.. a; J: . , ‘ .. 2.x. ‘ .5 e5 1., . r figmmmafimg . fiflmfi, a an». ‘ $5,. a. ‘ f .U oi .m... t L 19%.», 3.». _ . .. , .33., ._ 3ft nfigy ., . ‘ f 4 .V 4.. 1; (4 3143:). ‘r -fi#§fi§w¢ i ” 44mg; . .fiuumnr . A 3. 5L 4...: :«A 4;... .‘ or 45.3 .. L 5%.» 3 1“ I cry a. fi 3... u . 3.. . «(k i, . i p. s 9. an, S 5.23.. ., . Elsi 41.: z) .0: n7 3 .8. n. 1 A. g od...9u.£9 ,grt‘iti: 2:53.12. 1-1:. I!!!‘ it}.!.l [(rfl-h21‘ ‘ : vg'h. .Wmufimxi I. Lhfiflflflfin .. :x:|.f . . 9‘ This is to certify that the : thesis entitled ' RELATIONSHIP DIMENSION AND TYPE OF INEQUITY: RESTORATIONOF EQUITY IN COWORKER FRIENDSHIPS presented by CATHERINE Y. KINGSLEY has been accepted towards fulfillment of the requirements for the M. A. degree in Communication 7é/t/J/W Major Professor's Signature 0174/ 067/ 017094 Date MSU is an Affirmative Action/Equal Opportunity Institution LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE UK FEB 2‘11: 290.6 0 5 Lin I 58 e25 ~%-§tr~~ 6/01 c:/ClFlC/DateDue.p65-p.15 RELATIONSHIP DIMENSION AND TYPE OF IN EQUITY: RESTORATION OF EQUITY IN COWORKER FRIENDSHIPS By Catherine Y. Kingsley A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Communication 2004 ABSTRACT RELATIONSHIP DIMENSION AND TYPE OF INEQUITY: RESTORATION OF EQUITY IN COWORKER FRIENDSHIPS By Catherine Y. Kingsley Equity theory indicates that people feel distress in the face of inequity in their relationships. If people are friends at work in a task-oriented setting and also outside of work in a social setting, hypothetically they should react differently to inequity in one versus the other setting. This was tested in the setting of multidimensional friendships known as blended friendships. Small but significant differences were found in responses due to relationship dimension, and more substantial differences were found due to the type of inequity experienced (underreward/ overreward). Interaction effects between relationship dimension and type of inequity were found as well. Gender differences which were not hypothesized were also found. ACKNOWLEDGEMENTS First, I want to thank my committee members, Janet Lillie and John Beck, for their helpful comments and ideas for future studies. Second, a sample of working adults is not easy to achieve. Students are less difficult, but require generous professors and graduate students. I want to thank those who helped with both my student and working adult samples. Without my parents support, both emotional and monetary, for umpteen years of school, I wouldn’t even be here to write a thesis. Thanks for that, Mom and Dad. Unlike my parents, who witnessed the process from afar, some other people had to see it close up. At times, it wasn’t pretty! Particular thanks to Dave and to all others in the Communication department who were around to lighten my day or offer a supportive comment. Finally, Hee Sun, I can’t tell you how much your help on this thesis and in general has meant—you have been a sounding board, an editor, a mentor, a friend, an “academic” parent, and much more. I continue to appreciate your expertise in every aspect of research and the way you are willing to share it with me any time. Tim, thanks for sharing her. iii TABLE OF CONTENTS LIST OF TABLES ............................................................................................................. vi LIST OF FIGURES .......................................................................................................... vii Introduction ......................................................................................................................... 1 Literature review ................................................................................................................. 3 Equity Theory .................................................................................................................. 3 Equity theory: Task dimension .................................................................................... 9 Equity theory: Social dimension ............................................................................... 15 Friendship in the workplace ......................................................................................... 18 Rationale ....................................................................................................................... 20 Method .............................................................................................................................. 28 Pilot Study ..................................................................................................................... 28 Participants ............................................................................................................... 28 Design ....................................................................................................................... 28 Procedure .................................................................................................................. 31 Measurement ............................................................................................................. 3 1 Realism .................................................................................................................. 32 Fairness ................................................................................................................. 32 Pilot Study Results ........................................................................................................ 33 Realism ...................................................................................................................... 33 Fairness ..................................................................................................................... 35 Summary ................................................................................................................... 37 Primary Study ............................................................................................................... 38 Participants ............................................................................................................... 38 Design ....................................................................................................................... 39 Procedure .................................................................................................................. 42 Measurement ............................................................................................................. 42 Realism .................................................................................................................. 43 Fairness ................................................................................................................. 44 Liking .................................................................................................................... 45 Equity restoration ................................................................................................. 46 Change in own behavior ....................................................................................... 47 Changing other '3 behavior ................................................................................... 47 Manipulation check ................................................................................................... 48 Realism .................................................................................................................. 48 Fairness ................................................................................................................. 49 Results ............................................................................................................................... 51 Primary Study Results ................................................................................................... 51 Change own behavior and changing other '5 behavior ............................................. 52 Change in Liking ....................................................................................................... 55 Discussion ......................................................................................................................... 57 Relationship dimension ................................................................................................. 58 Behavioral restoration .............................................................................................. 58 Psychological restoration ......................................................................................... 61 Equity level .................................................................................................................... 61 iv Psychological restoration ......................................................................................... 61 Behavioral restoration .............................................................................................. 62 Relationship dimension and equity level ....................................................................... 63 Behavioral restoration .............................................................................................. 63 Psychological restoration ......................................................................................... 64 Gender and friendship composition .............................................................................. 66 Limitations .................................................................................................................... 68 Future directions ........................................................................................................... 72 Conclusion .................................................................................................................... 77 APPENDIX A ................................................................................................................... 80 APPENDIX B ................................................................................................................... 85 APPENDIX C ................................................................................................................... 86 APPENDIX D ................................................................................................................... 88 REFERENCES ............................................................................................................... 1 12 LIST OF TABLES Table 1: Means and Standard Deviations for Realism and Relationship foundation/ Relationship Dimension/ Equity Level ANO VA ........................................................ 88 Table 2: Means and Standard Deviations for Realism and Gender ANO VA ................... 90 Table 3: Means and Standard Deviations for Realism and Order ANO VA ..................... 91 Table 4: Means and Standard Deviations for Fairness and Relationship foundation/ Relationship Dimension/ Equity Level ANO VA ........................................................ 92 Table 5: Means and Standard Deviations for Fairness and Gender ANO VA .................. 94 Table 6: Means and Standard Deviations for Fairness and Order ANOVA .................... 95 Table 7: Means and standard deviations for change own behavior, attempt to change other ’5 behavior, and liking change (categorized by equity level and relationship dimension) ................................................................................................................. 96 Table 8: Zero-order correlations for change own behavior, attempt to change other 's behavior, and liking change (categorized by equity level and relationship dimension) ................................................................................................................. 97 Table 9: Means and Standard Deviations for Type of C hange/ Gender/ Equity level/ Relationship dimension ANOVA ............................................................................... 98 Table 10: Means and Standard Deviations for participant gender/ Chris 'gender/ relationship dimension/ equity level ANO VA on Liking ......................................... 