UNDERSTANDING THE AGGRESSIVE WORKPLACE: DEVELOPMENT OF THE WORKPLACE AGGRESSION TOLERANCE QUESTIONNAIRE
Coombs, W Timothy
This research reports the results of two studies designed to develop and assess the psychometric properties of the Workplace Aggression Tolerance Questionnaire (WATQ). The WATQ items assess perceptions of the tolerance of workplace aggression, the perceived appropriateness of the use of aggressive behaviors in organizations. The WATQ is offered as a means of assessing and possibly tracking a wide array of workplace aggression behaviors, including physical and verbal, active and passive, and direct and indirect forms. The first study examines the basic psychometric properties of the instrument. The second study provides evidence for its convergent and discriminant validity.
Fueled by mass media stories and statistics from governmental and labor sources, workplace aggression has begun to enter the consciousness of those in and studying organizations. The problem is now recognized as a global concern. The International Labor Organization’s workplace violence report to the United Nations confirmed workplace aggression was a problem all around the world, citing significant concerns in France, Argentina, Romania, and Canada (Fisher, 2001; Job, 1998; UN, 1998). Murder and physical assaults dominate the news media coverage of workplace aggression but distort its reality. Most instances of workplace aggression are verbal and passive in nature, such as talking behind someone’s back, spreading rumors, or giving someone the silent treatment (Baron & Neuman, 1996; Foran, 2001; Geddes & Baron, 1997; Perry, 2000; Stone, 1995).
Granted, shooting a co-worker is much more severe than spreading a rumor as it inflicts both physical and emotional pain. Rightly workplace violence programs have targeted and reduced the overt physical and verbal forms of workplace violence through their policies and prevention programs (Most effective, 2003). However, all forms of aggression, whether overt or covert, hold psychological ramifications for the targeted employee and productivity implications for the organization. Researchers repeatedly find that the covert (verbal and passive) forms of workplace aggression are prevalent in the workplace and exact a toll on individuals (Johnson & Indvik, 2001; Pearson, Andersson, & Porath, 2000). Efforts to examine and redress workplace aggression must address these verbal and passive variants in addition to current efforts targeting the physical, overt forms.
Communication scholars have been slow to bring their skills to the study of workplace aggression. Although they have examined related constructs such as verbal aggressiveness (Infante & Wigley, 1986) and employee emotional abuse (LutgenSandvik, 2003), they have not focused on workplace aggression directly. One of the largest and most widely publicized research projects involving communicative forms of workplace aggression was conducted by management scholars (Pearson et al., 2000). Although communication researchers have begun to examine emotions and emotional expression in the workplace (e.g., Kramer & Hess, 2002; Waldron, 2000; Waldron & Krone, 1991), scant attention is given to the topic of workplace aggression in the communication literature. Our work on aggression parallels Kreps’ (1993) observations about sexual harassment. Kreps observed: “communication is the primary medium through which sexual harassment is expressed; it is the means by which those who are harassed respond to harassment; and it is the primary means by which policies for eliminating sexual harassment in the workplace can be implemented” (p. 1). His words ring true for the relationship between communication and workplace aggression. Most workplace aggression is enacted through communication, communication is a mechanism for responding to workplace aggression, and it is a primary means for delivering policies designed to eliminate workplace aggression. Due to the communicative nature of workplace aggression, it should draw more attention than it has from communication researchers.
Typical workplace aggression polices target the most overt forms such as physical attacks and direct verbal threats (How can, 2008; Prince, 2003; Williams, 2003). Experts argue that focusing on these overt forms of aggression shifts workplace aggression from overt to covert forms (Baron & Neuman, 1996; Bjorkqvist, Osterman, & Hjelt-Back, 1994). Employees simply chose to enact forms of aggression that are not covered in the policies. These covert forms of aggression retain an element of ambiguity that make them difficult to confront. The aggressor can claim he or she did not mean anything by the act or it was misinterpreted. In some instances employees may use covert forms of workplace aggression to resist problematic organizational polices and practices they feel they cannot openly challenge (Collinson, 1994; O’Connell-Davidson, 1994; O’Leary-Kelly, Griffin, & Glew, 1996).
Our approaches to combating workplace aggression need to become more holistic. We must convince employees to view all forms of workplace aggression as inappropriate and report their occurrence. Toward that end a viable instrument for assessing attitudes about the wide array of workplace aggression behaviors-workplace aggression tolerance-would be valuable. The assessment of workplace aggression tolerance offers a means of evaluating organizational success or failure at reducing workplace aggression tendencies. Benchmarking can establish a pre-intervention workplace aggression tolerance assessment against which later measures can be compared. The key question is: “Are organizational members perceiving workplace aggression as less appropriate after the interventions have been implemented?” A positive answer indicates success while a negative answer indicates failure of anti-aggression interventions.
Assessing workplace aggression tolerance is consistent with the calls in risk management and human resources to assess the potential for on-site violence. A 2000 survey conducted by the Risk and Insurance Management Society and the American Society of Safety Engineers found 62% of corporations had written workplace violence policies but 70% did not have a formal risk assessment (Katz, 2000). Part of a risk assessment includes using surveys to collect information about employee concerns and attitudes (Braverman, 1999). We argue that an assessment of workplace aggression tolerance is an essential element of risk assessment.
This article presents the results of research designed to develop a viable measure of workplace aggression attitudes. We call the instrument the Workplace Aggression Tolerance Questionnaire (WATQ). We begin the paper by defining the breadth of workplace aggression and the role of employee attitude in understanding it. The present research describes the development of an instrument for assessing an organization’s workplace aggression tolerance.
