LEADER SOURCES OF POWER, REINFORCEMENT, PUNISHMENT, AND EMPLOYEE ATTITUDES AND BEHAVIORS ____________ A Thesis Presented to the Faculty of California State University, Chico ____________ In Partial Fulfillment of the Requirement for the Degree Master of Arts in Psychology Psychological Science Option ____________ by Scott T. Wallace Spring 2010 LEADER SOURCES OF POWER, REINFORCEMENT, PUNISHMENT, AND EMPLOYEE ATTITUDES AND BEHAVIORS A Thesis by Scott T. Wallace Spring 2010 APPROVED BY THE INTERIM DEAN OF THE SCHOOL OF GRADUATE, INTERNATIONAL, AND INTERDISCIPLINARY STUDIES: _________________________________ Mark J. Morlock, Ph.D. APPROVED BY THE GRADUATE ADVISORY COMMITTEE: _________________________________ Linda M. Kline, Ph.D., Chair _________________________________ Lawrence G. Herringer, Ph.D. TABLE OF CONTENTS PAGE List of Tables.............................................................................................................. iv List of Figures............................................................................................................. v Abstract....................................................................................................................... vi CHAPTER I. Introduction .............................................................................................. 1 II. Literature Review ..................................................................................... 5 Sources of Power.......................................................................... Supervisory Reinforcement and Punishment Behavior................ Workplace Deviance .................................................................... Hypotheses ................................................................................... 5 8 11 15 Methodology............................................................................................. 20 Participants ................................................................................... Measures....................................................................................... Procedure...................................................................................... 20 21 25 Results ...................................................................................................... 26 Testing Hypotheses ...................................................................... Discussion..................................................................................... Limitations.................................................................................... 31 40 47 Summary, Conclusions, and Recommendations ...................................... 50 References .................................................................................................................. 53 III. IV. V. iii LIST OF TABLES TABLE PAGE 1. Subordinates Descriptive Data.................................................................. 27 2. Supervisors Descriptive Data.................................................................... 28 3. Perceived Supervisor Use of Sources of Power........................................ 30 4. Perceived Supervisor Use of Leader Reward and Punishment Behavior ............................................................................................ 30 5. Self-Reported Employee Attitudes ........................................................... 31 6. Self-Reported Employee Behaviors.......................................................... 32 7. Sources of Power and Contingent/Non-Contingent Punishment and Reinforcement ............................................................................ 34 Employee Attitudes/Behaviors and Contingent/Non-Contingent Punishment and Reinforcement......................................................... 36 9. Employee Attitudes/Behaviors and Sources of Power ............................. 38 10. Employee Attitudes/Behaviors and Demographic Data ........................... 40 8. iv LIST OF FIGURES FIGURE 1. PAGE Hypothesis ................................................................................................ v 16 ABSTRACT LEADER SOURCES OF POWER, REINFORCEMENT, PUNISHMENT, AND EMPLOYEE ATTITUDES AND BEHAVIORS by Scott T. Wallace Master of Arts in Psychology Psychological Science Option California State University, Chico Spring 2010 Supervisors have five sources of power to lead subordinates, including legitimate, reward, referent, expert and coercive power. The use of each power source depends on situational variables and individual characteristics of the supervisor and subordinate. Supervisors can be more or less effective, depending on whether reinforcements and punishments utilized are contingent or not on the specified behavior. Contingent consequences are consistently more effective. The present study explored relationships between perceived supervisor bases of power, perceived supervisor reinforcement and punishment, and workplace deviance of subordinates. Ninety-two Sisco Enterprises subordinates and 30 Sisco Enterprises supervisors participated by completed inventories vi assessing demographic information, Sources of Power measure, Leader Reward and Punishment Questionnaire, and the Employee Perceptions Survey. Results supported the prediction that contingent reinforcement would be significantly correlated with all five sources of power (positively related to legitimate, reward, referent, and expert power, and negatively related to coercive power). Contingent punishment was also positively related with legitimate and expert power. Relationships with constructs related to job satisfaction and other employee attitudes and behaviors were also observed. Contingent reinforcement was positively related to company contentment and negatively related to dissatisfaction, whereas non-contingent punishment was positively related to dissatisfaction and negatively related to company contentment. Coercive power was positively related to theft approval, intentions to quit, and overall job dissatisfaction. Results also indicated that supervisors perceived themselves in a more positive light as compared to subordinate perceptions. These findings extend previously published work on bases of power and use of contingent reinforcement and punishment. vii CHAPTER I INTRODUCTION As the culture of business continues to evolve into a global market, competition remains fierce with companies continuing to search for ways to improve efficiency, increase profits, and achieve sustainability. An important component of any business is the dynamic relationship that supervisors share with their subordinates. This relationship can play an integral role in how effectively business operates. Standard responsibilities of any supervisor, such as ensuring product quality, implementing company policy, and controlling costs and profits are all directed through subordinates. In order to manage responsibilities, supervisors shape subordinate behavior by relying on multiple sources or methods of influence. A supervisor’s effectiveness often depends on the way in which he communicates and lead others and which power sources are relied upon to change or sustain subordinate behaviors (Hinkin & Schriesheim, 1994). French and Raven (1959) proposed five potential sources of interpersonal power: legitimate, coercive, reward, referent, and expert power. Supervisors rely on their power to influence subordinate behavior. To a certain degree, everyone holds each source of power, however a subordinate’s perception of the supervisor’s power ultimately determines effectiveness (French & Raven, 1959). A person who has the right, or the authority, to tell others what to do has Legitimate Power, and employees are obligated to comply with orders. For example, 1 2 when a supervisor writes a staff schedule, they are influencing subordinates by utilizing legitimate power, which is granted from their position. Someone who has Reward Power influences others because they control valued rewards, which generates employee compliance. A supervisor who grants an employee permission to leave early, or provides them with a raise is using reward power. Coercive Power is in the hands of a person who controls punishments and employees comply to avoid being punished. When supervisors administer corrective action they are using coercive power to change subordinates’ behavior. With Referent Power, a supervisor has personal characteristics that appeal to others. People comply because of admiration, a desire for approval, personal liking, or a desire to be like the leader. When supervisors create an enjoyable work environment for subordinates, while still leading by example, they are using referent power. Lastly, Expert Power exists when someone has certain expertise or knowledge, and employees comply because they believe in, can learn from, or can otherwise gain from that expertise (French & Raven, 1959; Hinkin & Schreisheim, 1994). In all supervisor-subordinate relationships, the five sources of power operate to a certain degree. How much, depends on the individual characteristics of those on both sides of the supervisor-subordinate relationship. More importantly, subordinates are generally more responsive to some sources of power compared to others. For example, the continued use of reward, referent, or expert power, tends to increase the attraction of the subordinate to the supervisor, increasing the extent of influence and overall effectiveness (French & Raven, 1959). Another approach to understanding leadership behavior is to observe the way supervisors punish and reward subordinates. One important variable includes whether or 3 not these leadership behaviors are being administered contingent on the targeted behavior. When supervisors reward subordinates, based on their performance, they are providing Contingent Reinforcement. Similarly, when supervisors punish subordinates based on their performance they are administering Contingent Punishment. Supervisors who reward subordinates on criteria that are not based on their performance are exercising Non-contingent Reinforcement, and supervisors who punish subordinates on criteria that are not based on their performance are using Non-contingent Punishment. Both punishment and reinforcement are most effective when they are applied contingent to the specific behavior intended to influence. Contingent reinforcement is supported as the most effective method in changing or sustaining behavior, as it is strongly associated with desirable affective and behavioral states such as job satisfaction and performance (Williams & Podsakoff, 1988). In comparison, contingent punishment has a small positive relationship with desirable affective and behavioral states. Noncontingent reinforcement has little effect on subordinate outcomes and no relationship with desirable affective and behavioral states, and non-contingent punishment has a very negative relationship with subordinate affect (Williams & Podsakoff, 1988). As supervisors, there are multiple ways to influence subordinates into complying with