100 Table 11: Means and Standard Deviations for participant gender/ Chris ’gender/ relationship dimension/ equity level ANO VA on changing own behavior .............. 103 Table 12: Means and Standard Deviations for participant gender/ Chris ’ gender/ relationship dimension/ equity level ANO VA on changing one ’s partner ’s behavior ................................................................................................................................. 105 vi LIST OF FIGURES Figure 1: Interaction between type of change and equity level ...................................... 107 Figure 2: Means for interaction between equity level, relationship dimension, and type of change ..................................................................................................................... 108 Figure 3: Interaction between type of change and gender .............................................. 109 Figure 4: Interaction between type of change, gender, and equity level ........................ 110 Figure 5: Interaction between task/social and over/underreward for change in liking. 11 1 vii Introduction The friendships that most people develop in high school or college may be difficult to maintain as those friends become more scattered. In today’s society, living in any part of the world is easily done, making moving to far-off locations more common, and causing people, out of necessity, to form friendships with the most available people: their coworkers. When one realizes that one of many known factors of attraction is mere exposure (Brockner & Swap, 1976; Saegert, Swap, & Zajonc, 1973), the amount of time spent with coworkers commands a little more importance. People spending 40 hours of their week in the workplace will develop some type of friendships there whether they want to or not. If people will form one relationship based on mere exposure, it seems likely that they would form many relationships based on mere exposure. Hence, not only single relationships, but social networks are formed based on coworker exposure. The fact that many people have not only one friendship with a coworker, but have developed social networks around coworkers, indicates that finding out more about coworker friendships that extend beyond the workplace is a worthy task. Friendships are interpersonal relationships characterized by mutual productivity and mutual positive regard. Expanding upon that, blended friendships are multidimensional relationships also characterized by mutual productivity and positive regard that encompass both workplace interaction and interaction outside of the workplace (Bridge and Baxter, 1992). A horizon yet to be examined is that of how the members of these blended friendships respond to inequity in different dimensions of their multidimensional friendships. Equity theory has been extended by Walster, Berscheid, and Walster (1973) to encompass not only economics as it has traditionally, but also social relationships. Further, equity theory also possesses predictive ability in task- oriented relationships, such as those of coworkers. This study seeks to test the predictive ability of equity theory when inequity exists in either the task (work) dimension or the social dimension of a blended friendship. An in-depth discussion of equity theory will be offered, followed by presentation of supporting and dissenting literature, elaboration on blended friendship, and a rationale and hypotheses. Following that will be a method section, and finally results and discussion of the results. Literature review Equity Theory According to Walster et a1. (1973), an equitable relationship is one in which either the participant or an outside observer perceives all participants to have equal relative outcomes. Outcomes can further be defined as either rewards (positive outcomes) or costs (negative outcomes). The participant in the relationship will also be referred to as the “perceiver”, because he/she is the one who perceives whether or not the relationship is equitable. For example, two coworkers, Heidi and Hannah, who have known each other for a long time, may consider themselves friends. Heidi spends time doing things like picking up lunch for Hannah, calling Hannah and providing support to Hannah regarding job-related issues. Hannah enjoys the friendship, but does not put in the same efforts as Heidi. Heidi may perceive this relationship to be inequitable because her rewards and costs from the friendship are not proportional to Hannah’s rewards and costs. In order for the friendship to be equitable from Heidi’s point of View, Heidi must perceive that the ratio of Hannah’s costs to her rewards is equal to the ratio of Heidi’s costs to her rewards. Walster et al. (1973) put forth four equity theory propositions regarding how individuals behave in relationships. First, “individuals will try to maximize their outcomes where outcomes equal rewards minus costs” (Walster et al., 1973, p. 151). This is based on the fundamental idea that humans are selfish creatures and without any governance, would arrange things to their advantage in any way possible. However, this unbounded desire for maximizing outcomes is tempered by socialization into methods of governance such as equity. The second proposition has two parts: “groups can maximize collective reward by evolving accepted systems for ‘equitably’ apportioning rewards and costs among members,” and “groups will generally reward members who treat others equitably and generally punish (increase the costs for) members who treat others inequitably” (Walster et al., 1973, p. 151). Following with this idea, an inequitable relationship violates two social norms: that maintaining equity is the way to receive rewards from the group and that creating inequity will induce punishment from the group. This indicates that discomfort due to inequity in relationships is socialized into humans. Following with the idea of discomfort due to inequity, proposition three states, “when individuals find themselves participating in inequitable relationships, they become distressed. The more inequitable the relationship, the more distress individuals feel” (Walster et al., 1973, p. 153). Perception of equitable treatment in a relationship should lead to satisfaction and positive assessments of the relationship; perception of inequitable treatment should lead to negative assessments and anger (Hegtvedt, 1990). Equity in a relationship can be disturbed in one of two ways: overreward and underreward, and distress may result from each. Overreward distress is expected to be experienced as guilt (Walster et al., 1973), and underreward is generally expected to be experienced as anger (Hegtvedt, 1990). Burgess and Nielsen (1974) explain the relationship between distress and equity by suggesting that it is curvilinear. In other words, low or high relative outcomes (inequity) can cause distress, while equal relative outcomes will not cause distress. Leventhal et a1. (1969) also showed that the more inequity experienced in a task relationship, the more distress participants feel. When inequity exists in a social relationship, the participants are likely to feel less satisfaction and more distress whether they are being overrewarded or underrewarded (Sprecher, 1992). Finally, proposition four notes, “individuals who discover they are in an inequitable relationship attempt to eliminate their distress by restoring equity. The greater the inequity that exists, the more distress they feel, and the harder they try to restore equity” (Walster et a1. 1973, p. 153). In a study using vignettes to evoke inequitable feelings in subjects, Sprecher (1992) found that subjects reported that in an underrewarding situation set up in a vignette, they would be more likely to engage in equity-restoring actions or leave the relationship rather than do nothing about restoring equity. In an overbenefiting situation, they would be more likely to attempt to restore equity than to do nothing. However, in the overbenefiting situation, subjects indicated that they would be more likely to do nothing than to leave the relationship. This indicates that overrewarded participants may be likely to remain in a relationship they know to be inequitable, and will probably attempt to change their level of attraction rather than attempt to gain compliance from their partner in changing his/her actions. One might ask why people would do something in response to inequity in their relationship, rather than simply living with the relationship as is. This question can be answered by cognitive dissonance theory (Festinger, 1957). The idea behind cognitive dissonance is that people do not relish living with conflicting ideas and reality in their lives. In other words, it is important for humans to have consistency between different ideas in their own minds and also between their ideas and their behaviors. Cognitive dissonance exists in relation to inequity because of the norms set up by society. As previously mentioned, feelings of distress because of inequity occur in humans because the inequity violates prevalent norms in our society. These societal norms suggest that it is undesirable to take advantage of others or to be so weak that others can take advantage of you. If the norms are violated, it is likely that consequences such as not receiving rewards or incurring punishments will result. Because of these norms, having inequity in our relationships can cause dissonance between our behaviors in the relationship and the socially embedded norms that tell us inequity is undesirable. According to equity theory, in response to this dissonance, people will either start to change their behaviors or their mindsets regarding the inequitable relationship. It is hypothesized that peOple will take one of four courses of action in order to eliminate the distress caused by an inequitable relationship. They may attempt to restore actual equity or psychological equity (Walster et al., 1973), terminate the relationship, or do nothing (Sprecher 1992). First, actual equity can be restored if one changes either one’s own or one’s partner’s outcomes or inputs. Recall the example of Heidi and Hannah. Heidi was doing most of the work in the fi'iendship, while Hannah enjoyed being friends, but did not put in the same effort as Heidi. In this case, Heidi could restore actual equity in two ways: by stopping her actions to benefit Hannah, such as no longer bringing her lunch or reducing the number of times she calls Hannah, or by asking Hannah to change her actions by doing something like bringing Heidi lunch or calling her more often. There are numerous ways to resolve inequity, particularly in a multidimensional friendship; these are just examples of possible ways Heidi could restore equity. By the same token, Hannah, feeling overbenefited, could also attempt to restore actual equity by changing her actions or asking Heidi to change hers. Second, psychological equity can be restored if one rationalizes the inequity in the relationship until one no longer perceives inequity. One does this by distorting one’s perception of one’s own or one’s partner’s outcomes or inputs (Walster et al., 1976). This can be done in two ways: by minimizing or exaggerating the inputs or outputs of one or the other member of the relationship. Given a new situation in which Hannah is married and pregnant and Heidi is single, Heidi might attempt to restore psychological equity by exaggerating her own inputs. She might think to herself that without her help, Hannah could not manage giving birth and taking care of a baby. By the same token, Hannah might minimize Heidi’s inputs by thinking to herself that Heidi is single and she needs Hannah more than Hannah needs her; hence it is okay that Heidi pays so much attention to Hannah. A third option is for a person to terminate the relationship altogether, thus eliminating the source of inequity (Sprecher, 1992). This may be difficult to achieve, particularly in a blended friendship, because of the built-in nature of the relationship. However, it is still within Heidi or Hannah’s options to end at least