WORKPLACE AGGRESSION
Workplace aggression refers to a much broader set of behaviors than the mass media publicize in reports of workplace murder or physical assault (Andersson & Pearson, 1999; Neuman & Baron, 1997). Because most acts of workplace aggression involve verbal assault rather than murder or physical assault and are more covert than overt (Geddes & Baron, 1997; Incivility, 2000; Neuman & Baron, 1997), workplace aggression is more subtle than media reports would have us believe. Research needs to include these “lesser forms” of aggression if we are to truly understand the phenomenon. Communication scholars have been cited for understudying indirect interpersonal aggression, such as spreading rumors, which is a part of the spectrum of workplace aggression (Beatty, Velencic, Rudd, & Dobos, 1999). Conceptualizations of workplace aggression should encompass the full range of aggression present in the workplace and should not overlook verbal and indirect forms of aggression. Some behaviors associated with aggression are similar to behaviors reported in the growing literature on verbal and behavioral organizational dissent and resistance (e.g., Deetz, 1995; Kassing, 1998). Organizational dissent and resistance are actions that signal disagreement with the organization or its representatives. As is the case with aggression, these actions can be overt or covert.
Workplace aggression is defined as perceived intentional “efforts by individuals to harm others with whom they work, or have worked, or the organizations in which they are currently, or were previously, employed” (Neuman & Baron, 1997, p. 38). This broad definition includes verbal aggression, which are messages that are perceived to be hurtful, bullying, and workplace incivility, “low-intensity deviant behavior with ambiguous intent to harm the target” (Andersson & Pearson, 1999, p. 457; Neuman, 2000). While no physical harm is inflicted, communicative-based workplace aggression is still harmful as it can damage the psyche, work relationships, employee performance, and turnover (Bensimon, 1997; Geddes & Baron, 1997; Incivility, 2000; Kinney, 1994; Top threat, 2001). Tolerance of verbal aggression may even be a precursor to more extreme behaviors like forms of physical violence (Andersson & Pearson, 1999; DeRidder, Schruijer, & Rijsman, 1999). An understanding of organization members’ perceptions of the appropriateness (or acceptability) of various forms of workplace aggression is central to managers’ and researchers’ attempts to assess and reduce the problem.
One application could be to the growing trend in workplace aggression training that is focused on the early detection of potential problems. People in supervisory roles are trained to identify the symptoms of workplace aggression. Once identified, the employee is referred to an employee assistance program or a special program for workplace aggression (Magyar, 2003). Many experts claim there are always warning signs before physical aggression occurs in the workplace and data indicate that training to identify cues does reduce employee-on-employee physical aggression (Most effective, 2003). Many of the symptoms are the verbal and passive forms of workplace aggression such as negative comments towards others (verbal aggression) and interrupting people (Magyar, 2003). If employees dismiss these lesser forms of workplace aggression and see them as appropriate, the employees are likely to miss potential warning signs for more serious physical workplace aggression.
Extant research (Baron & Neuman, 1996; Geddes & Baron, 1997) suggests Buss’s (1961) classification system for aggression effectively captures the complexity and variety of workplace aggression. Buss’s (1961) system uses three dimensions to classify aggressive acts: (1) physical/verbal, (2) active/passive, and (3) direct/indirect. Physical acts involve deeds (e.g., kicking or shoving) while verbal acts involve the use of words to inflict harm (e.g., insults or harsh criticism). The active dimension reflects how harm is produced by performing a behavior (e.g., actual assault or insults) while the passive category indicates harm is created by withholding some action (e.g., not providing needed information). The direct dimension indicates how aggression can be perpetrated directly toward the target (e.g., insulting a person to his or her face) while the indirect aspect denotes how aggression may be expressed through an intermediary or by attacking something valued by the target (e.g., spreading rumors behind his or her back). Baron and Neuman (1996) used Buss’s (1961) three dimensions to create a typology of eight workplace aggression categories: (1) physical-active-direct (PAD), (2) physical-active-indirect (PAI), (3) physical-passive-direct (PPD), (4) physical-passive-indirect (PPI), (5) verbal-active-direct (VAD), (6) verbal-active-indirect (VAI), (7) verbal-passive-direct (VPD), and (8) verbal-passive-indirect (VPI).
Previous workplace aggression research (Baron & Neuman, 1996; Geddes & Baron, 1997; Neuman & Baron, 1997) has found examples from all eight categories of aggression in the workplace. However, the underlying factor structure of the system has not been examined. Hence, the eight-category system has never been tested to determine if people actually discern between the types of aggressive behaviors and if they perceive these behaviors to be inappropriate in the workplace.
There are reasons to believe that a factor analysis of the eight category system will not yield a distinct factor structure. The categories are intertwined due to the crossing of three dimensions. We will illustrate this overlap using the PAI items. The PAD and PAI differ on direct/indirect dimension but share the physical and active dimensions. The PAI items share the physical dimension with the PPI items. Furthermore, the PAI items share the active dimension with the VAD and VAI items. It would be difficult to get a clean factor loading on any of the three dimensions because each item represents three separate dimensions. What binds all of the items together is the belief that each taps the workplace aggression construct.
There is a surprising dearth of information concerning people’s perceptions of the appropriateness of various forms of aggression. Employee perceptions of aggression have significant implications for research and aggression prevention programs in organizations. It is possible that employees perceive some forms of aggression as relatively benign and accept its occurrence in the workplace. For instance, “mild” forms of aggression (e.g., purposefully working slowly, giving dirty looks to a manager) may be perceived as appropriate while more extreme forms (e.g., verbally threatening a manager, hitting or kicking a manager) may be viewed as inappropriate. Because the expression of workplace aggression can damage workers’ morale, affect mental and/or physical health, and may lead to lawsuits against the organization, it is clear that a “healthy” workplace climate-one that discourages aggression-should be cultivated (Braverman, 1999; Kinney, 1994; Lambert, 1995; Levin, 1995). Moreover, if workplace aggression is instigated by management personnel, it is likely to be seen as acceptable and modeled by other organizational members and become socially contagious (O’Leary-Kelly et al., 1996; Robinson & O’Leary-Kelly, 1996). In some cases there may be a “mis-match” between employees’ and the organization’s guidelines for appropriateness. Programs designed to reduce workplace aggression, such as education programs and preventative practices, would be needed to bring employees’ perceptions into alignment with what is considered appropriate workplace behavior. Dougherty (2001) offers a similar argument concerning sexual harassment training. Organizations should consult with professional organizations, workplace aggression experts, and government recommendations to determine what should constitute appropriate and inappropriate behaviors.