orders. Since some methods are more effective than others, it is critical for researchers to identify the consequences and benefits of someone utilizing a particular source of power and with the different types of reinforcement and punishment. The purpose of this study was to explore relationships between leader sources of power and contingent/non-contingent reinforcement and punishment. Hinkin and Schreisheim (1994) previously established these relationships. I intended to replicate 4 their findings and extend the research to subordinate outcomes, as measured by their attitudes and behaviors related to the workplace. It is important to understand which sources of power most effective and to gauge the impact of contingency on behavior change. Another valuable component of this research study included a real-world business environment, which further added to the validity of the previous supporting findings. CHAPTER II LITERATURE REVIEW Sources of Power Over the last half century, the ideas French and Raven proposed in their sources of social power research have lead to a better understanding of these concepts and their contributions toward leadership. Legitimate, Coercive, and Reward power are distinguished as sources of position power since they are all granted by a person’s role or position. With these powers, the sources are attained externally, such as with rank and position, or the ability to administer rewards or punishments. Referent and Expert power are both uniquely different to the individual because they are distinguished as sources of personal power. With these powers, the sources are attained internally, such as with the ability to persuade people, or the possession of superior knowledge, judgment, or respect (Hinkin & Schriesheim, 1994). Determining which type of power to use is often times dependent on situational variables, such as current business needs, or individual characteristics of the supervisor or subordinate. Multiple power sources can also operate at one time. For example, a supervisor who lets an employee go home early because they worked hard is using both legitimate and reward power. French and Raven (1959) suggest that effective leadership is comprised of the balance use of personal power (Referent and Expert) and position power (Legitimate, Coercive, and Reward). 5 6 There are consistencies that occur across the five sources of power. For example, when the use of legitimate power is outside the range of what is warranted, the effectiveness of use decreases, and so does the responsiveness of the follower to the leader. With reward, referent, and expert power, use over time tends to increase the attraction of the follower to the leader by leading to an independent system. An independent system is advantageous because the associated region of behavior becomes positively valanced, often spreading to other regions of the life space, and in effect become stronger sources of reinforcement (French & Raven, 1959). Comparatively less effective over time, coercive power tends to decrease the attraction from follower to leader, continuing to be a dependent system. As a result, the associated region of behavior becomes negatively valanced; often spreading to other regions of the life space, and in effect lessening the strength of influence between the individual and the reinforcement mediator (French & Raven, 1959). Richardson, Swan, and Hutton (1995) examined how power distributors respond to using certain powers differently than using others. In their findings, the presence and use of reward power sources was positively related to distributor satisfaction. Consistent with previous research, the relationship also consistent from the other end of the spectrum with power distributors relying on coercive power sources also reporting less satisfaction. The sources of power have also been more recently differentiated according to soft sources of power, such as reward, referent and expert power, and harsh sources of power, such as legitimate and coercive power. Koslowsky, Schwarzwald, and Ashuri (2001), gathered data from 232 nurses and 32 supervisors from 2 municipal hospitals and 7 asked questions about power, job satisfaction, and organizational commitment. One objective was to explore compliance differences between harsh and soft bases of power. Compliance was argued to be a reflection of the quality of leadership. Previous research supports this rationale suggesting that compliance, which is attributed to fear of punishment or an expectation to fulfill contractual obligations, emphasizes subordination rather than voluntary acceptance (Avolio & Bass, 1988). The results of Koslowsky et al. suggested that job satisfaction is positively related to compliance with soft sources, such as reward, referent, and expert power, while also negatively related to compliance with harsh sources, such as legitimate and coercive power. The leadership objectives attained through soft sources were more effective because they were more task relevant and compliance was gained through personal rather than organizational resources. These soft sources de-emphasize the positional advantage of the supervisor and instead focus on the personal resources of the leader (2001). Carson, Carson, and Roe (1993) completed a meta-analysis examining the interrelationships and outcomes of the five social power bases developed and examined by French and Raven (1959). Each power base was examined considering their relatedness to satisfaction with one’s supervisor, job satisfaction, and job performance results. The results of their research indicated that Legitimate power exerts little influence on either satisfaction or performance. Reward power did not have a significant affect on satisfaction, however it did have a positive influence on performance. Coercive power appeared unrelated to performance but negatively affected satisfaction. Referent power strongly influenced satisfaction and expert power positively influenced all three outcomes. These results further support reliance on personal power bases (referent and 8 expert) to attain the most desirable outcomes. Other research suggests that leadership is most effective when the behavior modifier is applied contingent to the behavior that is intended to be sustained or changed. Supervisory Reinforcement and Punishment Behavior Contingent reinforcement is strongly and positively associated with desirable affective and behavioral states, while contingent punishment has a small positive relationship or none at all. Non-contingent reinforcement also has a small positive or no relationship with these dependent variables, and non-contingent punishment tends to be very negatively related (Hinkin & Schriesheim, 1994; Podsakoff & Schriesheim, 1986; Williams & Podsakoff, 1988). Hinkin and Schriesheim (1994) explored these concepts further and identified relationships between perceived leader reinforcement behavior and the perceived leader power sources by identifying correlates of the two constructs. Two-samples, one intraand one inter-organizational were used in their study, with participants responding to survey questionnaires. One sample included 375 full-time employees of a state-operated hospital located in the southern United States. Another sample included 297 mid-level managers employed at 100 financial institutions throughout the United States. The researchers used the Leader Reward and Punishment Questionnaire (LRPQ) to identify preferred leadership behaviors, and another measure developed by Hinkin and Schriesheim’s (1989) was used assessing French and Raven’s (1959) bases of social power. The researchers also collected demographic information and also tracked satisfaction levels through The Minnesota Satisfaction Questionnaire (MSQ). 9 Their findings showed perceived contingent reinforcement enhancing all five bases of power, especially referent and expert power. As expected, contingent reinforcement was also associated with positive subordinate affect. Perceived contingent reinforcement was also negatively related to the perception of coercive power, a power source that is generally viewed unfavorably by subordinates (Podsakoff & Schriesheim, 1986). Examples of contingent reinforcement within the LRPQ include supervisors responding to subordinates doing well, by giving praise, administering a reward, or recognizing improvements in their performance. These leadership reinforcement behaviors were found to be more associated with the soft bases of power. Perceived non-contingent punishment behavior showed a negative relationship with referent power, therefore suggesting that supervisors should avoid the use of this leadership behavior. Examples of non-contingent punishment within the LRPQ include supervisors responding incorrectly by being undeservingly critical, reprimanding, and displeased with subordinates. Perceived non-contingent reward behavior showed no relationships with perceived power sources, therefore suggesting the use of non-contingent reinforcement as ineffective. Previous research has supported that reinforcement and punishment appear to be related to the sources of social power. Hinkin and Schriesheim (1994) found a relationship between contingent reinforcement and the enhancement of all five sources of social power, and especially within referent and expert power. This is a valid association since contingent reinforcement is strongly associated with desirable affective and behavioral states, and so is referent and expert power. Also, with the enhancement of 10 power, a positive valance, or emotional value associated with a stimulus, develops which increases the supervisor’s extent of influence over the subordinate (French & Raven, 1959). On the other end of the spectrum, reliance on non-contingent punishment can be very negatively related to desirable behavioral and affective states and can decrease one’s power bases due to negative valances (Hinkin & Schriesheim, 1994). Similarly, the reliance on coercive power tends to decrease the attraction from the subordinate to the supervisor, often spreading negative valences onto other regions of the life space (French & Raven 1959). The findings of Hinkin and Schriesheim (1994) have shown relationships between perceived sources of power and perceived leader reward and punishment behavior. While a supervisor’s effectiveness can be reliably assessed according to these two constructs, equally as important should be the results of their leadership. I am interested in exploring subordinate attitudes that develop and behaviors that occur in response to the use of sources of power and supervisor reinforcement and punishment. My objective is to replicate the findings of Hinkin and Schriesheim (1994) and to extend the research scope by including comparison of relationships across subordinate attitudes and behaviors. For example, if supervisors can lose their power due to negative valances associated with the reliance of coercive power or by applying punishments noncontingently, then what types of behaviors and attitudes should we see from