one dimension of their relationship, and perhaps both. Either of them could decide to end their friendship outside of work, and if it was an extreme case, Heidi or Hannah could ask to be transferred to another division or simply move to a different location in their workplace to avoid unwanted interaction. Finally, the person may do nothing to eliminate the inequity (Sprecher, 1992). Sprecher simply asked participants if they would “do nothing” in response to a hypothetical inequitable relationship she set up. In some instances, this could be equated to psychological restoration because of the necessary mental adjustment required to do nothing when disadvantaged or advantaged in a relationship. Returning once again to the previous example, Heidi, being underrewarded, might think that Hannah could not manage giving birth and taking care of a baby without Heidi’s help. This rationalization would allow Heidi to “do nothing,” or make no changes in the relationship. Hannah could also avoid making changes in the relationship by thinking to herself that Heidi is single and she needs interaction with Hannah and her family more than Hannah needs her. The response of various individuals to inequity in different dimensions of their friendships could be useful to organizations in understanding how blended friendships in their organizations may work. For instance, organizations might like to know if their employees rationalize inequity in the task dimension of their blended friendships rather than taking action to change the inequity. The way in which individuals will attempt to restore equity in different dimensions is the main focus of this study. For the purposes of this study, equity will be studied in two arenas: task and social. These correspond with the two dimensions of the blended friendships to be examined here. Task equity refers to the equity that exists in the work relationship of blended friends. For instance, in an advertising firm, does one friend come up with all the creative ideas while the other friend does the finances for a given account or does one friend do all the work, creative and financial, while the other rides on his/her coattails? The latter example would be task inequity. Social equity refers to the equity that exists in the social dimension of the blended friendship. For example, do the two fiiends take turns planning weekend activities or does one friend do all the planning all the time? Again, the latter example is social inequity. It is important to note that this study will focus specifically on blended friends in equivalent positions in the workplace and exclude supervisor-subordinate friendships. Both task and social equity and inequity will be discussed in detail including a survey of studies done in each realm. Equity theory: Task dimension There have been numerous studies on the effects of inequity in task situations (e. g. Leventhal & Anderson, 1970; Leventhal & Bergman, 1969; Morgan & Sawyer, 1979; Ronen, 1986). These studies have involved everything from an examination of how children allocate rewards when completing a task with an imaginary partner (Leventhal & Anderson, 1970) to a study of organizational members’ choices of whose rewards they compare their own to, whether those “comparison others” be inside or outside the company (Ronen, 1986). These studies have encompassed a range of support and dissent for the predictive ability of equity theory. The idea that people will respond differently to overbenefit and underbenefit in a relationship has been explored by Sprecher (1992) and is further supported by Leventhal and Anderson (1970). In Sprecher’s study, subjects who read a vignette depicting them in an underbenefiting relationship said they would take some kind of action, whether restoring actual or psychological equity or terminating the relationship, rather than doing nothing. Overbenefited subjects considered themselves likely to engage in behaviors to restore actual or psychological equity. However, these subjects also said they were more likely to do nothing than leave the relationship. Leventhal and Anderson (1970) found that both male and female kindergarteners, when presented with evidence that their performance was inferior to their partner’s, took half the reward and psychologically distorted their performance. Psychological distortion was measured by self-report of performance. These children, who understood their performance to be inferior to a fictitious partner’s, greatly underestimated the difference between their performance and the partner’s performance. However, male children, when presented with evidence to their superior performance, took most of the reward and reported their performance accurately. Female children who were superior took no more than those who performed equally and accurately reported their and their partner’s performances. Although this study used kindergarteners as subjects and the current study is aimed at results which will be generalizable to working adults, these results illustrate just how early societal norms come into play, suggesting that even at the age of five, children have already been socialized to be uncomfortable with inequity. Clearly if the equity norm is embedded in children this early, it is even more strongly indicated that equity norms will govern adult relationships. In further study of responses to underbenefit task-related situations, Leventhal and Bergman (1969) hypothesized that extremely underrewarded individuals would decrease their rewards with self-depriving behavior. Subjects in the study performed a task jointly with a confederate, and each member of the team (confederate and subject) put in about the same amount of effort. Extremely underrewarded subjects received 5¢ of $1.40 earned by the team and could only increase or decrease their portion in increments of up to Sgt. Since the subjects had little power to increase their already-too-small rewards, the only way to restore power equity in the relationship was to deprive themselves of the small reward allotted. Leventhal and Bergman stated, “Person can increase his power through self-depriving behavior because such behavior demonstrates that the incentives which Other controls have little value for Person and, consequently, that Other cannot influence Person by administration of rewards and punishments” (p. 155). At moderate 10 levels of underreward, subjects who received 40¢ of the $1.40 would invariably increase their reward. More subjects increased their reward rather than self-depriving in either extreme or moderate underreward situations. These results indicate that individuals will act to change inequity and that they have differing responses to inequity in varying situations. Self-presentation may have an effect on how individuals in relationships choose to respond to inequity. Shapiro (1975) did a study where subjects performed a task with a partner and were then allowed to distribute the rewards as they chose. The distribution was a test to find if subjects who were either primed to expect future interaction or no future interaction with their partner would choose equity or equality as a governing principle. Shapiro found that for subjects who contributed more to the original task (and knew it), the choice between equity and equality was governed by whether or not they expected to have future interaction with their partner. This supports the idea that people choose their exchange responses carefully depending on the situation they are in; perhaps the response to inequity in a task relationship may depend on the need for self- presentation in the workplace. An interesting viewpoint of organizations holds that employees are not only governed by the standard of equity literal monetary pay, but also by the equity of other types of pay, specifically environmental pay, or the status conveyed by office space. Greenberg (1988) studied the effects of this phenomenon by changing employees’ office space. Some employees were moved to higher-status offices while others were moved to lower status office. Greenberg hypothesized that these employees would change their outputs proportionally to the status of their environment. That is, employees moved to 11 higher status offices would increase their outputs and employees moved to lower status offices would decrease their outputs, both to make up for the inequity created by the change of office. This contention was supported in Greenberg’s study, implying another facet in which blended friends could find inequity in their task relationship. Although it is unlikely that the friends set up in this study would have disparate office status, it is possible that actual blended fiiends could encounter this barrier to equity in their relationships. Equity theory was used to predict Major League Baseball (MLB) players’ propensity to initiate final-offer arbitration to determine their salaries. In this study, F izel, Krautmann, and Hadley (2002) examined MLB because of the public nature of MLB players’ salaries. They reasoned that because the players have knowledge of what others are being paid, they will easily develop opinions about whether their own salaries are inequitable and that these opinions would be predictors of their propensity to initiate arbitration. Specifically, a player who perceived his salary to be highly inequitable in comparison to other players’ salaries would be more likely to initiate arbitration. This propensity to arbitrate would develop under the conditions of low trust for the organization which was the source of the salary inequity and a belief that arbitration would force the organization to bargain fairly for the player’s salary. F izel et al. did indeed find that the perception of inequity was an indicator of the likelihood to file for arbitration. This indicates that in a task setting—baseball being the work at hand and salaries connected to baseball—individuals do adhere to the principles of equity theory by attempting to address inequity in their relationship with an organization. 12 Another instance in which an equity norm was studied in a sports team setting is in a study by Rees and Wechsler Segal (1984), who hypothesized that perceptions of equity would mediate the relationship between competition and attraction such that two competitors might, contrary to intuition, have high attraction for each other if they perceived their rewards to be equitable. Rees and Wechsler Segal studied this idea using two football teams. They counted as competitors any players who were listed as playing the same position, players who were reported by others to play the same position, and players who other players perceived as a threat to achieving their “individual goal for achievement level”. They measured equity levels by ranking all players for a given position and also by asking each player to rate the performance level of his competitors in relation to his performance level. This yielded a measure of equity such that if a player’s team ranking was higher than the respondent’s rating of his performance, perceived inequity existed. If the team ranking and respondent’s rating were similar, then equity was perceived. Interpersonal attraction between competitors was also measured. On one team, most players perceived equity between themselves and their competitors (eight of sixty-five perceived inequity). On the other team, a higher number of players perceived inequity between themselves and their competitors (seventeen of forty-four). There was no effect of perceptions of equity on attraction for competitors in the first team, but in the second team, competitors perceived to be inequitably ahead of the respondents were rated significantly lower on attraction than competitors who were perceived to be equitably ahead of the respondents. This is an interesting finding because of the inherent nature of competition. It seems obvious that competition leads to negative feelings toward the competitor. However, in this setting, it is shown that if equity is 13 perceived between two players, attraction between them may remain high, and if inequity is perceived, then attraction