Neuman and Baron (1997) articulate a model of workplace aggression that helps us to understand the events and interpretations that precede the decision to engage in aggressive or non-aggressive behavior. The model holds implications for identifying possible points of intervention. Their model identifies adverse antecedent conditions (stress, frustration) which lead to employee cognitions, affect, and arousal. These lead to an interpretation of the experience – what has occurred and what will occur in the future (primary appraisal) – and then an assessment of what can and should be done to deal with the event (secondary appraisal). It is in this step that the employee considers possible responses to the adverse event. The employee will consider three factors: (1) the consequences of aggression, (2) alternative coping strategies, and (3) additional information. The two solutions noted at the beginning of this paper, policies and training, mesh well with the model. Zero-tolerance shapes consequences; you engage in an action and you are fired. Training teaches alternative coping strategies such as various forms of dispute resolution. We posit that shaping attitudes is an essential part of the process as well. It is not enough to discourage some forms workplace aggression with punishment if other forms persist. Employees should consider all forms of workplace aggression as very inappropriate if the problem is to be addressed in its entirety. However, no realistic, enforceable zero-tolerance policy could encompass all forms of workplace aggression. Also, it is not enough to teach employees behavioral options. They must view those options as more appropriate than any form of workplace aggression. The first step toward understanding the effects of workplace aggression attitudes is to develop a useful measure of the concept.
DEVELOPMENT AND TESTING
The development and validation of the Workplace Aggression Tolerance Questionnaire (WATQ) involved two separate studies. Study 1 examined the basic psychometric properties of the newly-developed WATQ scale, including the underlying structure of the WATQ and item evaluation. Study 2 incorporated modifications and concentrated on the convergent and discriminant validity of the WATQ.
STUDY 1
Participants
Participants were 208 people (57% female and 43% male) who had at least 3 years of work experience and currently were not enrolled in college. Participants were solicited by college students at a moderate-sized Midwestern university. Students were given extra credit for finding qualified individuals to complete the questionnaire. The selection procedure was undertaken in order to create a non-student, work-experienced pool for the study. Students were supplied with a copy of the survey along with a cover sheet and written directions to be given to the respondents. Students were informed that the purpose of the survey was to learn about how people perceive behaviors in the workplace. They were instructed to locate non-student respondents who were willing to complete the surveys. Each respondent supplied contact information in the form of name, address, telephone number, and e-mail (if they had one). Only responses with complete contact information were used in the data analysis.
To verify participation, the contact person was called and asked if they had completed the survey. Respondents were told they might be contacted to verify their participation in the study. A graduate research assistant randomly selected 10% of the responses (n = 21) for verification. Of the 21 responses, 18 provided e-mail addresses. E-mail was used to set up a time to call those 18 individuals. Of the 18, 14 were reached on the first call while four required a second e-mail appointment and call. The procedure for the other three respondents was to use up to five attempts to reach the individual by phone. If the phone number was invalid or an individual could not be reached on the fifth call, that person would be replaced with another randomly selected respondent. One of the remaining numbers was no longer in service so it was replaced. One person was reached on the first call, one on the fourth call, and one on the fifth call. The verification rate was 100% for the final sample. The 208 participants exceeds Nunally’s (1978) recommendation of having a sample of at least five times the number of questionnaire items when examining the psychometric properties of an instrument (28 items were included on the initial version of the WATQ).
When developing his organizational dissent scale, Kassing (1998) chose to use respondents from a variety of organizations. His concern was that dissent could vary by organizational type. He selected a broad spectrum of organizations to avoid this problem while testing the basic psychometric properties of his scale. Later, the organizational dissent scale was applied to specific organizations to assess levels of organizational dissent (e.g., Valentine, Young, Bailey, Barhoum, LaBure, Glover, & Isaac, 2001). We followed a similar logic in our development of the WATQ. Because we believe that the perceived appropriateness of workplace aggression can vary by type of organization and from organization to organization, it was desirous to have a wide variety of different organizations in the initial assessment of the instrument and not just collect data from one or two organizations. Using students to find subjects casts a wide net so we believed a range of organizations would be found through this process The array of organizations included in the sample were: service, 25%; professional, 21.2%; education, 17.3%; retail, 10.1%; sales, 9.1%; manufacturing, 7.2%, construction, 5.3%; government, 3.4%, and financial, 1.4%.
The participants ranged in age from 20 to 61 years (M = 35.00, SD = 12.32). Work experience ranged from 3 to 45 years (M = 15.00, SD = 10.59). Around 25% of the sample had worked for 5 years or less while 30% had worked for 30 or more years.
Procedures
The participants were asked to complete a questionnaire concerning workplace behavior. The first section requested demographic information about age, sex, work tenure, and type of work. Next, participants were asked to read a workplace scenario involving a worker receiving negative performance feedback that was inaccurate. The worker discusses the inaccuracies with the manager but the manager refuses to make any corrections. It was noted that the inaccurate evaluation could affect potential raises and/or promotions for the worker. Using a single scenario as the stimulus to examine perceptions and responses is consistent with previous work on sexual harassment (Bingham & Burleson, 1996). The scenario selected for this investigation is similar to those used in previous studies of workplace aggression where a “triggering” situation is presented to respondents (e.g., Geddes & Baron, 1997).
Negative performance feedback involves messages which reflect unfavorably on a person’s task performance. Negative performance feedback was selected as a trigger because it has been linked to workplace aggression (Geddes & Baron, 1997; Geddes & Linneham, 1996). Inaccuracy was included to intensify the perceptions that the evaluation was unjust. Again, perceptions of injustice have been tied to workplace aggression (Geddes & Linneham, 1996; Mantel, 1994; Neuman & Baron, 1997; Skarlicki & Folger, 1997).