subordinates? Perhaps, more unfavorable actions such as absenteeism, privilege abuse, and theft are associated with coercive power and use of punishment. This information is 11 valuable for businesses because they struggle when employees are noncompliant with policies or steal; common forms of stealing include discount violations, theft of product, or grazing, which includes sampling or tasting of product (Hollinger & Clark, 1983). Workplace Deviance In 2004, when the National Food Service Security Council investigated 1,083 restaurants representing 10 companies, nearly half (44%) of the respondents reported some form of theft from their employer (Langton & Hollinger, 2005). Other estimates have suggested that between 1/4 and 1/2 of all employees steal to some degree from their employers, with annual costs ranging from $6 billion to $200 billion (Diefendorff & Mehta, 2007; Hollinger & Clark, 1983). Losses of this magnitude directly affect a company’s bottom line and threaten their sustainability. Workplace deviance is defined as “voluntary behavior that violates significant organizational norms and in so doing threatens the well-being of an organization, its members, or both” (Diefendorff & Mehta, 2007, p.556). Workplace deviance can be very costly as it includes theft, acts by employees against the property of the organization and counter-productivity, the violation of the norms regulating acceptable levels of production (Hollinger & Clark, 1983). Employee theft can be debilitating for businesses, however counterproductivity is often a more pervasive form of rule breaking by employees, and one that is more difficult to measure (Hollinger & Clark, 1983). Counterproductive behaviors exhibited by employee’s can be a major contributing factor to business failures (Hollinger & Clark, 1983; Kamp & Brooks, 1991). 12 Antecedents to employee deviance include negative interpersonal treatment, and workplace experiences such as frustration, injustices, and threats to self (Bennett & Robinson, as cited in Diefendorff & Mehta, 2007; Robinson & Greenberg, 1998). Mitchell and Ambrose (2007) conducted survey research and showed how abusive supervision is positively related to all types of workplace deviance. Abusive supervision has been defined as the subordinates’ perceptions of the extent to which their supervisors engage in the sustained display of verbal and nonverbal behaviors, excluding physical contact (Tepper, 2000). Abusive managers have also been described as supervisors who callously and arbitrarily use their power and authority to mistreat employees (Ashforth, 1997). Perceived non-contingent punishment could be considered abusive supervision. Data was collected from 427 individuals who were called for jury duty by a county circuit court in Southeastern United States. The researchers did not address anything related to the jury or court system, but instead asked the potential jurors information about sensitive issues related to their regular job. The survey results suggested that abusive supervision negatively affects employee attitudes. In reaction to supervisor abuse, employees became less willing to engage in positive behaviors, such as collaboration on tasks or recognition of issues, and more willing to engage in negative behaviors, such as organizational deviance, supervisor-directed deviance, or reciprocating negatively in response to others behaviors. In other studies, abusive supervision has also been shown to be related to lower levels of satisfaction, lower levels of commitment, altered perceptions of justice that is fair, higher levels of turnover, and psychological distress (Mitchell & Ambrose, 2007). 13 Many researchers have looked at other variables related to deviant employee attitudes and behaviors. Deviant employee attitudes, such as the intention to quit and job dissatisfaction, have been widely used to predict employee compliance and productivity behaviors (Bolin & Heatherly, 2001). Mangoine and Quinn (1975) identified psychological predictors of employee counterproductivity. Their findings suggested that tolerant attitudes toward violence and dissatisfaction significantly predicted on-the-job damage by employees. Also, condoning attitudes towards violence or dishonesty and dissatisfaction with co-workers reliably predicted on the job waste by employees. Relationships between demographic variables, such as age and employee deviance have also been observed. In an extensive research study, Hollinger and Clark (1983) explored correlates of employee deviance across three industrial sectors; retail, hospital, and manufacturing, and three metropolitan areas; Minneapolis-St. Paul, DallasFort Worth, and Cleveland. Self-report questionnaires were completed by 9,175 participants from their homes answering questions about their involvement in property theft activities within the formal employment setting. Hollinger and Clark reported negative relationships between employee deviance and job satisfaction, age, education, continuous employment, and tenure. A few of their findings in particular were strongly supported. In all three job sectors, employees who were actively looking for a new job were significantly more likely to participate in behaviors that were considered theft. A similar relationship was found for tenure. Again, in all three sectors, employees were less likely to engage in theft behaviors the longer they were at their occupation. As a whole, their findings suggest that employee deviance is more likely to be found with employees who were less satisfied, younger in age, less 14 educated, less tenured, and also with those who had not maintained continuous employment with their employer. Bennett and Robinson (2003) identified three distinct research trends within workplace deviance. One includes studies where deviance was conceptualized as a reaction to experiences at work. Another focuses on examining deviance as a reflection of employee’s personality, and a third investigates deviance as an adaptation to the social context of work. The three research trends were conceptualized based on reviewing a collection of reviewed research studies. In the first research trend, job dissatisfaction was studied in terms of reactions to experiences at work. Research has established job dissatisfaction as being related to measures of deviance (Bennett & Robinson, 2003) and also withdrawal behaviors (Hulin, 1991). Those who reported being more dissatisfied than others also reported higher incident of both deviant and withdrawing behaviors. The subordinate’s reaction was within the elements of the task at hand, or something contingent specifically of their job performance. The effects of their own job dissatisfaction generally only interfere with their own job performance and not others. With the second research trend, evidence suggested that anger and hostility, constructs within the individuals personality, were positively related to deviance and counterproductive behaviors (Fox & Spector, 1999; Lee and Allen, 2002). This research focus concerned detrimental or potentially destructive acts that hurt colleagues or organizations. This knowledge is invaluable because many forms of counterproductive deviance such as insubordination or corporate sabotage can be very damaging to the infrastructure of a business system. 15 In the third research trend, Robinson and Greenberg’s (1998) findings suggested that unfair interpersonal treatment is prominent in social influence on deviance. Those who perceived themselves as being mistreated or disrespected, and were within a work culture fostering deviance, were more inclined to be counter-productive, deviant, and have more negative attitudes. All three research trends are compelling and particularly relevant focus areas for my current topic at hand. The preceding research findings support the notion that leadership and supervisory behavior can have a tremendous impact on employee attitudes such as satisfaction, which can then affect employee behaviors. It becomes essential to evaluate the interactions, which occur between supervisor and subordinate, in order to gain a better understanding of workplace deviance. Hypotheses The general purpose of this study is to explore relationships between perceived supervisory reinforcement and punishment behavior, sources of power, and employee attitudes and behaviors. Another objective is also to further understand sources of social power further in terms of soft and harsh bases. This study is also designed to follow up the work of Hinkin and Schrieshiem (1994) by exploring relationships between perceived sources of power and perceived leader reward and punishment behavior. The present study sought to extend their findings to deviant employee attitudes and behaviors. Considering other relationships of employee deviance, mentioned earlier (Bolin & Heatherly, 2001; Hollinger & Clark, 1983; Mangoine & Quinn, 1975), additional demographic variables, such as age, sex, 16 education, job tenure, continuous employment, supervisory experience, work schedule, and familiarity of workshop classes will also be included in the research design. Refer to Figure 1 for a diagram of all the hypotheses. Figure 1. Hypothesis. Hypotheses #1: Supervisor Behavior and Power Sources In line with Hinkin and Schriesheim (1994), I predicted that relationships would exist between perceived supervisor reward and punishment behaviors and perceived power sources. I predicted positive relationships between contingent reinforcement and legitimate, reward, referent, and expert sources of power, and a 17 negative relationship between contingent reinforcement and coercive sources of power. Referent and expert power, the personal power bases, should show the strongest correlations with contingent reinforcement. Carson et al.’s (1993) meta-analysis provides supportive data linking referent power with positive satisfaction with one’s job and one’s supervisor. Their research also showed relationships with the other personal power base, expert power, with positive satisfaction with one’s job, one’s supervisor, and better individual performance at their job. I predicted that perceived supervisory personal power would be negatively correlated with non-contingent reinforcement. Regarding management contingent punishment, I predicted positive relationships between contingent punishment and legitimate sources of power as well as expert sources of power. Non-contingent punishment was predicted to be negatively correlated with legitimate, reward, referent, and expert sources of powers. Noncontingent punishment was predicted to be positively correlated with coercive sources of power. Hypothesis #2: Leader Behavior and Employee Attitudes and Behaviors I predicted significant relationships between perceived supervisory reward and punishment behavior and both employee attitudes and employee behaviors. For employee attitudes, I predicted significant negative relationships between contingent reinforcement and approval of theft, intent to quit, and job dissatisfaction, as well as a positive relationship with being content with the company for which they work. With regard to employee behaviors, I predicted negative relationships between contingent reinforcement and absenteeism, substance abuse, and theft. 