will be low. How does a football study translate into a discussion of blended friendship? This study illustrates an instrumental difference between the task and social dimensions of blended friendship. It seems clear that two people who work for the same company in essentially the same job will compete at some point. They may compete for a certain project or even for a promotion; either way, competition is indicated in a work setting. If these people happen to be blended friends, it is also clear that there is a social dimension to their friendship in which competition is neither necessary nor imminent. Coworkers who are part of a blended fiiendship may have difficulty dealing with inequity, particularly in the task dimension, and remaining friends. However, they may more readily weather inequity in the social dimension where competition does not inherently exist. Not many studies have attempted to examine inequity in task or social dimensions where both dimensions were present in a relationship. It is our contention that there will be a difference in how subjects will respond to inequity depending on which dimension of their blended friendship it is manifested in. Along these lines, Morgan and Sawyer (1979) studied the amount of competition in dyadic interactions between strangers and friends. They found that strangers increased competition when they perceived inequity, while friends who perceived inequity did not. If we align the “stranger” in Morgan and Sawyer’s study with our conception of a task dimension, (a less emotional, more businesslike relationship) and Morgan and Sawyer’s “friend” with our social dimension (a more emotional, personal relationship), some light can be shed on responses to 14 inequity in either dimension. This supports the idea that if blended friends have separation between the task and social dimensions of their friendships, they may respond differently given that the inequity is in one dimension rather than the other. For instance, if inequity exists in the task dimension, blended friends may be more likely to attempt to restore actual equity while if inequity exists in the social dimension, blended friends may be more likely to restore psychological equity in an effort to retain the friendship. Equity theory: Social dimension Mixed support exists for the idea that equity theory is applicable to social relationships. However, there are a number of researchers (Hatfield, Traupmann, Sprecher, Utne, & Hay, 1985; Hatfield, Utne, & Traupmann, 1979; Walster, Walster, & Berscheid, 1978) who believe that equity theory does hold predictive power in social relationships as well as relationships that can be measured through economic means, such as task relationships. A number of researchers have done tests of equity theory in social settings which indicate that equity theory can be used to successfully predict outcomes in social relationships. For instance, Davidson, Balswick, and Halverson (1983) tested the relationship between affective self-disclosure and marital adjustment and found support for the predictive ability of equity theory, which predicted that couples who had inequitable relationships in terms of their self-disclosure would experience distress, and hence have difficulty adjusting to marriage. Davidson et a1. did find that in the perception of individuals, there was lower marital adjustment when the perceiver was under- or overrewarded in self-disclosure from their partner. 15 Vogl-Bauer, Kalbfleisch, and Beatty (1998), in a study of parent-adolescent relationships, found support for equity theory in the parents’ reports of the relationship. There was a statistically significant difference between the reports of satisfaction from parents who perceived themselves to be underbenefited and those who thought they were in an equitable relationship with their adolescent child. There was no difference between overbenefited parents and equitable parents, however. The adolescents’ reports also supported equity theory: adolescents were most satisfied with overbenefited relationships and least satisfied with underbenefited relationships. Buunk and Bakker (1997) also tested equity theory in a study examining subjects’ responses to their partner’s unprotected extradyadic sex. They hypothesized that subjects would respond strongly to a violation of the norm of reciprocity by their partners. This prediction proved to be true and supported equity theory because in each case, the norm of reciprocity was violated by the unprotected extradyadic sex, hence upsetting the balance of equity. The subjects’ responses after reading a vignette about their partner’s unprotected extradyadic sex, to items such as “I would, despite everything, keep loving my partner” and “I would not want to see my partner for a while” were, in effect, ways of rebalancing equity in the relationship. In a dissenting study, Michaels, Edwards, and Acock (1984) studied the effects of inequities on satisfaction in intimate relationships. They found that inequity was not a strong predictor of satisfaction. Michaels et al. used a formula to measure the balance of the relationship, and also measured participants’ outcomes and their outcomes relative to their expected outcomes or comparison level. The better predictors of relational satisfaction were outcome level and outcome relative to expectation or comparison level, 16 indicating the possibility that inequity within a relationship is not so important in social relationships and inequity with outside comparisons may be more important. In a similar but task-oriented study, Ronen (1986) found that employees’ satisfaction with pay was affected more when using outside comparisons than when using comparisons to other employees in the same organization. Equity theory will make different predictions depending on the type of relationship under the hypothetical microscope. Winn, Crawford, and Fischer (1991) studied both dating relationships and same-sex friendships. They looked at both equity and commitment in both types of relationship, questioning if fiiends would be more tolerant of inequity than romantic partners. This comparison was at the behest of Wright (1985) and Davis ( 1985), who both suggested that there may be differences between the way friends and romantic partners experience equity and inequity. Winn et al. found that individuals in inequitable friendships were less distressed than those in inequitable dating relationships, underlining the difference between those types of relationship. In a similar contrasting study, Kollock, Blumstein, and Schwartz (1994) found differences between the ways that cohabiting versus married couples perceived equity in their relationships. In studying actual couples, they used a global measure of equity to assess the members’ perceptions of equity in their relationships, while also gathering data about the members’ actual inputs and outcomes in terms of things like income or hours spent on housework. Though there were varying results across various dependent variables, the overall importance of studying the type of relationship was underlined by the specific differences between cohabiting and married couples. l7 These studies do not indicate that there will be a similar difference between coworker and outside of work relationships, particularly in multidimensional relationships containing both types. However, they do provide support for the possibility that there will be differences in these types of relationships when inequity is present. Further, they suggest the value of studying the differences between multiple types of relationship. Friendship in the workplace People who work together often develop friendships based on organizational roles. Friendships which develop because of similar organizational roles or propinquity may develop further into relationships known as blended friendships (Bridge & Baxter, 1992). Blended friendships are fiiendships that combine interaction in the workplace with interaction outside the workplace (Bridge & Baxter, 1992). Bochner (1984) notes a dichotomy of role and personal bonds in social relationships. Role and personal bonds can be likened to task and social dimensions, respectively. Blended friendship involves not only a work dimension, but also an outside dimension; the two could be categorized into task and social dimensions. Sias and Cahill (1998) found when they asked respondents to identify the shift from close friend to “almost best” fiiend in the workplace, respondents indicated that the change included extra-organizational socializing, experiencing life events during the friendship, and work-related problems. It is wise to note the dichotomy of task versus social dimensions, which both exist in blended friendships. Katz and Kahn (1978) dichotomize these by noting that production, or task accomplishment, is the goal in the task dimension of a relationship, while keeping up the social network of an organization, or social maintenance, is the goal 18 in the social dimension of a multidimensional relationship. The task dimension demands efficiency, while the social dimension values cohesiveness. The task dimension, also known as the technical or production-orientation dimension, evokes a more economic slant, such that equity theory is often applied in a task-oriented setting. Because using equity theory in predicting outcomes of social dimensions is less intuitive, it is a unique and hence valuable lens for interpretation. The existence of these two dimensions means that not only do blended friends have the potential for inequity, they have the potential for inequity in both the task and relational dimensions of their friendships. Bridge and Baxter (1992) note that tensions can develop as a result of this type of multidimensional relationship. Friends whose blended relationship grew from a working relationship may have a difficult time knowing which dimension of the relationship to enact at different times. They may also have difficulty knowing where to resolve inequity if it becomes apparent in either dimension of their fiiendship. Organizations want satisfied workers; since their workers are no doubt in blended friendships due to proximity, knowing more about these how these types of friendships relate to worker satisfaction could be quite helpful. Winstead, Derlega, and Montgomery (1995) did a study on friendship at work and job satisfaction which indicated that quality of friendships at work and job satisfaction are related. Interestingly, Winstead et al. also found that spending time with friends outside of work was negatively related to job satisfaction. Although having fiiends at work may improve job satisfaction, apparently a blended friendship can actually lower job satisfaction. More information about blended friendships is needed before organizations decide to take actions such as offering help in 19 managing blended friendships, encouraging them, or banning them entirely. Organizations might like to know how individuals deal with the problem of inequity in the social or task dimension of their blended friendship; this could be helpful information for an organization to possess. Rationale Equity is traditionally an economic concept and intuitively seems to associate best with relationships that involve strictly a task dimension, such as work relationships. Despite this, many researchers have used equity theory to make predictions about social and romantic relationships (e. g. Sprecher, 1992; Winstead et. al., 1995; Winn et al., 1991). However, equity has also been used to evaluate things like job satisfaction (e. g. Pritchard, Dunnette, & Jorgenson, 1972) or workplace status (Greenberg, 1988). Equity intuitively associates with the task dimension because it is an impersonal way to operate relationships. It is more comfortable and socially acceptable to seek out an “equitable” relationship when it does not involve friends. The idea of using equity as a rule for evaluating friendships is not desirable to some researchers and probably to most laypeople because one premise of equity theory is that individuals are selfish and self- motivated (Adams, 