Our choice of scenario was guided by the desire to use a stimulus from previous research that was known to elicit workplace aggression (strong ecological validity). The selected scenario involves both negative feedback and perceptions of injustice, known triggers of workplace aggression in real-life settings. We wanted to select a strong stimulus but wanted to avoid scenarios involving retaliation for physical violence or overt threats. We selected a scenario that is likely to spawn workplace aggression without itself being a direct form of aggression. The manager in the scenario could simply disagree with the worker’s perception. Admittedly it is a strong stimulus but we felt it was warranted because of its ecological validity. If the stimulus was too mild, it could be too easy to dismiss. The stimulus needed to be strong enough to evoke thoughts about aggression. Inaccurate, negative performance feedback and injustice are recognized as strong motivators for workplace aggression. A copy of the stimulus is provided in the Appendix.
Measure
Research participants were asked to evaluate the appropriateness of 28 different worker responses to the interaction which comprise the Workplace Aggression Tolerance Questionnaire (WATQ). The items included on the WATQ were derived from Buss’s (1961) classification framework for aggression and past research applying his framework to workplace aggression (Baron & Neuman, 1996; Geddes & Baron, 1997; Neuman & Baron, 1997). Eight types of workplace aggression were represented by two to five items, with each item on the WATQ providing a different example of workplace aggression. The number of items on the WATQ reflects the number of examples provided by Neuman and Baron (1997). That is to say, the more examples given for a type of workplace aggression, the more items that appear on the WATQ to tap that type of workplace aggression. Table 1 lists the WATQ items organized according to Buss’ eight categories of aggression.
Participants were asked to indicate the appropriateness of each action in the workplace. Responses were recorded on a 5-point Likert scale ranging from “very inappropriate” to “very appropriate.” An evaluation of appropriateness was selected because it would reveal the extent to which attitudes in the workplace were conducive to peoples’ acceptance of aggressive acts. The scenario indicates that the workplace aggression would be directed at the manager, not the person answering the items. People could have a much different view of appropriateness if they were to be the target of the aggressive behavior. We were interested in perceptions of the general use of aggression in the workplace, not for aggression specifically targeting the respondent.
Item Evaluation and Reliability
Each WATQ item was examined to evaluate its contribution to the consistency of the scale. An item was eliminated if it reduced the Cronbach’s alpha to below .90. In addition, item-total correlations for the items were reviewed. The minimum criterion for item retention in the scale was .40. No item-total scale correlation fell below .40 and the scale demonstrated an acceptable inter-item correlation of .42. It was unnecessary to delete any items from the scale. The overall reliability for the WATQ was .95 (standardized Cronbach’s alpha).
Factor Analysis
A principal components factor analysis using oblique rotation was performed on the WATQ. If the WATQ contained sub-scales they probably would be related to one another, hence the use of the oblique rotation. (Varimax rotation was performed with essentially the same results.) Following the advice of Burgoon and Hale (1987), a set of four criteria were used to evaluate the factor structure: (1) eigen values had to exceed 1.0; (2) a factor had to have at least 3 items meeting the .50/.30 strength/purity standard for factor loadings; (3) the scree test had to show each additional factor was making a reasonable improvement in the variance accounted for; and (4) any item in a given factor had to have a primary factor loading of .50 or better.
Five factors having eigenvalues of 1 or better emerged and accounted for 67.8% of the total variance. The five factors accounted for 44.5%, 10.0%, 5.4%, 4.3%, and 3.6% of the variance, respectively. Because there was a great deal of cross loading between factors, only seven items could meet the .50/.30 strength/purity standard for factor loadings. The factors were not clearly interpretable. Moreover, the scree test criteria suggested a single factor solution should be retained. The analyses did not seem to justify converting the factors into sub-scales. Forcing a single factor solution resulted in factor loadings of .45 or higher which are acceptable (Stuckless & Goranson, 1992). Table 2 presents the factor loadings for the single factor solution as well as the means and standard deviations for each item. Overall, the analyses suggest that the WATQ is fundamentally uni-dimensional. All 28 items were retained for use in Study 2 because they met the item evaluation and reliability criteria.
Additional Analysis
A t-test was used to examine sex differences in responses. A composite WATQ score was created by summing the scores on the items and dividing by 28. The WATQ scores are ascending such that higher scores indicate that respondents perceive the aggression as more appropriate. There was a significant gender difference for the composite WATQ score (t(207) = 8.61, p
Another point to address was whether or not the results obtained through the WATQ were consistent with previous research that had measured aggressive behaviors in the workplace using the Buss categorization system. Two previous studies used the Buss categorization system to assess the perceived occurrence of workplace aggression acts. Both studies compared the reported incidence of verbal to physical behaviors, passive to active behaviors, and direct to indirect behaviors. Geddes and Baron (1997) used the Bliss category system to assess workplace aggression following negative feedback. Their sample reported significantly more verbal than physical behaviors and more passive than active behaviors. However, there was no significant difference between direct and indirect behaviors. Baron and Neuman (1996) examined the occurrence of workplace aggression in general (i.e., not limiting it to feedback sessions). Using Buss’s category system, they measured both witnessed and experienced workplace aggression. They found significantly more verbal than physical behaviors, more passive than active, and more direct than indirect behaviors. The finding for more direct than indirect behaviors was the only surprising finding, given that they had predicted the opposite result.
The analysis of the WATQ data revealed similar results when comparing the three behavior categories. The data were analyzed using t-tests. Following recommendations by Stevens (1980), the alpha level was set to a stricter .001 level to protect against the possibility of Type I error. The t-test results found verbal behaviors (M = 1.48, SD = .47) to be perceived as significantly more appropriate than physical (M = 1.35, SD = .61), t(208) = -8.98, p
Summary
Study 1 explored the basic psychometric properties of the WATQ. The item evaluation indicated all 28 items could be retained. The questionnaire demonstrated a solid .95 reliability. The factor analysis suggested the WATQ was uni-dimensional-all items seemed to be tapping a similar construct. In addition, the WATQ produced consistent results with past workplace applications of the Buss categorization. Verbal behaviors were found to be more acceptable than physical behaviors and passive behaviors were deemed more acceptable than active behaviors.
