seemed to be a dominant response to aggression. This pattern begs the questions as to whether or not retaliation is a primary reaction, and whether the lack of findings for the other dependent variables might be explained by an order effect. For example, perhaps individuals initially try to retaliate; however, if they continue to be angry, they may then displace aggression or withdraw.
If they become worried or fearful, they might engage in constructive problem-solving. This is an issue for future research to consider.
A second and problematic trend across both studies was the lack of and mixed findings for many of the predicted relationships. Overall, the unexpected findings suggest aspects of the theoretical model deserve further consideration. Two issues, in particular, pose limitations to the conceptual framework: (1) measurement of the dependent variables and (2) the influence of nonbehavioral reactions to aggression (e.g., emotional and psychological reactions). With regard to measurement, it could be argued that the lack of findings resulted from the measurement of only limited forms of the dependent variables. Results for retaliation may have been more dominant because retaliation can take on one general form: intentionally harmful behaviors targeted against the source of the harm. In contrast, other forms of displaced aggression, constructive problem-solving and withdrawal might not have been measured and assessed in the studies. One example of this is displaced aggression. Post hoc tests of other forms of displaced aggression within the experiment suggest that, depending on the target of the displaced aggression, the predictions may or may not have been supported. The post hoc evaluations demonstrated that one form of displaced aggression (evaluations of the overall fairness of the Management Department) was consistent with Dollard et al.’s (1939) contentions. Therefore, other forms of displaced aggression (e.g., those targeted against family, strangers) might also support Dollard et al.’s arguments. Similarly, it could be argued that because I only assessed a limited form of constructive problem-solving and withdrawal, the results did not support the predicted relationships. For example, students may not have engaged in the form of constructive problemsolving assessed in the experiment, but they might have engaged in other forms (e.g., e-mailing the instructor or dropping by the instructor’s office). Further, for withdrawal, the survey design only focused on task withdrawal, but other forms of withdrawal may have been influenced (e.g., job withdrawal—leaving the organization, transferring, absenteeism).
To explore this further, I also conducted post hoc analysis with the survey study data on turnover intentions. In general, turnover intentions do not represent actually-enacted behaviors;
however, they are perceptions about whether or not the individual intends to remain in the organization. The results show that the fear of retaliation x perceived supervisor aggression interaction was moderately significant (β=.11, p.10). Figure 1 shows that the pattern of the slope is consistent with my predictions. Specifically, fear of retaliation strengthens the positive relationship between perceived supervisor aggression and turnover intentions, such that for high fear of retaliation, as perceptions of supervisor aggression increased, intentions to leave the organization also increased. Therefore, the lack of findings may be explained by the limited forms of the dependent variables that were assessed in each of the studies.