18 With regard to non-contingent punishment and employee attitudes, I predicted positive relationships with approval of theft, intent to quit, and job dissatisfaction. I also predicted a negative relationship between non-contingent punishment and company content, a construct which considers an employee’s overall satisfaction with a company and their policies. Considering employee behaviors, I also predicted positive relationships between non-contingent punishment and absenteeism, substance abuse, privilege abuse, and theft. No predictions were made for both contingent punishment and non-contingent reinforcement. Hypothesis #3: Sources of Power and Employee Attitudes and Behaviors Relationships should exist between power sources and employee attitudes. With regard to reward, referent, and expert sources of power, I predicted negative relationships between these power sources and theft approval, intent to quit, and dissatisfaction. I predicted a positive relationship between the power sources and company content. I predicted positive relationships between coercive power source and theft approval, intent to quit, and dissatisfaction, as well as a negative relationship between coercive power source and company content. No predictions were made for legitimate power. Relationships should also exist between power sources and employee behaviors. For reward, referent, and expert sources of power, I predicted negative relationships between these power sources and absenteeism, substance abuse, privilege abuse, and theft. With regard to coercive power, I predicted positive relationships 19 between absenteeism, substance abuse, privilege abuse, and theft. No predictions were made for legitimate power. Hypothesis #4: Demographic Data and Employee Attitudes and Behaviors Relationships should exist between demographic variables and employee attitudes. With regard to age, education, continuous employment, and tenure, I predicted negative relationships between these and theft approval, intent to quit and dissatisfaction. I predicted positive relationships between company content, age, education, tenure, and a negative relationship with non-continuous employment. Relationships should also exist between demographic variables and employee behaviors. With regard to age, education, and tenure, I predicted negative relationships between these demographic variables and absenteeism, substance abuse, privilege abuse, and theft. Non-continuous employment will also have a positive relationship with absenteeism, substance abuse, privilege abuse, and theft. Hypothesis #5: Store Location and Employee Attitudes and Behaviors Relationships should exist between store location and employee attitudes and behaviors. I predicted that pre-identified, high-deviance stores would report high levels of theft approval, intent to quit, dissatisfaction, absenteeism, substance abuse, privilege abuse, and theft, when compared to the pre-identified, low-deviance stores. The preidentified high-deviance stores will also report lower levels of company content compared to the pre-identified low-deviance stores. CHAPTER III METHODOLOGY Participants The participants consisted of 122 Sisco Enterprises employees from six Round Table Pizza locations. Of the 122 participants, 92 were subordinates and 30 were supervisors. Slightly more than half of the total sample (56.2%) was male and the mean age was 23.45 years (SD = 8.45). The youngest participant was 16 years of age and the oldest participant was 55 years of age with a variance of 71.48. The level of education was fairly representative with 24.6% at a high school level, 22.1% with a high school diploma or equivalent, 44.3% with some college education, 6.6% with a college degree, and 1.6% were graduate students. The average tenure was 32 months (SD = 49.37). The least tenured participant had been employed for 1 month, and the most tenured participant had been employed for 26 years, with a variance of 2,437.14 months. The majority (73.8%) of employees at Sisco Enterprises reported continuous employment. Of Sisco Enterprises’ 23 store locations, three were pre-identified as highdeviance locations, based on recent prevalence of incidents. These three stores were included in the target population at the request of Sisco Enterprises. Four locations were eliminated from the selection procedure to avoid potential biases due to the researchers 20 21 familiarity of employees. Out of the remaining 16 locations, three target locations were randomly selected and classified as low-deviance locations. Measures To assess the respondents’ perceptions of leader power, Hinkin and Schriesheim’s (1989) measure of the five bases of social power was used. This measure has four items for each of the five power bases (i.e., Legitimate, Coercive, Reward, Referent, and Expert). Also, it uses a 5-point Likert response format (1 = Strongly agree, 5 = Strongly disagree). This measure has reported coefficient alpha reliabilities ranging from.77-.90 based on three samples. One consisted of 251 upper-level undergraduates enrolled in two different business courses at a large southern U.S. university. Another sample consisted of 375 full-time employees of a large southern U.S. psychiatric hospital. A third consisted of 220 part-time MBA students taking organizational behavior and business policy classes at a medium-sized southern U.S. university (Hinkin & Schriesheim, 1989; Hinkin & Schreisheim, 1994). The supervisors evaluated their own behavior while the subordinates evaluated their supervisor’s behavior. This measure is most typically used as a self-report instrument of one’s behavior. Content validity of the social power measure developed by Hinkin and Scheisheim (1989) was established. Originally, 53 items were selected and retained for further assessment. Two independent panels of judges (N = 37 and N = 42) were presented with the construct definitions and were instructed to classify the randomly ordered items into one or more power categories. The 42 items, which were assigned to 22 the proper a priori category more than 60% of the time by both panels, were retained, thus supporting appropriate content validity for the measure. Scores were calculated by averaging the responses to the 20 questions in groups of five composite scales according to the sources of power: legitimate, reward, coercive, referent, and expert. For example, one of the items measuring legitimate power asked, “To what degree can your supervisors make you feel like you have tasks to accomplish?” Measuring coercive power, another question asked, “To what degree can your supervisor make work difficult?” Perceived leader reward and punishment behavior was assessed using the 23item Leader Reward and Punishment Questionnaire (LRPQ), which was developed by Podsakoff and Skov (1980). The instrument consists of four psychometrically sound subscales (i.e., Contingent Reinforcement, Contingent Punishment, Non-contingent Reinforcement, and Non-contingent punishment), with a 5-point Likert response format (1 = Strongly agree, 5 = Strongly disagree). This measure has demonstrated the appropriate internal consistency levels and reliabilities, all above .70 based on two samples. One consisted of 375 full-time employees of a large southern hospital, and another included 297 mid-level managers employed at over 100 US financial institutions ( Hinkin & Schriesheim, 1994; Podsakoff & Schriesheim, 1984, 1986; Schriesheim, Hinkin, & Tetrault, 1991). The supervisors evaluated their own behavior while the subordinates evaluated the perceptions they have of their supervisor’s behavior. This measure is most typically used as a self-report of one’s behavior. Scores for the Leader Reward and Punishment Questionnaire measure were calculated by averaging the responses to the 23 questions in groups of four composite 23 scales according to reinforcement, punishment, contingency and non-contingency. For example, one of the items measuring contingent reinforcement asked, “To what degree can your supervisors give you positive feedback when you do well?” To measure contingent punishment, another item asked, “To what degree can your supervisors let you know if you perform poorly?” The 39-item Employee Perceptions Survey (EPS) (Bolin & Heatherly, 2001) was used to assess employee attitudes and employee behaviors. This measure was divided into two parts (Attitudes and Behaviors) and consists of four subscales for each section. The employee attitudes measure assessed using four subscales; Theft Approval, Company Content, Intent to Quit, and Dissatisfaction. Employee behaviors included Absenteeism, Substance Abuse, Privilege Abuse, and Theft. For the Attitudes section, persons could respond to each statement on a 5-point scale, ranging from 1 (Strongly agree) to 5(Strongly disagree), and for the behaviors section, responses could range from 1 (Never) to 5 (Always). Five point Likert scales were used for both parts of the EPS (Attitudes and Behaviors). The EPS has reported coefficient alpha reliabilities ranging from .62 - .88 based on a sample of 1,759 entry-level restaurant employees and 2,763 entry-level supermarket and grocery store employees (Bolin & Heatherly, 2001). Robinson and Bennett (1995) provided supportive evidence for the construct validity of the EPS. They developed a list of 45 deviant workplace behaviors. Participants then rated the similarity of these behaviors to each of the other 44 behaviors. The researchers concluded that deviant workplace behaviors vary along two dimensions: 1) a 24 minor/serious dimension, and 2) a dimension that identifies behaviors that harm the organization and those that harm the individual (Bolin & Heatherly, 2001). Scores for the EPS measure were calculated by averaging the responses to the 20 questions from the attitudes section into groups of four composite scales, according to theft approval, company content, intent to quit, and dissatisfaction. For example, to measure theft approval an item suggests, “I could steal from my employer anytime,” and asks for a response from 1 to 5 to indicate the strength of agreement. Other items are weighted for more than one category. Question #9 states, “I will leave my current job in the next three months.” This item measures both intent to quit and dissatisfaction. The EPS behaviors section was also calculated by averaging the responses to the 19 questions into groups of four composite scales according to absenteeism, substance abuse, privilege abuse, and theft. For example, to measure absenteeism, one item suggests, “I have been absent without a legitimate excuse,” and asks for a response from 1 to 5 to indicate the strength of agreement. Similar to the attitudes section, some items also measure more than one construct. Item #11 suggests, “I have eaten food at work without paying for it,” which is a measure of both privilege abuse and theft. To