STUDY 2
Study 2 extends the research on the WATQ by examining convergent and discriminant validity. Moreover, the inclusion of additional measurement instruments affords the opportunity for a more extensive item evaluation.
Participants
The participants were 201 students enrolled in a moderate-sized, Midwestern university. The sample was comprised of 37% males and 63% females. The respondents ranged in age from 18 to 60 years (M = 23, SD = 6.55). Work experience ranged from 1 year or less to 40 years with a mean of 7 years work experience (SD = 6.00). Around 45% of the sample had worked for 5 years or less while 18% had worked for 10 or more years. As expected, the sample in Study 2 had less work experience than the sample in Study 1. However, Study 2 focused on convergent and discriminant validity issues with the WATQ so a student-based population was deemed viable.
Procedures
As in Study 1, participants were told the study involved workplace behaviors related to aggression. The same negative performance feedback scenario from Study 1 was used as the stimulus. Participation was voluntary and most students received some extra credit for their participation. Each participant was asked to read the scenario and the directions to each section. The survey had four parts: (1) demographic information, (2) the WATQ scale, (3) Vengeance Scale and Marlowe-Crowne Social Desirability Scale, and (4) Verbal Aggression Scale.
Measures
WATQ. Participants completed the 28-item WATQ instrument used in Study 1.
Social Desirability. It is important to consider the problem of social desirability with self-report data which involves negative behaviors. The Marlowe-Crowne social desirability scale is an accepted tool for evaluating this problem. A 13-item version of the Marlowe-Crowne Scale was used to assess social desirability (Reynolds, 1982). The reliability coefficient for the scale was .67 (Cronbach’s alpha). By dropping one item, the reliability coefficient was raised to .74 thereby meeting Nunally’s (1978) .70 level for acceptability. Therefore, 12 items were used to create the Social Desirability composite variable. The WATQ was anticipated to have no or a very weak correlation with the social desirability scale.
Vengeance Scale. Stuckless and Goranson’s (1992) 20-item vengeance scale measures people’s attitudes toward the appropriateness of taking vengeful acts and was used to assess vengeance. The reliability coefficient for the scale was .88 (Cronbach’s alpha).
Verbal Aggressiveness Scale. Verbal aggression is “a personality trait that predisposes persons to attack the self-concepts of other people instead of, or in addition to, their positions on topics of communication” (Infante & Wigley, 1986, p. 61). Infante and Wigley (1986) developed the Verbal Aggressiveness Scale (VAS) to measure this trait. A 10-item version of this scale, used in previous research, was employed to assess verbal aggression (Infante & Gorden, 1989; Infante, Myers, & Buerkel, 1994; Martin & Anderson, 1996). The reliability for the scale was .76 (Cronbach’s alpha). Both the vengeance scale and verbal aggressiveness scale were anticipated to correlate positively with the WATQ.
The Social Desirability, Vengeance, and Verbal Aggressiveness scales demonstrated acceptable reliabilities which were consistent with results from past uses of those scales. Given the acceptable reliabilities, composite variables were created by averaging the items on each scale.
Results
Because of the newness of the WATQ, item evaluation for item retention and factor analysis to examine possible sub-scales were performed to further examine its psychometric properties. Correlational analyses were used to test the convergent validity and discriminant validity of the instrument.
Item Evaluation and Reliability
A three-step process was used to evaluate item retention for the WATQ. A more extensive item evaluation was possible in Study 2 because of the inclusion of additional scales in the data collection instrument. In the first step, each WATQ item was examined to evaluate its contribution to the consistency of the scale. An item was eliminated if it reduced the Cronbach’s alpha to below .90. In addition, item-total correlations for each item were reviewed. The minimum criterion for retention in the scale was .40. It was unnecessary to delete any items from the scale. The reliability for the WATQ was .97 (standardized Cronbach’s alpha). The scale also demonstrated an acceptable mean inter-item correlation of .55 (Briggs & Cheek, 1986).
In the second step, tests for contamination for social desirability were conducted by checking correlations between the individual WATQ items and the Marlowe-Crowne Social Desirability Scale. An item was retained only if it had a correlation of less than .30 with the Social Desirability Scale. All of the correlations between the WATQ items and the social desirability items were less than .25.
In the third step, items were examined to see if their item-total correlation was lower than their correlation with either the Vengeance or Verbal Aggression Scales. An item was deleted if its item-total correlations were lower than its correlations with either of the other two scales (Jackson, 1970). The highest WATQ item correlation with another scale was .37, well below the lowest item-total correlation of .60. The results again suggested that all 28 items should be retained in the scale.
Factor Analysis
A principal components factor analysis using oblique rotation was performed on the WATQ. Once more the four criteria recommended by Burgoon and Hale (1987) for factor evaluation were employed. Three factors having eigenvalues of 1 or better emerged and accounted for 69.1% of the total variance. Factor one accounted for 56.8% of the total variance, factor two 7.5%, and factor three 5.6%. The scree test criteria suggested a single factor solution should be retained. Forcing a single factor solution resulted in factor loadings of .60 or higher which are acceptable (Stuckless & Goranson, 1992). Table 2 presents the factor loadings for the single factor solution as well as the means and standard deviations for each item. The analyses did not seem to justify converting the factors into sub-scales. Again, the analyses suggest that the WATQ is fundamentally a uni-dimensional measure.