1965; Roloff, 1981). However, it has been shown by numerous researchers (Davidson et al., 1983; Vogl-Bauer et al., 1998; Buunk and Bakker, 1997; Winn et al., 1991; Kollock et al., 1994) that people do use equity as a governing principle in social relationships, whether it is socially desirable or not. The difference between using exchange orientations such as equity theory in task versus social settings was noted by Kollock et a1. (1994), “The accounting system of friends and lovers is likely to be slower in its response and often not very fine-grained in 20 its attention to detail” (p. 341). Although we do not consciously evaluate our social relationships on the basis of equity theory, we unconsciously use the lens of equity to evaluate our relationships and instinctively respond to violations of our socialized rules, as shown by various researchers (Buunk & Bakker, 1997; Davidson et al., 1983; Vogl- Bauer et al., 1998). When the two dimensions, task and social, are juxtaposed with inequity in one or the other, how the members of the relationship will respond to the inequity may be dictated by the dimension and the social norms attached to it. In discussing the two dimensions, the assumption is made that members of blended friendship possess and employ the ability to separate the two dimensions of their friendships. Why might people respond differently to inequity in different dimensions of their blended friendships? In Morgan and Sawyer’s (1979) study, strangers proceeded differently in response to inequity than friends. If we consider the task dimension to be similar to the “stranger” condition in this study (a less emotional, more businesslike relationship), and the social dimension to be akin to the “friend” condition in the study (a more emotional, personal relationship), then it seems logical that blended friends will respond differently to inequity in the businesslike part of their relationship than they will to inequity in the personal part of their relationship. Two other studies found differences in responses to different relationship dimensions (Kollock et al., 1994; Winn et al., 1991). Winn et al. found that individuals in inequitable friendships were less distressed than those in inequitable dating relationships, while Kollock et al. found various differences in responses between cohabiting couples and manied couples. All these parallels indicate that differences may also exist between the task and social dimensions of relationships. 21 Production is a goal in the task dimension of a fiiendship, while relational or social maintenance is a goal in the social dimension of a blended friendship (Katz & Kahn, 1978). The task dimension of a friendship focuses on productivity, efficiency, and competition, while the social dimension values cohesion, emotional connection, and vulnerability. Competition in a relationship where inequity is present causes attraction to lessen (Rees & Wechsler Segal, 1984). In other words, we like our competitors less if there is inequity. On the other hand, if no competition is present, inequity does not cause a great change in attraction. The fact that competition is neither necessary nor likely between friends, while it is almost unavoidable between coworkers, indicates that people will react differently to inequity in the task dimension than to inequity in the social dimension. It is also postulated that there is a difference between how people will respond to being under- and overrewarded. Hegtvedt (1990) suggested that overrewarded individuals would tend to justify (psychologically resolve) receiving extra rewards in a relationship. In a vignette study similar to this one, she found that underrewarded subjects were the most distressed, followed by equitably treated subjects, and finally overrewarded subjects, who experienced the least distress. This indicates that although distress is indicated in the case of both underreward and overreward, there is a difference in how strongly the distress is felt, and that responses to the two kinds of inequity may very likely be different. Differing responses to over- and underreward were also indicated in Sprecher’s (1992) study, which also used vignettes. Underrewarded subjects wanted to restore equity or leave the relationship, while overrewarded subjects were more likely to restore equity than do nothing. Overrewarded subjects were also more likely to do 22 nothing than leave the relationship, meaning that they would be happy simply receiving extra benefits and not doing anything about it rather than ending an unfair relationship. It has been shown that both social- and task-oriented relationships are often governed by social exchange rules, and by equity in particular. However, there are many different ways for inequity to exist in either type of relationship, and hence many different ways to resolve inequity in these relationships. A prominent and measurable currency in both task and social relationships is attraction, or liking, for the relational partner. This is one currency over which both relational partners have control: either one can at any time decide to increase or decrease their liking for the other. Interpersonal attraction has been measured and used in other equity studies (Clark & Mills, 1979; Rees & Wechsler Segal, 1984). Cognitive dissonance theory (F estinger, 1957) indicates the use of increase or decrease in attraction in the face of inequity. Dissonance exists in a relationship in which one person is overbenefited or underbenefited while the two are friends. Since cognitive dissonance theory suggests that in the face of dissonance, people will make some adjustment to reduce the dissonance, it seems logical that some kind of dissonance reduction will occur in inequitable blended friendships. One way to reduce the dissonance of having an unfair relationship with a friend would be to either increase or decrease liking for them, and this is the method of measurement that will be used. Restoring psychological equity will be operationalized as making changes in interpersonal attraction. Interpersonal attraction occurs when one individual is drawn to another due to a positive evaluation of the other. One way people may restore psychological equity is by changing their level of attraction to their friend. For instance, if Hannah bakes cookies for Heidi three times a week and Heidi does nothing in return, 23 Heidi may increase her liking for Hannah as a way of resolving the inequity in their relationship. By the same token, Hannah may decrease her liking for Heidi as a way of resolving her underrewarded situation. Another way to restore equity is to make an actual change in the relationship. Because of the nature of this study, measuring the actual change is not possible—only the intent to change one’s own behavior or intent to gain compliance from one’s partner in changing his/her behavior can be measured. Based on answers to an open-ended question in our pilot study, when people encounter inequity, they will be likely either to change their own behavior or to try to get their partner to change his/her behavior. Restoring actual equity will be operationalized in two ways: first, as intent to change own behavior, and second as intent to gain compliance from the relational partner, or changing other’s behavior. People in relationships will change their outputs in response to inequity in their relationships (Greenberg, 1988). For instance, Hannah may change her contribution to the relationship by no longer baking cookies for Heidi. If we cannot (or do not want to) change our own behavior, the next logical step would be to try to effect change in the behavior of our partners. Compliance-gaining is a manipulation of others to achieve one’s own goals (Marwell & Schmitt, 1967). Rather than changing her behavior, Hannah may attempt compliance-gaining by asking Heidi if she will run some errands for her in order to remove the inequity created by Hannah’s cookie—baking. Although it seems fairly clear that in order to restore actual equity, people will either change their own behavior or attempt to gain compliance in changing their friend’s behavior, it is not yet clear which of those two will prevail. Because of this, the 24 hypotheses pertaining to behavior change will be duplicated: once for changing own behavior and once for changing other’s behavior. The result is a research question about this particular construct. RQl: Which type of behavior change will be prevalent as a means of restoring equity in inequitable blended fiiendships? Liking is less crucial in the task dimension of the friendship because of its more impersonal and businesslike nature. Because of this, when underreward exists in the task dimension, the perceiver will be more likely to instigate behavior change, either changing his own behavior or attempting to change his partner’s behavior, to restore equity than to decrease his liking for his partner. Because liking may not be a necessity in a task relationship, decreasing liking for a partner would not be an effective means of restoring equity in the task dimension. For example, if two lawyers, Joe and Sam, were good friends and also worked for the same firm, there would be a difference in how Joe responded to inequity in one dimension of their friendship versus the other. If Joe noticed that Sam was asking him to take over half of Sam’s cases at work, Joe would be more likely to refuse to do the work or ask Sam to take his cases back than to simply think, “I certainly don’t like Sam very much anymore.” Because the work relationship is expected to be more economic in nature, it is postulated that asserting an actual change to restore equity will be more prevalent than changing liking in task versus social dimensions. H13: If underreward is perceived in the task dimension of a blended friendship, the perceiver will be more likely to instigate a change in his/her own behavior than if underreward is perceived in the social dimension. 25 Hlb: If underreward is perceived in the task dimension of a blended friendship, the perceiver will be more likely to instigate changing in other’s behavior than if underreward is perceived in the social dimension. In the case of underreward in the social dimension, Joe would be likely to change his level of attraction for Sam, liking him less because he is taking advantage of Joe. Because the social dimension of the blended friendship is not socially expected to be judged by economic means, it is postulated that resolving the inequity will be done in a way that circumvents the natural tendency to distress in the face of inequity. As previously discussed, liking is a currency in social relationships. A social fiiendship is built on liking for each other. Because of this, it seems that lowering or increasing liking will be a more effective method of equity restoration when underreward is perceived in the social dimension. H2: If underreward is perceived in the social dimension of a blended friendship, the perceiver will be more likely to change his level of attraction than if it is perceived in the task dimension. If overreward is perceived in the task dimension of the relationship, the perceiver will be likely to change his level of attraction for the partner. Returning to the example of Joe and Sam, if Joe notices that Sam will do a couple of his cases if Joe puts them on his desk, he will be more likely to start disliking Sam than to change his behavior, which would not be beneficial to him. Regarding the cases, Joe might tell himself, “Sam is such a pushover.” So, in the case of overreward in the task dimension, according to the premise that individuals evaluating relationships using rules such as equity are selfish and 26 self-motivated, it is postulated that the perceiver, to his advantage, will be more likely to change his level of attraction than if the inequity was in the social dimension. H3: If overreward is perceived in the task dimension, the perceiver will be more likely to change his level of attraction than if overreward is perceived in the social dimension. In the case of overreward in the social dimension, it is expected that individuals will feel the distress of inequity, but rather than “taking advantage” of their relational partner, will change their own behavior or attempt gain compliance from their partners in changing their behaviors. For example, if Sam makes dinner for Joe every night, some inequity exists in their relationship. Because a social relationship is built on liking for each other, and Joe is the one receiving extra benefits, Joe would be uncomfortable changing his liking for Sam, so he would choose to resolve it by changing his own behavior or attempting to gain compliance from Sam in changing Sam’s behavior, such as by asking Sam to stop inviting him every night, hence restoring actual equity in the relationship. H43: If overreward is perceived in the social dimension of the relationship, the perceiver will be more likely to instigate a change in his/her own behavior than if overreward is perceived in the task dimension. H4b: If overreward is perceived in the social dimension of the relationship, the perceiver will be more likely to instigate changing in other’s behavior than if overreward is perceived in the task dimension. 