analyze the data, the responses were reduced into composite variables based on the intended subscales of each measure. The responses for perceived leader power were reduced into five subscales: Legitimate power, Reward power, Coercive power, Referent power, and Expert power. The Leader Reward and Punishment Questionnaire was reduced according to four subscales; Contingent reward behavior (CR), Contingent punishment behavior (CP), Non-contingent reward behavior (NCR), and Non-contingent punishment behavior (NCP). The Employee Perceptions Survey: 25 Attitudes section was reduced into the four intended subscales: Theft approval, Company Content, Intent to Quit, and Dissatisfaction. The Employee Perceptions Survey: Behaviors section was also reduced into four intended subscales: Absenteeism, Substance Abuse, Privilege Abuse, and Theft. Procedure The daily procedure for collecting data consisted of pulling voluntary participants from the daily workforce to complete the measures. All data collected occurred during open business hours and all participants individually completed the measures with the proctor. Data collection occurred over the time span of 30 days. After the informed consent process, the participants completed the measures in the following order: Demographic Data, Power Sources, Leader Reward and Punishment Questionnaire (LRPQ), and The Employee Perceptions Survey (EPS): Attitudes (A) & Behaviors (B). Supervisors were asked to report on perceptions of their own behaviors for the Sources of Power measure, the Leader Reward and Punishment Questionnaire and for the Employee Perceptions Survey. Subordinates were asked to report on perceptions of their supervisor’s behaviors for the Sources of Power measure and the Leader Reward and Punishment Questionnaire. However, the subordinates in slight difference did report on perceptions of their own attitudes and behaviors for the Employee Perceptions Survey. Following completion of the survey, the participants were debriefed and their names were entered for a $50 Best Buy gift certificate drawing. The entire running procedure was 30 minutes long. CHAPTER IV RESULTS Responses from supervisors and subordinates were compared on all measures. There were several t-tests with each one looking at one dependent variable. The independent variable was supervisor or subordinate, while the dependent variables included the constructs measured according to their appropriate subscales. For the sources of power measure, supervisory and subordinate differences in response to the subscale questions were compared for legitimate, reward, coercive, referent, and expert powers. For the Leader Reward and Punishment Questionnaire, supervisory and subordinate differences in response to the subscale questions were compared for contingent reinforcement, non-contingent reinforcement, contingent punishment, and non-contingent punishment. For the Employee Perceptions Survey, supervisory and subordinate differences in response to the subscale questions were compare for theft approval, intent to quit, dissatisfaction, company content, absenteeism, theft, privilege abuse, and substance abuse. Although the number of t-test comparisons is large (25), the total number of significant findings exceeds chance levels for multiple tests. For descriptive statistics behind each of the measures see Table 1 and Table 2. One objective of the study was to determine how accurately supervisors view their own leadership behaviors in terms of effectively motivating their subordinates. Since the effectiveness of a power source is determined by the receiving agent it is 26 27 Table 1 Subordinates Descriptive Data Subordinates (N = 92) Mean (SD) Minimum Maximum Legitimate Power 4.26 (SD = .61) 3 5 Reward Power 3.36 (SD = .77) 1.5 5 Coercive Power 3.07 (SD = 1.16) 1 5 Referent Power 4.30 (SD = .72) 2.5 5 Expert Power 4.21 (SD = .63) 2.5 5 Non-Contingent Reinforcement 2.45 (SD = .74) 1 4.5 Non-Contingent Punishment 2.17 (SD = .77) 1 4.25 Contingent Punishment 2.96 (SD = .48) 1.6 4 Contingent Reinforcement 3.44 (SD = .66) 1.6 4.5 Theft Approval 2.15 (SD = .84) 1 4.33 Company Content 3.37 (SD = .32) 2.44 4.22 Intent to Quit 2.92 (SD = .76) 1 4.67 Absenteeism 1.48 (SD = .45) 1 3.2 Substance Abuse 1.25 (SD = .47) 1 3.75 Privilege Abuse 1.59 (SD = .62) 1 3.83 Theft 1.32 (SD = .42) 1 3.6 Dissatisfaction 2.97 (SD = .62) 1.33 4.42 28 Table 2 Supervisors Descriptive Data Supervisors (N = 29) Mean (SD) Minimum Maximum Legitimate Power 4.32 (SD = .57) 3 5 Reward Power 3.27 (SD = .82) 1 5 Coercive Power 3.23 (SD = 1.39) 1 5 Referent Power 4.63 (SD = .44) 3.5 5 Expert Power 4.58 (SD = .44) 3.75 5 Non-Contingent Reinforcement 2.47 (SD = .73) 1.25 4 Non-Contingent Punishment 1.68 (SD = .41) 1 2.5 Contingent Punishment 3.04 (SD = .54) 2 4 Contingent Reinforcement 3.97 (SD = .32) 3.1 4.5 Theft Approval 2.68 (SD = 1.03) 1 4.67 Company Content 3.41 (SD = .35) 2.67 4.11 Intent to Quit 2.78 (SD = .83) 1 4.17 Absenteeism 1.45 (SD = .40) 1 2.4 Substance Abuse 1.32 (SD = .52) 1 3 Privilege Abuse 1.84 (SD = .82) 1 3.67 Theft 1.49 (SD = .51) 1 3.1 Dissatisfaction 2.81 (SD = .68) 1.33 4.08 29 critical for supervisors to understand if their directives are being accurately understood by the subordinates. Because of this, I decided to group the supervisor responses together and the subordinate responses together, to compare perceptual differences. The two groups showed notable differences in the Sources of Power measure and the Leader Reward and Punishment Questionnaire. Independent samples t-tests revealed that 30 supervisors self-reported significantly more use of referent power than perceived by subordinates, t(82) = 2.98, p < .005. The supervisors also self-reported significantly more expert power, than perceived by subordinates, t(71) = 3.57, p < .001. There were no significant differences with regard to legitimate power, reward power nor coercive power. Some of the participants chose not to answer some of the survey questions; therefore, some of the data reports lower participant levels than total participants. See Table 3 for the means and standard deviations. A second set of independent samples t-tests explored possible differences between supervisors and subordinates regarding contingent reinforcement, contingent punishment, non-contingent reinforcement, and non-contingent punishment. An independent samples t-test revealed that supervisors self-reported significantly less non-contingent punishment than perceived by subordinates, t(89) = 4.46, p < .001. The supervisors also self-reported significantly more contingent reinforcement than subordinates perceived, t(120) = 2.81, p < .01. There were no significant differences with regard to contingent punishment and non-contingent reinforcement. Please refer to Table 4 for the means and standard deviations. 30 Table 3 Perceived Supervisor Use of Sources of Power Power Source Supervisors N = 30 Subordinates N = 92 Legitimate Power 4.32 (SD = .57) 4.26 (SD = .61) Reward Power 3.27 (SD = .82) 3.36 (SD = .77) Coercive Power 3.23 (SD = 1.39) 3.07 (SD = 1.16) Referent Power* 4.63 (SD = .44) 4.31 (SD = .72) Expert Power** 4.58 (SD = .44) 4.21 (SD = .63) *p < .01 **p < .001 Table 4 Perceived Supervisor Use of Leader Reward and Punishment Behavior Leader Behavior Supervisors N = 29 Subordinates N = 91 Contingent Reinforcement* 3.97 (SD = .32) 3.44 (SD = .66) Contingent Punishment 3.05 (SD = .54) 2.95 (SD = .48) Non-contingent Reinforcement 2.47 (SD = .73) 2.45 (SD = .74) Non-contingent Punishment** 1.68 (SD = .41) 2.17 (SD = .77) *p < .01 **p < .001 31 A series of independent samples t-tests explored possible differences in deviant attitudes and behavior between supervisors and subordinates. There were no significant differences. Please refer to Tables 5 and 6 for means and standard deviations. Table 5 Self-Reported Employee Attitudes Employee Attitudes Supervisors N = 30 Subordinates N = 92 Theft Approval 2.68 (SD = 1.03) 2.15 (SD = .84) Intent to Quit 2.78 (SD = .83) 2.92 (SD = .76) Dissatisfaction 2.81 (SD = .68) 2.97 (SD = .62) Company Content 3.41 (SD = .35) 3.37 (SD = .32) Testing Hypotheses Hypothesis #1 predicted that relationships would exist between perceived supervisor reward and punishment behaviors and perceived power sources. Positive relationships should exist between contingent reinforcement and legitimate, reward, referent, and expert sources of power, and a negative relationship between contingent reinforcement and coercive sources of power. Referent and expert power also should have the strongest correlations with contingent reinforcement. Referent and expert should be negatively correlated with non-contingent reinforcement. With management contingent punishment, positive relationships were predicted between contingent 32 Table 6 Self-Reported Employee Behaviors Employee Behaviors Supervisors N = 29 Subordinates N = 92 Absenteeism 1.45 (SD = .40) 1.48 (SD = .45) Substance Abuse 1.32 (SD = .52) 1.25 (SD = .47) Privilege Abuse 1.83 (SD = .82) 1.59 (SD = .62) Theft 1.49 (SD = .51) 1.32 (SD = .42) punishment and legitimate sources of power as well as expert sources of power. Noncontingent punishment was also predicted to be negatively correlated with legitimate, reward, referent, and expert sources of powers. Non-contingent punishment was predicted to be positively correlated with coercive sources of power. Pearson product moment correlations explored possible relationships between contingent reinforcement and each source of power. For this section, the subordinate data was analyzed separate from the supervisor data because the subordinates provide the most relevant responses in terms of perceived supervisory sources of power. For all the hypotheses, bivariate correlations were used. Although the number of separate correlations is large (25), the total number of significant findings exceeded chance levels for multiple tests. Results indicated there were significant positive correlations between contingent reinforcement and legitimate power, r(92) = .335, p < .005, reward power, r(92) = .255, p < .005, referent power, r(92) = .367, p < .001, and expert power, r(92) = 33 .337, p < .005. A Pearson correlation analysis also indicated that there was a significant negative correlation between contingent reinforcement and coercive power, r(92) = -.273, p < .01. Subordinates who perceived supervisors as displaying contingent reinforcement also perceived supervisors as using legitimate, reward, referent, and expert power, as well as using less coercive power. Pearson product moment correlations explored possible relationships between non-contingent reinforcement and the personal bases of power, referent and expert power. There were no significant findings Pearson product moment correlations explored possible relationships between contingent punishment and each source of power; legitimate, coercive, reward, referent, and expert. A significant positive correlation was found for contingent punishment and both legitimate power, r(91) = .303, p < .005, and expert power, r(92) = .238, p < .01. Subordinates who