Assessments of Validity
Tests for convergent validity should be performed when developing a new scale (DeVellis, 1991). The WATQ’s convergent validity was assessed by correlating it with two measures that addressed conceptually relevant variables, the Vengeance Scale (Stuckless & Goranson, 1992) and the Verbal Aggressiveness Scale (Infante & Wigley, 1986). Vengeance and verbal aggression are two variables that should be positively related to workplace aggression. A proclivity towards vengeance or verbal aggression could facilitate an acceptance of workplace aggression. The WATQ correlated with both the Vengeance Scale (r = .37, p
Discriminant validity was assessed using the 12-item short form of the Marlowe-Crowne social desirability scale (Reynolds, 1982). Using the 12-item short form is consistent with past use of the scale to assess concepts related to vengeance (Stuckless & Goranson, 1992).1 The WATQ correlated at r = -.18 (p
Summary
Study 2 extended the item evaluation and examined the convergent and discriminant validity of the WATQ. The extended item evaluation indicated all 28 items could be retained. The reliability for the 28-item version of the WATQ was computed to be .97 (Cronbach’s alpha). The results of the factor analysis suggest that the WATQ is uni-dimensional. The results indicated that the WATQ was not contaminated by social desirability as assessed by the Marlowe-Crowne scale. This small correlation between the instruments provides evidence for the WATQ’s discriminant validity. As expected, the WATQ was weakly correlated with the Vengeance and Verbal Aggression scales, lending support to the convergent validity of the WATQ scale. The WATQ items cover all of Buss’s (1961) eight categories of aggressive behavior. The 28 items reflect evaluations of the appropriateness of different acts of workplace aggression.
DISCUSSION
Overall, the original 28-item WATQ was found to be reliable and valid. The WATQ demonstrated strong reliability in both studies. Study 2 provided support for both convergent and discriminate validity. Basing the scale on previous applications of Buss’s (1960) aggression typology to the workplace adds to the WATQ’s face validity. All behaviors on the WATQ have been reported to occur in the workplace. The items were found to be uni-dimensional in both studies, supporting the belief that people perceive a common conceptual thread running through the eight types of workplace aggression. These results confirm the importance of recognizing a wide range of behaviors as representing forms of workplace aggression and as being inappropriate for the workplace. Workers need to understand it is not just physical assaults and verbal threats that comprise workplace aggression. In fact, most instances of workplace aggression will be more covert than overt. This is in contrast with the physical-direct-active forms of workplace aggression that have been hyped by the news media.
Moving beyond the sensational headlines of murder in the workplace leads to the larger subject of workplace aggression. Most workplace aggression is subtle and delivered verbally and passively, not in physical assaults. These covert forms of workplace aggression lack the physical violence associated with their high profile counterparts but are still a concern because they can create serious problems for individuals and organizations. Moreover, verbal aggression can be a precursor to violent actions (Andersson & Pearson, 1999; Infante, Chandler, & Rudd, 1989; Pearson et al., 2000). Understanding workplace aggression is enriched when communication insights are added to the mix. The full range of workplace aggression warrants careful study. A potentially rich topic is workplace aggression tolerance. Overall, the WATQ’s convergent validity and reliability indicate it can supply data relevant to perceptions of the appropriateness of a variety of behaviors associated with workplace aggression.
The results from Study 1 indicate that verbal and passive forms of aggression are perceived to be the most acceptable forms of workplace aggression. The results suggest people are most likely to view passive forms of aggression as appropriate in the workplace. Previous research indicates that such aggression is the most common form found in organizations (Andersson & Pearson, 1999). These subtle forms of aggression reflect the danger of workplace aggression “going underground.” Employees may avoid the obvious forms of aggression, ones typically covered in zero-tolerance workplace aggression policies. Why use a type of aggression that can lead you to be fired under a zero-tolerance policy? People naturally gravitate toward those types of aggression that are harder to detect and to punish. As Neuman and Baron (1997) warned, punishment simply suppresses direct forms of workplace aggression and increases the likelihood of more covert forms being used. If workplace aggression behaviors do shift to more subtle forms, the punishment doled out by most workplace aggression policies will have little effect on the overall incidence of workplace aggression. The high-profile behaviors will diminish but the problem of workplace aggression will persist. We can expect communication-based workplace aggression to rise as the overt, physical forms are strongly discouraged through organizational policies. The uni-dimensionality of the WATQ suggests it can be used to evaluate the perceived appropriateness of the gestalt of workplace aggression behaviors, including the too often ignored covert forms. The factor structure suggests the WATQ may be best-suited to assessments of employees’ overall tolerance of aggression.
Ideally, anti-aggression interventions should be directed toward attitude change and not the mere suppression of certain behaviors. Eliminating assaults and overt verbal abuse does not mean that workplace aggression has been eliminated or even discouraged. When developing programs to reduce aggression, it is important to explore possible underlying causes of aggressive behavior that are interconnected with other aspects of the organization. Managers need to understand how aggressive behavior may signal organizational problems such as stress, resistance to policies, and feelings of powerlessness and be perceived as functional coping mechanisms for some workers. Dougherty (2003) offers a similar argument in her analysis of how sexual harassment behaviors and training to eliminate harassment must acknowledge the functions and sensemaking surrounding the behavior. The effectiveness of training interventions will be enhanced when the problems and solutions are understood from a systemic perspective.
Through benchmarking, the WATQ can help establish if anti-aggression interventions have encouraged people to perceive aggression as a less appropriate option. The overall WATQ scores should decline after an effective intervention/training. The belief is that as people view aggressive acts as less appropriate, they may choose to redirect their aggression in more positive ways. Training in conflict management and forgiveness has been used successfully to create positive alternatives to aggressive actions (Prince, 2003; Williams, 2003).
Furthermore, the WATQ can identify problems within different organizational units. A unit with an especially high WATQ score warrants closer scrutiny to determine the source of the problem. Higher scores for particular units may reveal problems stemming from organizational factors including structural reporting relationships, managerial style, work load stress, burnout, and diversity-related tensions. Does the aggression reflect a type of dissent or resistance to organizational practices or policies that need to be re-evaluated? Organizations must also question how employees are being educated about the policies. To what extent are anti-aggression policies being communicated or implemented properly? Is aggression somehow being rewarded in the unit? Is modeling of aggression occurring? Do some norms in the unit seem favorable toward aggression?