27 Method Overview The study consisted of a pilot study followed by a primary study. Participants in the two studies initially read vignettes describing different blended friendships involving inequity in one or the other dimension. Following each vignette, participants reported their responses to items as if they were a member of the friendship described. The pilot study tested vignette realism, and the equity manipulation for effectiveness. This initial study was intended to test the manipulation to see that it was realistic to the participant and perceived in the way intended (i.e. the underrewarded vignette prompted the participant to feel underrewarded). The primary study involved surveying students and employed adults. This part of the study also included the realism and equity manipulation check, but was specifically intended to test the effect of the manipulation on the participants’ choice of method of equity restoration. Pilot Study Participants. One hundred and eighty-five undergraduate students participated in the pilot study. Participants were 38% male and 62% female. The sample consisted of 50.8% Caucasian, 25.9% African-American, 10.3% Asian American, 3.2% Hispanic, 0.5 %, Native American, 0.5% Pacific Islander, 1.1% Mixed, and 7.0% of other ethnicities. Ages of the participants ranged from 17 to 38 years old (M: 21.5, SD= 2.2), with 72.4% of the participants between the ages of 20 and 22. One participant did not enter an ethnicity and one participant neglected to denote his age. Design. The design was a 2 (relationship foundation: coworker vs. friend) x 2 (relationship dimension: inequity in task or social dimension) between subjects x 3 28 (equity level: underreward, equitable, overreward) repeated. This design yielded twelve different vignettes. Each participant received three vignettes. These vignettes presented all three equity levels while keeping the same relationship foundation by relationship dimension cross, so that a participant might receive a Coworker-Task-Underreward, Coworker-Task-Equitable, Coworker-Task-Overreward combination. In turn, another participant might receive a Friend-Social-Underreward, Friend-Social-Equitable, Friend- Social-Overreward combination. The first factor, coworker vs. friend, manipulated participants’ perception of the relationship foundation by changing the first sentence of the vignette. This sentence led the participant to believe either that their relationship had developed through being coworkers who became friends or friends who also happened to be coworkers. In the friend condition, the sentence stated, “You and Sam are friends who work together.” In the coworker condition, the sentence stated, “You and Sam are coworkers who are also friends.” This dimension was added because there might be a disparity in responses to inequity depending on whether the participants believed their loyalties were owed initially to a coworker or friend. The second factor, task versus social, manipulated relationship dimension by placing the inequity in either a task setting, such as projects in which the work was unevenly distributed, or in a social setting, such as unequal sharing of the driving responsibilities for a shared hobby outside of work. In the social condition, the manipulation was as follows: “Outside of work, you and Sam mountain bike together. You often spend weekends biking trails together, and you also do other things together, like going out for a drink after work.” In the task condition, the manipulation stated, “At 29 work, you both keep up with your projects without any trouble. Neither of you has any difficulty completing projects on time. You discuss the projects and share ideas back and forth freely.” The third factor, equity level, was manipulated by either setting up the participant to receive less benefit than Sam, the fictitious fiiend, more benefit than Sam or about the same benefit as Sam. In the underrewarded condition, the vignette contained the following sentences which indicate that the participant receives less benefit than Sam: “Sam often asks you to drive to the trails when biking, even though Sam’s truck is better for transporting your bikes. Not only that, but Sam usually depends on you to bring water and snacks along for the day also.” In the overrewarded condition, a section of the vignette reads, “You often ask Sam to drive to the trails when biking, even though your truck is better for transporting your bikes. Not only that, but you usually depend on Sam to bring water and snacks along for the day also,” indicating that the participant is receiving more benefit than Sam. Because I believed that the order of the vignettes given to participants might affect the responses I received, I counterbalanced the vignettes so that each participant randomly received a different order of the vignettes. For example, in the Coworker-Task- Underreward (CTU), Coworker-Task-Equitable (CTE), Coworker-Task-Overreward (CTO) set of vignettes, participants could receive one of six different orders: CTU, CTE, CTO; CTU, CTO, CTE; CTE, CTO, CTU; CTE, CTU, CTO; CTO, CTE, CTU; CTO, CTU, CTE. The same pattern was followed for each cross of relationship foundation by relationship dimension (Coworker-Task, Coworker-Social, Friend-Task, Friend-Social), 30 yielding twenty-four differently ordered versions of the survey. The different crosses of relationship foundation by dimension were also randomly distributed. A vignette that exemplifies the coworker, task, and equitable conditions is as follows: “You and Sam are coworkers who are also friends. Not only are you friends, but you also work on similar projects and hold identical positions in the same department of the same company. You have been friends ever since you met at work. At work, you both keep up with your projects without any trouble. Neither of you has any difficulty completing projects on time. You discuss the projects and share ideas back and forth freely. You lend each other a hand if needed. You and Sam are supportive of each other, but you both work on your own projects alone.” All the vignette crosses are listed in Appendix A. Procedure. Surveys were distributed during various communication classes in which students received either class credit or extra credit for their participation. They were collected directly following completion by the researcher. Measurement. Participants filled out two scales after reading each of three vignettes, including one each of the underreward, overreward, and equitable conditions, and either the coworker or friend condition and the task or social condition. The Hunter program was used to perform confirmatory factor analysis (CF A) on each scale, two of which consist for each condition: underbenefit, equitable, and overbenefit. The two scales measure realism and fairness, and all items for each scale are listed in Appendix B. Participants rated the items on a Likert-type scale of 1 (strongly disagree) to 7 (strongly agree). Three reliabilities were calculated for each scale, one for each equity condition: underbenefit, equitable, and overbenefit. 31 Realism. To test the realism of the vignette, participants indicated the degree to which they agree with each of the following four items “A friendship like this could develop in real life,” “It is possible to have a coworker like Sam,” “You can imagine being in a friendship like this one,” and “This type of relationship with Sam in the story can happen in real life.” None of the items were recoded. Hunter’s C onfirmatory Factor Analysis (CFA) program was used to perform a test for internal consistency, which involves computing a predicted correlation for each combination of items based on factor loadings, and then calculating deviations between predicted and observed correlations. When performed on the equitable realism scale, the test yielded no significant deviations, indicating data consistent with a unidimensional factor. The reliability (Cronbach’s or) for the equitable scale was .91. For the underbenefit realism scale Hunter’s test again showed no significant deviations and the reliability was .83. Finally, for the overbenefit realism scale, the test again showed no significant deviations and the reliability was .80. Based on the CFA performed, all four items were included in the analysis. Fairness. Fairness was measured using participants’ indication of the degree of their agreement with seven items, which stated, “Sam is taking advantage of you,” “You are getting more benefits from the relationship,” “This friendship seems fair,” “This work relationship seems fair,” “You and Sam both get the similar amount of benefits from the relationship,” “You are taking advantage of Sam,” and “Sam is getting more benefits from the relationship.” Again, according to the CF A performed, all items were included in the analysis. For the equitable fairness scale, Hunter’s unidimensionality test showed no significant deviations. For the underbenefit fairness scale, Hunter’s test again showed 32 no significant deviations. Finally, for the overbenefit fairness scale, Hunter’s test also showed no significant deviations. Because of the nature of the conditions (underbenefit, equitable, overbenefit), the items were recoded differently in each condition. In the underbenefit condition, the following items were recoded: “Sam is taking advantage of you” and “Sam is getting more benefits from the relationship.” On this scale in general, the higher the score given, the more fair the participant believed the vignette to be. In the underbenefit condition, however, a higher score on these two items would indicate that the relationship was more unfair. Hence, those two items were recoded. In the equitable condition, four items were 9, recoded: “Sam is taking advantage of you, “You are getting more benefits from the relationship,” “You are taking advantage of Sam,” and “Sam is getting more benefits from the relationship.” Again, higher scores on these items indicate that the relationship is more unfair, necessitating the recoding. Finally, in the overbenefit condition, two items were recoded: “You are getting more benefits from the relationship” and “You are taking advantage of Sam.” In the overbenefit condition, higher scores on these items would indicate more unfairness. The reliability for the underbenefit scale was .81. For the equitable scale, the reliability was .90, and for the overbenefit scale, the reliability was .75. Pilot Study Results Realism. Realism of the vignettes was tested with a one-sample t-test. The comparison value was 4, or the midpoint of the scale. The test of the equitable vignette (M=6.21, SD= 0.89) was significant, t (180) = 33.33,p < .01, 172: .86, indicating that the vignette was very realistic. The test of the underbenefit vignette (M=5.70, SD=1.07) was 33 also significant, t (183) = 21.48, p < .01, 772: .72, indicating this vignette was also realistic. Finally, the overbenefit vignette (M=5.63, SD=1.04) was also significant, t (179) = 20.98, p < .01 , 