perceived supervisors as displaying contingent punishment also perceived supervisors as using legitimate and expert power. Contingent punishment was not significantly correlated with reward, coercive, and referent power. Pearson product moment correlations explored possible relationships between non-contingent punishment and bases of power. A significant correlation was found for non-contingent punishment and legitimate power, r(92) = -.218, p < .05. Subordinates who perceived supervisors as displaying non-contingent punishment also perceived supervisors as not using legitimate power. Non-contingent punishment was not significantly correlated with coercive, reward, referent, and expert powers. See Table 7 for supporting data. Hypothesis 2 predicted significant relationships between perceived supervisory reward and punishment behavior and both employee attitudes and employee 34 Table 7 Sources of Power and Contingent/Non-Contingent Punishment and Reinforcement Legitimate Power Contingent Reinforcement Contingent Punishment NonContingent Reinforcement NonContingent Punishment .335** .303** .055 -.218 Coercive Power - .273** .125 -.066 .186 Reward Power .255* -.041 .232 -.132 Referent Power .367** .123 .081 -.158 Expert Power .337** .238* .038 -.156 **p < .001 *p < .01 behaviors. For employee attitudes, I predicted significant negative relationships between contingent reinforcement and approval of theft, intent to quit, and job dissatisfaction, as well as a positive relationship with being content with the company they work for. For employee behaviors, I predicted negative relationships between contingent reinforcement and absenteeism, substance abuse, and theft. With regard to non-contingent punishment and employee attitudes, I predicted positive relationships with approval of theft, intent to quit, and job dissatisfaction. I also predicted a negative relationship between non-contingent punishment and company content, which is a construct which considers an employee’s overall satisfaction with a company an their policies. Considering employee behaviors, I also predicted positive 35 relationships between non-contingent punishment and absenteeism, substance abuse, privilege abuse, and theft. Pearson product moment correlations explored possible relationships between contingent reinforcement and each employee attitude. Results indicated there was a significant negative correlation between contingent reinforcement and dissatisfaction, r(92) = -.325, p < .005. There was a significant positive correlation between contingent reinforcement and contentment with the company, r(92) = .335, p < .001. Subordinates who perceived supervisors as displaying contingent reinforcement, reported less dissatisfaction and greater contentment with the company. There were no significant correlations between contingent reinforcement and theft approval and intent to quit. Pearson product moment correlations explored possible relationships between non-contingent punishment and employee attitudes. Results indicated a significant positive correlation between non-contingent punishment and dissatisfaction, r(92) = .314, p < .005. Subordinates who perceived supervisors as displaying non-contingent punishment, reported more dissatisfaction. There were no significant correlations between non-contingent punishment and theft approval, intent to quit, and company content. Pearson product moment correlations explored possible relationships between contingent reinforcement and each deviant employee behavior. There were no significant correlations between perceived leader reinforcement behavior and deviant employee behaviors. Pearson product moment correlations also explored possible relationships between non-contingent punishment and each deviant employee behavior, which also revealed no further significant correlations. See Table 8 for supporting data. 36 Table 8 Employee Attitudes/Behaviors and Contingent/Non-Contingent Punishment and Reinforcement Dissatisfaction Company Content Contingent Reinforcement Contingent Punishment NonContingent Reinforcement NonContingent Punishment -.325* .163 .020 .335* .208 .301* -.246 .314* Intent to Quit -.202 .217 .066 .227 Theft Approval -.134 -.043 .118 .186 Absenteeism -.009 .132 .015 .055 Substance Abuse .019 .074 .155 .094 Privilege Abuse .022 .132 .131 .074 -.077 .117 .057 .103 Theft **p < .001, *p < .01 Hypothesis 3 predicted relationships between power sources and employee attitudes. With regard to reward, referent, and expert sources of power, I predicted negative relationships between these power sources and theft approval, intent to quit, and dissatisfaction. I predicted a positive relationship between the power sources and company content. I predicted positive relationships between coercive power source and theft approval, intent to quit, and dissatisfaction, as well as a negative relationship between coercive power source and company content. 37 Relationships should also exist between power sources and employee behaviors. For reward, referent, and expert sources of power, I predicted negative relationships between these power sources and absenteeism, substance abuse, privilege abuse, and theft. With regard to coercive power, I predicted positive relationships between absenteeism, substance abuse, privilege abuse, and theft. Pearson product moment correlations were completed between bases of power and attitudes. Results revealed significant correlations between coercive power and theft approval, r(92) = .348, p < .005, intent to quit, r(92) = .444, p < .001, and dissatisfaction, r(92) = .458, p < .001. Subordinates who reported and perceived supervisors as using coercive power tended to be more approving of theft, have more intentions to quit, and be more dissatisfied with their job. Pearson product moment correlations were also conducted to analyze relationships between employee behaviors and power sources. No significant correlations were found for relationships between referent, expert, legitimate and reward power and the constructs measuring employee attitudes and behaviors. See Table 9 for supporting data. Hypothesis 4 predicted relationships between identified demographic variables and employee attitudes. With regard to age, education, continuous employment, and tenure, I predicted negative relationships between these and theft approval, intent to quit and dissatisfaction. I predicted positive relationships between company content, age, education, continuous employment, and tenure. Relationships were also predicted to exist between demographic variables and employee behaviors. With regard to age, education, continuous employment, and tenure, I predicted negative relationships between these demographic variables and absenteeism, substance abuse, privilege abuse, and theft. 38 Table 9 Employee Attitudes/Behaviors and Sources of Power Legitimate Coercive Power Power Dissatisfaction Company Content -.148 .263 .458** .039 Reward Power Referent Power Expert Power -.187 -.268 -.144 .073 .173 .167 Intent to Quit -.065 .444** -.139 -.129 -.071 Theft Approval -.176 .348* -.121 -.140 -.096 Absenteeism -.090 .130 -.023 -.112 .051 Substance Abuse -.022 .195 .116 -.053 .081 Privilege Abuse -.223 .147 -.131 -.171 -.115 Theft -.221 .256 -.070 -.186 -.098 **p < .001 *p < .01 Supervisors and subordinates were analyzed as group in these analyses because the content of the questions were related to each individual’s attitudes and personal demographic information. This approach was considered appropriate because these variables were not related to the interactions that occur between the supervisors and subordinates in regard to sources of power or reinforcement and punishment behaviors. Pearson product moment correlations revealed significant correlations for age and intent to quit, r(121) = -.323, p < .001 and dissatisfaction, r(121) = -.344, p < .001. Data from supervisors and subordinates suggested that older persons had less intent to quit and less dissatisfaction. 39 Pearson product moment correlations also revealed significant correlations for non-continuous employment and absenteeism, r(122) = .323, p < .001, privilege abuse, r(122) = .236, p < .001, and theft, r(122) = .349, p < .001. Continuous employment was associated with less absenteeism, privilege abuse, and theft. The predicted relationships with education, continuous employment, and tenure did not result in significant findings with theft approval, intent to quit, and dissatisfaction. Additionally, the predicted relationships with age, education, and tenure did not result in significant findings with absenteeism, substance abuse, privilege abuse, and theft. Tenure, interestingly was positively correlated with theft approval, opposite as predicted, r(122) = .301, p < .001. Refer to Table 10 for further supporting data. Hypothesis 5 predicted that relationships should exist between store location and employee attitudes and behaviors. I predicted that pre-identified, high-deviance stores would report higher levels of theft approval, intent to quit, dissatisfaction, absenteeism, substance abuse, privilege abuse, and theft, when compared to the preidentified, low-deviance stores. The pre-identified high-deviance stores were also predicted to report lower levels of company content compared to the pre-identified lowdeviance stores. Independent-samples t-tests were conducted with high/low deviance stores as the independent variable and employee attitudes and behaviors as the dependent variables. The results indicated no significant differences in employee attitudes and deviant employee behaviors across store locations. 40 Table 10 Employee Attitudes/Behaviors and Demographic Data Age Gender Dissatisfaction -.344** -.089 .080 -.204 .164 Company Content -.171 .000 .142 -.091 .101 Intent to Quit -.318** .004 .055 -.197 .086 .140 -.086 .158 Absenteeism -.012 -.027 Substance Abuse -.077 Privilege Abuse Theft Theft Approval Education Tenure NonContinuous Employment .301** .082 -.088 .176 .323** -.131 .006 .122 .195 -.198 -.004 .006 .097 .236** -.062 -.059 .040 .173 .349** **p < .001 *p < .01 Discussion Supervisors can influence subordinate behavior by administering rewards and punishments. To most effectively influence behavior, rewards and punishments should be administered contingent to the specific behavior (Hinkin & Schreisheim, 1994). Supervisors can also influence subordinate behavior through any of their five power bases: Legitimate, Coercive, Reward, Referent, and Expert Power, (French & Raven, 1959). 