There are several limitations to this research. First, this paper reports the initial stages of WATQ development. Additional research on its psychometric properties is warranted. Second, the scenario stimulus is limited to a particular type of role relationship, the superior-subordinate relationship. Respondents were asked to report their perceptions of “how appropriate” the various actions “would be in the workplace.” They were asked to evaluate the subordinate’s behavior in response to the performance appraisal scenario. The majority of the behaviors they reported on were directed toward the manager, with a few being directed toward the workplace itself (e.g., sabotaging the workplace, wasting resources). It is possible that different patterns of responses would be obtained with the inclusion of peer-to-peer relationships or superior-to-subordinate relationships where they report on a superior’s aggressive actions toward a subordinate. Third, respondents were asked to report their perceptions of what they see as “appropriate” “in the workplace.” It may be that their answers would differ if they were asked specifically to report on what they see as appropriate “in their workplace.” Future research could address these issues.
In sum, the WATQ is offered as a diagnostic tool for discovering possible tolerances for aggressive behavior in the workplace. Since finding the problem is a prelude to solving it, the WATQ can provide a useful service to organizations looking to improve the health of its workplace by reducing all forms of workplace aggression.
Communication scholars should make a more concerted effort to apply their expertise to the problem of workplace aggression. Our skills can provide new insights into the assessment and amelioration of this serious, global concern. The WATQ offers a step in that direction as it identifies problems which communication-based concepts can help to solve. As a field, communication has a great deal to offer the efforts to reduce workplace aggression. We can and should do more to realize that potential.
NOTE
1 When assessing the discriminant validity of the Vengeance Scale, Stuckless and Goranson (1992) found one item from the social desirability scale (“Sometimes I try to get even rather than forgive and forget”) correlated highly with the Vengeance Scale. Because the item was very similar to items on the Vengeance Scale, it was dropped from the analysis. A similar concern existed with the WACI items. The 12-item version of the Marlowe-Crowne Scale correlated r = .98 with the 13-item version of the Marlowe-Crowne Scale so the 12-item version was used for the analysis.
REFERENCES
Andersson, L. M., & Pearson, C. M. (1999). Tit for tat? The spiraling effect of incivility in the workplace. Academy of Management Review, 24, 452-471.
Baron, R. A., & Neuman, J. H. (1996). Workplace violence and workplace aggression: Evidence on their relative frequency and potential causes. Aggressive Behavior, 22, 161-178.
Beatty, M. J., Valencic, K. M, Rudd, J. E., & Dobos, J. A. (1999). A “dark side” of communication avoidance: Indirect interpersonal aggressiveness. Communication Research Reports, 16(2), 108-109.
Bensimon, H. F. (1997, Sept.). What to do about anger in the workplace. Training and Development, 51, 28-32.
Bingham, S. G, & Burleson, B. R. (1996). The development of the sexual harassment proclivity scale: Construct validation and relationship to communication competence. Communication Quarterly, 44, 308-325.
Bjorkqvist, K., Osterman, K., & Hjelt-Back, M. (1994). Aggression among university employees. Aggressive Behavior, 20, 173-184.
Braverman, M. (1999, June). Seven steps to preventing workplace violence. Access Control and Security Systems Integration, (42)7, 41-42.
Briggs, S. R., & Cheek, J. M. (1986). The role of factor analysis in the development and evaluation of personality scales. Journal of Personality, 54, 106-148.
Burgoon, J. K., & Hale, J. L. (1987). Validation and measurement of the fundamental themes of relational communication. Communication Monographs, 54, 19-41.
Buss, A. H. (1961). The psychology of aggression. New York: John Wiley.
Collinson, D. (1994). Strategies of resistance: Power, knowledge and subjectivity in the workplace. In J. M. Jermier, D. Knights, & W. R. Nord (Eds.), Resistance and power in organizations (pp. 25-68). London: Routledge.
Deetz, S. A. (1995). Transforming communication transforming business: Building responsive and responsible workplaces. Cresskill, NJ: Hampton.
DeRidder, R., Schruijer, S. G. L., & Rijsman, J. B. (1999). Retaliation to personalistic attack. Aggressive Behavior, 25, 91-96.
DeVellis, R. F. (1991). Scale development: Theory and applications. Newbury Park, CA: Sage Publications.
Dougherty, D. S. (2001). Sexual harassment as [dys]functional process: A feminist standpoint analysis. Journal of Applied Communication Research, 29, 372-402.
Exclusive HRfocus survey: Safe, not sorry, is the best approach. (2000, April). HR Focus, 77(4), S1, S4.
Fisher, S. K. (2001, May). Global workplace violence. Monthly Labor Review, 124(5), 53.
Koran, V. (2001, September). Employees concerned about workplace violence. Occupational Hazards, 63(9), 27.
Geddes, D., & Baron, R. A. (1997). Workplace aggression as a consequence of negative performance feedback. Management Communication Quarterly, 10, 433-454.
Geddes, D., & Linneham, F. (1996). Exploring the dimensionality of positive and negative feedback. Communication Quarterly, 44, 326-344.
How can HR help address the threat of workplace violence? (2003, October). HR Focus, 80(10), 8.
Incivility: Gateway to violence? (2000, February). Security, 37(2), 8.
Infante, D. A., Chandler, T. A., & Rudd, J. E. (1989). Test of an argumentative skill deficiency model of interspouse violence. Communication Monographs, 56, 163-177.
Infante, D. A., & Gorden, W. I. (1989). Argumentativeness and affirming communicator style as predictors of satisfaction/dissatisfaction with subordinates. Communication Quarterly, 37, 81-90.
Infante, D. A., Myers, S. A., & Buerkel, R. A. (1994). Argument and verbal aggression in constructive and destructive family and organizational disagreements. Western Journal of Communication, 58, 73-84.
Infante, D. A. , & Wigley, C. J., III. (1986). Verbal aggressiveness: An interpersonal model and measure. Communication Monographs, 53, 61-69.