172: .71 , indicating this vignette was also quite realistic. Realism was also analyzed in a relationship foundation (coworker vs. friend) X relationship dimension (task vs. social) X equitable status (underbenefit vs. equitable vs. overbenefit) mixed factorial analysis of variance (ANOVA). ANOVA was used to see if there were significant differences among the participants’ perception of realism for the three vignettes. Relationship foundation and relationship dimension were between subjects factors and equitable status was a repeated measure. Cell means and standard deviations are listed in Table 1. The ANOVA yielded a significant difference for equity level, F (2, 352) = 32.82, p < .01, n2 = .01 between the realism of the three vignettes (underbenefit, equitable, overbenefit). There was a significant difference in how realistic each vignette seemed depending on the equity level portrayed. The ANOVA for relationship foundation (fiiend M = 5.83, coworker M = 5.88) was not significant, F ( 1, 176) = .14, p = .71, n2 = .00. The ANOVA for relationship dimension (task M = 5.83, social M = 5.88) was also not significant, F (1, 176) = .20, p = .66, n2 = .00. These two ANOVAs reveal that there was not a significant difference in the realism of the vignettes based on either the type of relationship or the dimension of the relationship. Significant interactions between these factors were not expected and none of the interactions were significant. A 2 (gender: male and female) X 3 (equity level) repeated ANOVA was performed to check the effects of gender on realism. Cell means and standard deviations are reported in Table 2. The effects for gender were not significant, F (l , 178) = 0.76, p = 34 .39, n2 = .00. So, gender of the participant did not make a difference in their perception of the realism of the different equity levels. A 6 [order: underbenefit, overbenefit, equitable (UOE, UEO, EUO, EOU, OUE, OEU)] X 3 (equity level) repeated ANOVA was performed to check for order effects in perception of realism. Cell means and standard deviations are reported in Table 3. The effects for order were not significant, F (5, 174) = 0.92, p = .47, n2 = .00, meaning that the order of the vignettes in the survey did not make a difference in the perception of realism. Fairness. Fairness of the vignettes was also tested with a one-sample t-test for each vignette. The comparison value was again 4, or the midpoint of the scale; higher than four would indicate the vignette was fair, while lower than four would indicate it was unfair. The equitable condition (M=6.02, SD=1.03) showed scores that were significantly higher than the mean, t (180) = 26.5, p < .01, n2 = .80, indicating participants viewed this vignette as fair. The underbenefit condition (M=3.00, SD=1.03) showed scores that were significantly lower than the midpoint, t (183) = -13.18, p < .01, n2 = .49, indicating that this vignette was viewed as unfair. Finally, in the overbenefit condition, (M=3.00, SD= 0.94), scores were again significantly lower than the midpoint, t (179) = -14.29, p < .01, n2 = .53, indicating this vignette was also viewed as unfair. Fairness was also analyzed in a relationship foundation (coworker vs. friend) X relationship dimension (task vs. social) X equitable status (underbenefit vs. equitable vs. overbenefit) mixed factorial analysis of variance (ANOVA) to determine whether significant differences existed among the three vignettes’ fairness. Relationship foundation and relationship dimension were between subjects factors and equitable status 35 was a repeated measure. Cell means and standard deviations are listed in table 4. Mauchly’s test of sphericity was significant, p < .01, indicating that the differences between scores for each participant across three conditions of equity level were not homogeneous, so the Greenhouse-Geisser correction was used. The ANOVA for equity level yielded a significant difference, F (1.25, 220.00) = 569.77, p < .01, n2 = .25 between the realism of the three vignettes (underbenefit, equitable, overbenefit), indicating that participants did perceive a difference between the three equity level manipulations. In other words, the manipulation worked. The ANOVA for relationship foundation (friend M = 4.01, coworker M = 4.01) was not significant for fairness, F (1, 176) = 0.004, p = .95, n2 = .00. Because this test result was insignificant, it seems that the possible difference originally postulated between the two types of relationship does not exist. The AN OVA for relationship dimension (task M = 3.97, social M = 4.05) was also not significant F (1, 176) = 0.79, p = .38, n2 = .00. This means participants did not differ in their perceptions of fairness according to the dimension in which the inequity existed. Again, I was not particularly interested in the interactions between these factors and none of them were significant. A 2 (gender: male and female) X 3 (equitable status: underbenefit, equitable, overbenefit) repeated ANOVA was performed to check the effects of gender on fairness. Cell means and standard deviations are reported in Table 5. The effects for gender were not significant, F (1, 178) = 0.27, p = .61, n2 = .00. So, gender of the participant did not make a difference in participants’ perception of the fairness of the different equity levels. A 6 [order: underbenefit, overbenefit, equitable (UOE, UEO, EUO, EOU, OUE, OEU)] X 3 (equity level) repeated ANOVA was performed to check for order effects in 36 perception of fairness. Cell means and standard deviations are reported in Table 6. The effects for order were not significant, F (5, 174) = 1.91 , p = .10, n2 = .00, meaning that the order of the vignettes in the survey did not make a difference in the perception of fairness. Summary. The manipulation of different levels of equity was successful in all vignettes. All the vignettes were judged by participants to be quite realistic. Further, the equitable vignette was judged to be fair, and the underreward and overreward vignettes were judged to be unfair. However, no difference was found due to the relationship foundation manipulation: there was not a significant difference in either realism or fairness because of the perception of being either fiiends first and coworkers second or coworkers first and friends second. This indicates that it doesn’t make any difference which dimension of a blended friendship developed first; people will have similar perceptions of equity in a relationship regardless. Because of the lack of significant differences, variation of relationship foundation will be omitted from the primary study. Although there were no significant differences between task and social dimensions for realism and fairness, this factor will remain because of our interest in the differences between these dimensions when looking at the dependent variables added in the main study. One final note is necessary in wrapping up the pilot study. The social vignette gave a very specific example of where inequity existed in the mountain biking relationship, down to who was bringing snacks and water. The task vignette, on the other hand, was quite vague in its description of where inequity existed, only mentioning projects that were worked on and ideas that were shared or not shared. This difference became 37 apparent after the pilot study had been undertaken. The task vignettes will be modified for the main study to reflect detail similar to the social vignettes. Since the relationship foundation factor will be removed, all vignettes will be modified to reflect that change. Also, it was determined that since Sam’s gender is supposed to be neutral or decided by the participant, it should be a more gender-neutral name. The name was changed to Chris instead of Sam. The new vignettes will be included in Appendix A with all the original pilot study vignettes. Primary Study Participants. Two hundred and eighty-eight undergraduate students and twenty- nine employed adults participated. Students were recruited through various university classes, while employed adults were a snowball sample. Of the entire 317 participants, 38.8 % were male, 59 % were female, and 2.2 % did not indicate their sex. The sample consisted of 75.7 % Caucasian, 7.6 % African-American, 3.8 % Asian American, 2.8 % Hispanic, 0.9 % Pacific Islander, 0.3 %, Native American, 2.2 % Mixed, and 2.2 % of other ethnicities. Ages of the participants ranged from 18 to 52 years old (M= 23.24, SD= 5.92), with 60 % of the participants between the ages of 20 and 22. Eight participants did not enter an ethnicity and eight participants did not denote their ages. Two hundred and eighty five participants indicated the type of job they currently hold. Of those participants, 11.9 % hold administrative/support jobs, 11.6 % hold sales jobs, 9.5 % hold public service jobs, 7.0 % hold hourly/skilled jobs, 6.7 % hold management jobs, 4.2 % hold human resource jobs, 1.1 % hold healthcare jobs, and 0.4 % hold technology jobs. 25.3 % indicated they had jobs that did not fit into any of the previous categories and 19.3 % indicated that their jobs fit into multiple categories. The other thirty-two 38 participants did not indicate a type of job. Of 227 participants who entered their hours per week in their current job, 30.3 %, or 70 participants, indicated they work forty or more hours per week, while 68.8 % (including those above 40 hours) indicated working twenty or more hours per week currently. When asked if they have a person who they identify as both a friend and coworker, in other words, a blended fiiend, 79.5 % of the participants indicated that they did. Based on the combination of participants’ gender and the participants’ guesses at Chris’ gender, it was determined that 50 % of the friendships in the study were cross-sex and 50 % were same-sex friendships. More specifically, 31 % of the sample consisted of male participants who believed Chris also to be male, 9 % of the sample was male participants who believed Chris to be female, 19 % of the sample was female participants who believed Chris also to be female, and 42 % of the sample was female participants who believed Chris to be male. Design. The design of this study is 2 (relationship dimension: task and social) X 2 (equity level: underbenefit, overbenefit). Since the hypotheses deal with the differences between underrewarded and overrewarded situations, the equitable condition was included as a control to which either the underrewarded or overrewarded conditions are compared. Because the aim was to measure change, each participant first received an equitable vignette followed by either an underbenefited or overbenefited vignette. This enabled me to establish a baseline for responses in an equitable relationship and compare that to responses in an inequitable relationship. This means a participant may have received Task-Equitable, Task-Underbenefit or Task-Equitable, Task Overbenefit. By the 39 same token, a participant may also have received Social-Equitable, Social-Underbenefit or Social-Equitable, Social-Overbenefit. The vignettes involved the same two main characters, but with varying situations and states of equity or inequity. A generic introduction has replaced the coworker/ fiiend specific introduction. This introduction simply introduces the blended friendship, “You and Chris are both friends and coworkers. You work in identical positions in the same department of the same company. Not only do you work together, you are also friends outside of the workplace.” The name of the fictional friend has been changed to the more gender- neutral “Chris.” The social condition of all three (underreward, equitable, overreward) vignettes was also been modified to make the activity engaged in outside of work more gender- neutral. The new social vignette contained almost the same manipulation, except with the substitution of rollerblading for mountain biking. For instance, in the social equitable condition, the vignette reads as follows, “Outside of work, you and Chris rollerblade together. You often spend time on the weekends rollerblading just for fun and you also do other things together, like