41 The primary focus behind this study was to explore relationships between employee attitudes and behaviors with respect to perceptions of supervisory use of reinforcement, punishment, and sources of power. Relationships between perceived supervisory reinforcement behavior and perceived supervisory sources of power were also assessed, extending the research of Hinkin and Schrieshiem (1994). Support was found for many of the hypothesized relationships. Supervisor and Subordinate Perceptions An important focus within this study was exploring the perceptions of both subordinates and supervisors concerning supervisory use of reinforcement behavior and bases of power. This is an important subject because perception is reality when considering the interactions between this dyad. A supervisor’s use of power or their issuance of a reward or punishment is only as effective as the subordinate’s perception. Some common themes relative to this sample include supervisor’s overestimating their effectiveness. Supervisors interestingly rated themselves significantly higher in both of the personal-interaction based powers, referent and expert power, when compared to their subordinate’s assessments. Supervisors also rated themselves having used contingent reinforcement significantly more than the subordinates had reported. These are all particularly interesting results because differences in responses ultimately reflect how accurately the supervisors perceive their own leadership behaviors. Since accuracy is “in the eyes of the beholder,” the responses which, most relevantly reflect the power sources and reinforcement and punishment behaviors, are the subordinate responses. 42 Hypothesis 1: Supervisor Behaviors and Power Sources Perceived Contingent Reward Behavior. Hinkin and Schreisheim (1994) specifically examined relationships between subordinate perceptions of supervisory reinforcement behaviors and the five French and Raven (1959) bases of social power in employees of a hospital and mid-level managers in 100 financial institutions. The present study replicated their findings. Hypothesis 1 was primarily supported. Consistent with Hinkin and Schriesheim (1994), the use of contingent rewards had the greatest effect on enhancing the four most favorable power bases, legitimate, reward, referent, and expert, while also reducing the least favorable power base, coercive power. Perceived contingent reward behavior significantly correlated with legitimate power, reward power, referent power, expert power, and negatively with coercive power. The strongest effect was with one of the more positively perceived power sources, referent power, which is developed based on personal interactions (Podsakoff & Schriesheim, 1986). These finding reflects the notion that referent power is a soft power that is developed through personal interactions. Reliance on these soft powers is viewed more favorably by subordinates, eliciting compliance in a positive form. These findings are consistent with previous research suggesting contingent reinforcement behavior as being related to the sources of power and most similarly related to soft bases of power, which are also most associated with positive subordinate satisfaction (Williams & Podsakoff, 1988). The results are also consistent with Carson et al.’s (1993) meta-analysis examining the interrelationships and outcomes of the five French and Raven (1959) social power 43 bases. When researching supervisor satisfaction, job satisfaction, and job performance, the power sources developed through personal-interactions, referent and expert powers, generated the most desirable outcomes. Perceived Contingent Punishment Behavior. Consistent with the hypotheses, a supervisor’s perceived contingent punishment behavior was positively associated with both legitimate and expert sources of power. Punishment appears to enhance legitimate power and expert power when it is contingent on behavior. This suggests that supervisors should not resist punishing inappropriate subordinate behavior just as long as the punishment is administered contingent upon the specific inappropriate behavior. Perceived Non-contingent Punishment and Non-contingent Reinforcement Behavior. The perceived used of non-contingent punishment and non-contingent reinforcement did not have the predicted relationship with the power sources. Consistent with past research (Hinkin & Schreisheim, 1994) noncontingent consequences were predicted to be less connected with power bases, however, several predictions were made and not supported by the data. Continued research on this topic might better elucidate the relationship between non-contingent consequences and bases of power. Hypothesis 2:Leader Behavior and Employee Attitudes and Behaviors Hypothesis 2 predicted relationships between supervisor reinforcement and punishment behaviors with employee attitudes and behaviors. As predicted, contingent reinforcement was both negatively correlated with employee reported dissatisfaction, and positively related to employee reported contentment with the company. Essentially, when 44 supervisors use contingent reinforcement, employees are less dissatisfied and are more content with the company they work for. The results also revealed non-contingent punishment was positively correlated with dissatisfaction, which supported the second hypothesis. Interestingly non-contingent punishment was perceived significantly less by supervisors compared to subordinates while contingent reinforcement was perceived significantly more by supervisors compared to subordinates. This result suggests that supervisors must improve on the selfawareness of their own behavior. Having a skewed perspective of their effectiveness, as leaders, will hinder the supervisor’s ability to reinforce and change subordinate behavior. Some of the hypothesized predictions did not surface potentially because of a self-report bias, most particularly with the Employee Perceptions Survey. Many of the questions within the survey were controversial by nature, with the content including topics of employee deviance. Some participants chose not to answer some of these questions, and during data collection a few participants asked for further clarification of their confidentiality rights. Hypothesis 3: Sources of Power and Employee Attitudes and Behaviors With respect to the sources of power, one of the most revealing findings from this study includes strong positive correlations between the perceived supervisory use of coercive power and subordinate theft approval, intent to quit, and dissatisfaction. Previous research suggests coercive power to be associated with a decreased attraction from the subordinate to the supervisor, often spreading negative valences onto other regions of the life space (French & Raven, 1959). 45 Theft approval, intent to quit, and dissatisfaction are all negatively oriented constructs, generally intended to be at a minimum. The use of coercive power should be strongly discouraged and generally seen a last resort option. Results of the present study support the negative connection between perceived coercive power and negative employee attitudes and behaviors. Some of the hypotheses from this section were not supported and it is believed to again be primarily due to self-report biases for the Employee Perceptions Survey. There was more support for the predicted hypotheses concerning the Attitudes section, as opposed to the Behaviors section. This could suggest that participants were more comfortable and honest when answering the Attitudes section of The Employee Perceptions Survey, compared to the Behaviors section. Hypothesis 4: Demographic Data and Employee Attitudes and Behaviors Other considerations for this project were demographic constructs such as age and employment status. These items were analyzed across the subordinate attitudes and behaviors. The findings from this sample showed age to be negatively correlated with intent to quit and also dissatisfaction. This is consistent with previous research (Hollinger & Clark, 1983), and fairly logical considering the natural progression of life including acceptance, consistency, and more satisfaction. The findings from this sample also supported Hollinger and Clark (1983), showing non-continuous employees as more likely to be absent, abuse their privileges, and steal, when compared to employees who have had continuous employment. 46 Employers should be cautious when considering bringing back previous employees. Diligence in their process and careful consideration of ethics would be best. Other relationships that were hypothesized were not supported. Based on Hollinger and Clark’s (1983) research, education level was predicted to be related to more positive employee attitudes and behaviors, such as job satisfaction. No significant relationships were found in either direction for education level, which likely is related the relatively low education level of the demographic majority. Tenure at Sisco Enterprises was also predicted to be related to the more positive constructs within employee attitudes and behaviors and not related to the less favorable items. However, the results were surprisingly supportive in the opposite direction with tenure positively correlated with theft approval. This would be a valuable direction for future research. Hypothesis 5: Store Location and Employee Attitudes and Behaviors In this hypothesis, relationships were expected between store location and employee attitudes and behaviors. Pre-identified, high-deviance stores were expected to report higher levels of theft approval, higher intentions to quit, more dissatisfaction, more absenteeism, more substance and privilege abuse, and higher incidents of theft, when compared to pre-identified, low-deviance stores. The results of the study did not support any of these hypotheses indicating no difference between the pre-identified high and low deviance stores relative to the constructs measured. The Employee Perceptions Survey presented reporting accuracy challenges, as previously stated. It is also possible that some of the deviant employees 47 were no longer working at the stores, compared to the original time the stores were preidentified as high-deviance. Limitations The design of the study included a homogeneous population of 122 Sisco Enterprises employees from six Round Table Pizza locations. Of the 122 participants, 92 were subordinates and 30 were supervisors. Of their 23 store locations, three preidentified high-deviance locations were included in the target population along with three randomly selected locations. All participants completed the Sources of Power measure, the Leader Reward and Punishment Questionnaire, the Employee Perceptions Survey, along with additional demographic data. The participants completed the surveys in a 30-minute time period, in a one-on-one basis with the proctor, while on the job. This method strengthened the external validity of the study, however a response bias was also a strong possibility. Given the nature of the topic, a homogeneous population became necessary with respect to the context in which I measured perceived sources of power, perceived leader reward and punishment behavior, and employee attitudes and behaviors. This was important so that each participant could be measured consistently across the same scale. One of the strengths of this study includes having achieved this goal. All participants responded to the survey questions from the perspective and experience of working for their current employer, Sisco Enterprises. They also responded to these questions while at the workplace, thus further strengthening the results of my study with a real-world environment. 