Jackson, D. (1970). A sequential system for personality scale development. In C. D. Spielberger (Ed.), Current topics in clinical and community psychology, (pp. 61-96). New York: Academic Press.
Job violence: A global survey. (1998, Aug. 17). Business Week, 3591, p. 35.
Johnson, P. R., & Indvik, J. (2001). Rudeness at work: Impulse over restraint. Public Personnel Management, 30(4), 457-465.
Kassing, J. W. (1998). Development and validation of the organizational dissent scale. Management Communication Quarterly, 12(2), 183-229.
Katz, D. M. (2000, January 17). Study finds lag in violence prevention. National Underwriter: Property & Casualty/Risk & Benefits Management Edition, 704(3), 9.
Kinney, T. A. (1994). An inductively derived typology of verbal aggression and its association to distress. Human Communication Research, 27, 183-222.
Kramer, M. W., & Hess, J. A. (2002). Communication rules for the display of emotions in organizational settings. Management Communication Quarterly, 16, 66-80.
Kreps, G. E. (1993). Introduction: Sexual harassment and communication. In G. L. Kreps (Ed.). Sexual harassment: Communication implications (pp. 1-5). Cresskill, NJ: Hampton.
Lambert, B. (1995, September). Give your company a checkup. Personnel Journal, 74, 143-144.
Levin, R. L. (1995). Workplace violence: Sources of liability, warning signs, and ways to mitigate damages. Labor Law Journal, 46, 418-428.
Lutgen-Sandvik, P. (2003). The communicative cycle of employee emotional abuse: Generation and regeneration of workplace mistreatment. Management Communication Quarterly, 16, 471-501.
Maccoby, E., & Jacklin, C. N. (1980). Sex differences in aggression: A rejoinder and reprise. Child Development, 51, 946-980.
Magyar, S. V. Jr. (2003, June). Preventing workplace violence. Occupational Health & Safety, 72(6), 64-66.
Mantel, M. (1994). Ticking bombs: Defusing violence in the workplace. New York: Irwin Professional Publishing.
Martin, M. M., & Anderson, C. M. (1996). Argumentativeness and verbal aggressiveness. Journal of Social Behavior and Personality, 71, 547-554.
Most effective tool against workplace violence. (2003, February). HR Focus 80(1), 11-14.
Neuman, J. H. (2000, January). Injustice, stress, and bullying can be expensive! Paper presented at the meeting of the Workplace Bullying Conference, Oakland, CA.
Neuman, J. H., & Baron, R. A. (1997). Aggression in the workplace. In R. A. Giacalone and J. Greenberg (Eds.), Antisocial behavior in organizations (pp. 37-67). Thousand Oaks, CA: Sage Publications.
Nunnally, J. (1978). Psychometric theory. New York: McGraw-Hill.
O’Connell-Davidson, J. (1994). The sources and limits of resistance in a privatized utility. In J. M. Jermier, D. Knights, & W. R. Nord (Eds.), Resistance and power in organizations (pp. 69-101). London: Routledge.
O’Leary-Kelly, A. M., Griffin, R. W., & Glew, D. J. (1996). Organization-motivated aggression: A research framework. Academy of Management Review, 27, 225-253.
Pearson, C. M., Andersson, L. M., & Porath, C. L. (2000). Assessing and attacking workplace civility. Organizational Dynamics, 25(2), 123-137.
Perry, P. M. (2000, May). Zero tolerance. Industrial Distribution, 05(5), 56-60.
Prince, M. (2003, May 12). Violence in the workplace on the rise: Training, zero-tolerance can prevent aggression. Business Insurance, 37(9), 1-2.
Reynolds, W. M. (1982). Development of reliable and valid short forms of the Marlowe-Crowne social desirability scale. Journal of Clinical Psychology, 38, 119-125.
Robinson, S. L., & O’Leary-Kelly, A. (1996). Monkey see, Monkey do: The role of role models in predicting workplace aggression. In J. B. Keys and L. N. Dosier (Eds.), Academy of Management Best Papers Proceedings (pp. 284-287).
Skarlicki, D. P., & Folger, R. (1997). Retaliation in the workplace: The roles of distributive, procedural, and interactional justice. Journal of Applied Psychology, 82, 434-443.
Stevens, J. (1980). Intermediate statistics: A modern approach. Hillsdale, NJ: Erlbaum.
Stone, R. A. (1995, March-April). Workplace homicide: A time for action. Business Horizons, 38, 3-10.
Stuckless, N., & Goranson, R. (1992). The vengeance scale: Development of a measure of attitudes toward revenge. Journal of Social Behavior and Personality, 7, 25-42.
Top threat: Workplace violence remains no. 1 nemesis. (2001, June). Security, 38(6), 9-12.
U.N.: Job fears fueling violence. (1998, July 19). Associated Press. Retrieved July 20, 1998 from the World Wide Web: http://biz.yahoo.com
Valentine, S., Young, K., Bailey, L., Barhoun, N., LaBure, S., Glover, N, & Isaac, P. (2001). Expressing organizational dissent: The role of ethical ideology and corporate ethical values. Journal of Business Strategies, 18, 71-89.
Waldron, V. R. (2000). Relational experiences and emotion at work. In S. Fineman (Ed.), Emotion in organizations (2nd ed.) (pp. 64-82). Thousand Oaks, CA: Sage Publications.
Waldron, V. R., & Krone, K.J. (1991). The experience and expression of emotion in the workplace: A study of a corrections organization. Management Communication Quarterly, 4, 287-309.
Williams, K. (2003, July). Preventing workplace violence. Plants Sites and Parks, 30(3), 18-19.
Please direct correspondence to the first author (wtcoombs@hotmail.com)
An earlier version of this paper was presented to the Applied Communication Division of the National Communication Association, November 2002.
Copyright Central States Speech Association Fall 2004
Provided by ProQuest Information and Learning Company. All rights Reserved