48 While a homogeneous population was achieved, this procedure does however introduce potential limitations to the study. The participants represent only one organization, Sisco Enterprises These findings should easily generalize to other customer service type environments but they may not be valid for other comparisons. The design of the study still attains sound internal consistency, based on the measures used in previous studies, the generalizability of the study remains weak based on the sample population being from one employer in one demographic. Many of the research findings were compelling, yet did not meet the minimum qualifications determining statistical significance. Many different variables can contribute error into any study, which merits the appropriateness of exploring these relationships again in future studies. While coercive power did not show the specific relationships as expected with some of the more negative employee attitudes and behaviors, such as theft, absenteeism, or privilege abuse, I believe this is a relationship worth exploring further. I believe this because one of the weaknesses of my study includes the potential for incorrect responses specifically with the deviance measure, via self-report biases or dishonest responses. The Employee Perceptions Survey asked many questions that were controversial by nature and could be viewing as self-incriminating. Due to this, self-report biases are likely a source of error within some of the results. I would suggest for further researchers to consider either using a different measure for assessing deviance or developing an alternative, sound procedure that is even more assuring that responses remain anonymous and non-reciprocated. 49 Understanding deviance behaviors is important. It is recommended that future researchers include a more appropriate method of gathering deviance information. The current instrument was effective in gathering information relevant to the study, however there was a large potential for self-report biases and dishonest responses based on fears of retaliation from one’s employer. I also suggest modifying the environment or context in which the questions are asked. While asking responses while employees were on the clock provided strong validity, it may have introduced error into some of the measures and damaged reliability, most particularly the more controversial measure, The Employee Perceptions Survey. CHAPTER V SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS The findings of this study contribute further towards a better understanding of the relationships that exist between supervisors and subordinates. The quality of this dyad can have a major impact on the success of a business. One of the objectives behind this study was to explore workplace deviance in relation to supervisors and subordinates. The concept follows Robinson and Greenberg’s (1998) findings, which suggest that unfair interpersonal treatment is prominent in social influence on deviance. One of the findings in the current study suggested that perceived use of coercive power was positively correlated with theft approval, intent to quit, and dissatisfaction. The results of the study replicated Hinkin and Schriesheim’s (1994) findings and showed perceived supervisory reinforcement and punishment to be clearly related to perceived sources of power, and when used effectively and appropriately, perceived power is enhanced. Contingent reinforcement shared a positive relationship with legitimate power, reward power, referent power, and expert power, including a negative relationship with coercive power. Contingent punishment also shared a positive relationship with legitimate power and expert power. Reinforcement is a form of reward within itself and when used appropriately it merits legitimacy and respect, as reflected in referent and expert power. It is no surprise 50 51 that as predicted, the personal powers, referent and expert power, showed the strongest correlations with contingent reinforcement. There were no significant findings for noncontingent punishment and non-contingent reinforcement. While the predicted relationships were found between supervisors and subordinates in regard to sources of power and reinforcement and punishment behaviors, the predicted relationships with employee deviance were not as clearly supported. Many of the predicted relationships with employee attitudes were supported, however none of the predicted relationships with employee behaviors were backed by the data. As for employee attitudes, the perceived use of contingent reinforcement appears to be consistent with employees who are satisfied and content with the company. The perceived use of non-contingent punishment appears to be consistent with employees who are dissatisfied and not content with the company. The perceived use of coercive power also had strong relationships with undesirable employee attitudes, including theft approval, intent to quit, and dissatisfaction. I would encourage business owners to businesses utilize the more responsive forms of social power and encourage their supervisors to use their reward, referent, and expert powers. It is also important to focus on having reinforcements and punishments contingent on the specific behavior. Both increasing the extent of a supervisor’s influence and these leadership behaviors are associated with higher job satisfaction and better performance (French & Raven, 1959; Williams & Podsakoff, 1988). Hinkin and Schriesheim (1994) appropriately tie the concepts together by suggesting, “it would be beneficial for managers to take advantage of the opportunities to 52 provide contingent rewards, to avoid the use of non-contingent punishment, and to not be hesitant to use contingent punishment” (p. 791). REFERENCES REFERENCES Ashforth, B. E. (1997). Petty tyranny in organizations: A preliminary examination of antecedents and consequences. Canadian Journal of Administrative Sciences, 14(2), 126-140. Ashforth, B. E., Sluss, D. M., & Harrison, S. H. (2007). Socialization in organizational contexts. International Review of Industrial and Organizational Psychology, 22, 1-70. Avolio, B. J., & Bass, B. M. (1988). Transformational leadership, charisma, and beyond. In J. G. Hunt, B. R. Baliga, H. P. Dachler, & C. H. Schriesheim (Eds.), Emerging leadership vistas (pp.29-50). Lexington, MA: Lexington Books. Bennett, R. J., & Robinson, S. L. (2003). The past, present, and future of workplace deviance research. In J. M. Greenberg (Ed.), Organizational behavior: The state of the science (2nd ed., pp. 247-281). Hillsdale, NJ: Lawrence Erlbaum Associates. Bolin, A., & Heatherly, L. (2001). Predictors of employee deviance: The relationship between bad attitudes and bad behavior. Journal of Business and Psychology, 15(3), 405-418. Carson, P. P, Carson, K. D., & Roe, W. (1993). Social power bases: A meta-analytic examination of interrelationships and outcomes. Journal of Applied Social Psychology, 23(14), 1150-1169. 54 55 Diefendorff, J. M., & Mehta, K. (2007). The relations of motivational traits with workplace deviance. Journal of Applied Psychology, 9(4), 967-977. Fox, S., & Spector, P. E. (1999). A model of work frustration-aggression. Journal of Organizational Behavior, 20(6), 915-931. French, J. R. P., Jr., & Raven, B. H. (1959). The bases of social power. In D. Cartwright (Ed.), Studies in social power (pp. 150-167). Ann Arbor, MI: Institute for Social Research. Hinkin, T. R., & Schriesheim, C. A. (1989). Development and Application of New Scales to Measure the French and Raven (1959) Bases of Social Power. Journal of Applied Psychology, 74, 4, 561-567. Hinkin, T. R., & Schriesheim, C. A. (1994). An examination of subordinate-perceived relationships between leader reward and punishment behavior and leader bases of power. Human Relations, 47(7), 779-791. Hollinger, R. C., & Clark, J. P. (1983). Theft by employees. Lexington, MA: Lexington Books. Hulin, C. (1991). Adaptation, persistence, and commitment in organizations. In M. D. Dunnette & L. M. Hough (Eds.), Handbook of industrial and organizational psychology (2nd ed., Vol. 2, pp. 445-505). Palo Alto, CA: Consulting Psychologists Press. Kamp, J., & Brooks, P. (1991). Perceived organizational climate and employee counterproductivity. Journal of Business and Psychology, 5(4), 447-458. 56 Koslowsky, M., Schwarzwald, J., & Ashuri, S. (2001). On the relationship between subordinates compliance to power sources and organizational attitudes. Applied Psychology: An International Review, 50(3), 455-476. Langton, L., & Hollinger, R. C. (2005). An empirical test of the relationship between employee theft and low self-control. Deviant Behavior, 27(5), 537-565. Lee, K., & Allen, M. (2002). Organizational citizenship behavior and workplace deviance: The role of affect and cognitions. Journal of Applied Psychology, 87(1), 131-142. Mangoine, T. W., & Quinn, R. P. (1975). Job satisfaction, counterproductive behavior, and drug use at work. Journal of Applied Psychology, 60(1), 114-116. Mitchell, M. S., & Ambrose, M. L. (2007). Interpersonal deviance, organizational deviance, and their common correlates: A review and meta-analysis. Journal of Applied Psychology, 92(4), 1159-1168. Podsakoff, P. M., & Schriesheim, C. A. (1984). Leader reward and punishment behavior: A methodological and substantive review. Unpublished manuscript, Graduate School of Business, Indiana University, Bloomington, IN. Podsakoff, P. M., & Schriesheim, C. A. (1986). Leader reward and punishment behavior: A conceptual, methodological, and integrative review. Bloomington, IN: Graduate School of Business, Indiana University. Podsakoff, P. M., & Skov, R. (1980). Leader reward and punishment behavior scales. Unpublished research, Graduate School of Business, Indiana University, Bloomington, IN. 57 Richardson, L. D., Swan, J. E., & Hutton, J. D. (1995). The effects of the presence and use of channel power sources on distributor satisfaction. The International Review of Retail, Distribution and Consumer Research, 5(2), 185-201. Robinson, S. L., & Bennett, R. J. (1995). A typology of deviant workplace behaviors: A multi-dimensional scaling study. Academy of Management Journals, 38(2), 555-572. Robinson, S. L., & Greenberg, J. (1998). Employees behaving badly: Dimensions, determinants and dilemmas in the study of workplace deviance. Trends in Organizational Behavior, 5, 1-30. Schriesheim, C. A., Hinkin, T. R., & Tetrault, L. A. (1991). The discriminant validity of the leader reward and punishment questionnaire (LRPQ) and satisfaction with supervision: A two-sample, factor analytic investigation. Journal of Occupational Psychology, 64(2), 159-166. Tepper, D. T. (2000). Responses to Cohen’s (2000) case study. Consulting Psychology Journal: Practice and Research, 52(2), 150-152. Williams, M. L., & Podsakoff, P. M. (1988). A meta-analysis of attitudinal and behavioral correlates of leader reward and punishment behaviors. Southern Management Association Proceedings, 161-163.