going out for a drink after work. You take turns picking each other up to go to the park, which is halfway between your houses. By the same token, you both take turns bringing water and snacks for after your workout.” In the underrewarded condition, the vignette reads as follows, “Outside of work, you and Chris rollerblade together. You often spend time on the weekends rollerblading just for fun and you also do other things together, like going out for a drink after work. Chris often asks you to drive and go to a park that is near Chris’s house, even though it’s quite far out of your way and you have to drive through a lot of traffic to get there. Not only that, but Chris usually 40 depends on you to bring water and snacks for after the workout.” In the overrewarded condition, the vignette reads as follows, “Outside of work, you and Chris rollerblade together. You often spend time on the weekends rollerblading just for fun and you also do other things together, like going out for a drink after work. You often ask Chris to pick you up and drive to a park that is closer to your house, even though it’s quite far out of Chris’s way and Chris has to drive through a lot of traffic to get there. Not only that, but you usually depend on Chris to bring water and snacks for after your workout.” In the task conditions, the participant shares a task with a fictitious fiiend, Chris. The equitable condition is set up as follows, “At work, you and Chris must jointly come up with a monthly budget report for your company’s supply closet. You are supposed to do an inventory of everything in the supply closet and Chris is supposed to do the math to f1 gure out what needs to be replaced and how much the replacements will cost. This arrangement seems to work out fine. You do your part and Chris does Chris’s part.” In the underrewarded task condition, the task is set up as follows, “At work, you and Chris must jointly come up with a monthly budget report for your company’s supply closet. You are supposed to do an inventory of everything in the supply closet and Chris is supposed to do the math to figure out what needs to be replaced and how much the replacements will cost. However, sometimes Chris can’t seem to get his part done and you end up doing the entire report. You never complain about helping Chris, but simply do the report and move on.” In the overrewarded task condition, the task is set up as follows, “At work, you and Chris must jointly come up with a monthly budget report for y our company’s supply closet. You are supposed to do an inventory of everything in the Supply closet and Chris is supposed to do the math to figure out what needs to be 41 replaced and how much the replacements will cost. However, sometimes you can’t seem to get your part done and Chris ends up doing the entire report. Chris never complains about helping you, but simply does the report and moves on.” Procedure. Surveys were distributed to undergraduate students during varied communication classes in which students volunteered or received class credit or extra credit for their participation. The surveys were collected directly following completion by the researcher. The employed adult surveys were distributed by various means, including directly by the researcher with self-addressed, stamped envelopes for return and by snowball sample. Measurement. Each participant received two vignettes, each of which was followed by various scales. The equitable version was followed by scales for realism, fairness, liking, doing nothing, and satisfaction. Because behavior changes are not applicable in an equitable relationship, as there is no need to change anything, the changing other’s behavior and terminating scales were not included with the equitable vignette. The second vignette received, which was either underbenefit or overbenefit, was followed by scales for realism, fairness, liking, changing other’s behavior, doing nothing, termination, and satisfaction. Along with other descriptive information such as gender, type of job held and hours worked weekly, participants were requested to give their guess as to the gender of Chris following the equitable vignette. Although there was no prediction regarding differences due to friendship makeup (same-sex or cross-sex), this allows for tests of possible effects of friendship makeup. Realism and fairness were measured in the same manner as in the pilot study. A few items were re-worded to remove awkwardness. For instance, “This type of 42 relationship with Chris in the story can happen in real life,” was changed to “The relationship with Chris could happen in real life.” Also, for continuity across all items, the wording was changed to replace “You” with “I,” reasoning being that if I am asking the subjects to imagine that they are in the relationship described, it makes more sense for the items to be about “I” rather than “You.” For example, “Chris is taking advantage of you” was changed to “Chris is taking advantage of me.” All scales were a Likert-type format ranging from 1 (strongly disagree) to 7 (strongly agree). Realism. Two realism scales were administered: one for the equitable vignette and one for the inequitable vignette. Each scale included the same four items, which can be found in Appendix C. None of the items were recoded. Hunter’s Confirmatory Factor Analysis (CFA) program was used to perform a test for internal consistency, which involves computing a predicted correlation for each combination of items based on factor loadings, and then calculating deviations between predicted and observed correlations. When performed on the equitable realism scale, the test yielded no significant deviations, indicating data consistent with a unidimensional factor. The reliability (Cronbach’s or) for the equitable realism scale was .91. When CFA was performed on the inequitable realism scale, again the test yielded no significant deviations, indicating data consistent with a unidimensional factor. The reliability (Cronbach’s or) for the inequitable realism scale was .84. Based on the CFA performed, all four items were included in both realism scales for further analysis. Fairness. Fairness was measured with two scales: one for the equitable vignette and one for the inequitable vignette. Both scales were identical and included seven items: “Chris is taking advantage of me,” “I am getting more benefits from the relationship,” 43 “Chris and I both get a similar amount of benefits from the relationship,” “I am taking advantage of Chris,” and “Chris is getting more benefits from the relationship.” For the scale following the equitable vignette, items 1, 2, 4, and 5 were recoded; a high score on this scale indicates fairness; however, high scores on any of these items would indicate unfairness. Therefore, they were recoded so that high scores on these items would indicate fairness. A Hunter’s CFA was performed on this scale and was consistent with unidimensionality of all items. The reliability (Cronbach’s or) of the scale was .81. For the (identical) scale following the inequitable vignette, recoding was more complex because of the nature of the conditions (underbenefit, overbenefit). In the underbenefit condition, the following items were recoded: “Chris is taking advantage of me” and “Chris is getting more benefits from the relationship.” On this scale, the higher the score given, the more fair the participant believed the vignette to be. In the underbenefit condition, however, a higher score on these two items would indicate that the relationship was more unfair. Hence, those two items were recoded. In the overbenefit condition, two items were recoded: “I am getting more benefits from the relationship” and “I am taking advantage of Chris.” In the overbenefit condition, higher scores on these items would indicate more unfairness. A Hunter’s CFA was performed on each scale and both the underbenefit and overbenefit scales were consistent with unidimensionality. The reliability (Cronbach’s or) for the underbenefit scale was .83 and for the overbenefit scale the reliability was .82. Liking. For the purposes of this study, restoring psychological equity was operationalized as making changes in interpersonal attraction. I believe that one way subjects will choose to psychologically restore equity to an inequitable relationship is by 44 adjusting their liking for their fiiend. For example, if Sam and Joe are blended friends, and Sam cooks dinner for Joe every night, Joe may resolve the inequity by thinking to ’9 ! himself, “Boy, Sam is a great friend to have! I really like him a lot By the same token, Sam may resolve the inequity by thinking to himself, “I don’t really like Joe very much since he always comes over for dinner and never offers anything in return.” The scale for liking was included twice in this survey: once following the equitable vignette and once following the inequitable vignette. The scale for liking includes some original items as well as some items from Rubin’s (1970) liking scale. The original items include ones such as “I like Chris a lot.” The items used from Rubin’s scale were chosen for face validity with our items and because they fit best with vignettes. The scale was originally designed to be applied to actual relationships; some of the items were regarding information our subjects do not have about their relationship with Chris. For example, “When I am with _[Chris]___, we are almost always in the same mood.” The items chosen for our study from the scale are as follows, “Chris is the sort of person who I myself would like to be,” “I think that Chris and I are quite similar to each other,” “I would vote for Chris in a class or group election,” and “I would highly recommend Chris for a responsible job.” All items for the liking scales following the equitable and inequitable vignette are listed in Appendix C and designated as either original items or Rubin’s items. Hunter’s CFA was performed on the liking scale for the equitable vignette to check for unidimensionality, and as a result of large deviations between predicted and observed correlations, some items were removed from the analysis. The items removed include items numbered 1, 3, 5, and 9 (see appendix C). With the remaining five items, CFA produced a unidimensional model; hence, these items were used for further 45 analyses. The reliability (Cronbach’s or) of the liking scale for the equitable vignette was .84. Hunter’s CF A also showed large deviations between predicted and observed correlations for the inequitable liking scale, so items 1, 3, 5, and 9 (see appendix C) were also removed from that scale. The remaining five items produced a unidimensional model and were included for use in further analyses. The reliability (Cronbach’s or) of the liking scale for the inequitable vignette was .87. Equity restoration. Restoring actual equity was operationalized as changing one’s own behavior and changing other’s behavior. One way subjects may attempt to restore equity is by changing their own behavior. This could involve increasing or decreasing what they are putting into the relationship. Another way subjects will attempt to restore actual equity will be by attempting to change the behavior of their fiiend. Marwell and Schmitt (1967), when studying differing approaches to changing other’s behavior, found that seven compliance- gaining behaviors loaded to a factor they named “activation of impersonal commitments” which refers to appeals to obligations to behave in a certain way for social reasons. This factor consisted of invoking behaviors through the use of moral appeal, appeals to positive or negative self-feelings, appeals to positive or negative altercasting, and appeals to positive or negative self-esteem. These particular behaviors seem to lend themselves to someone who is appealing to a social norm such as equity theory to effect change in a partner. As a result, a scale was composed for each of these two behavior changes to measure “activation of impersonal commitments” in the participants of this study. Change in own behavior. Items written to measure a subject’s change in his/her own behavior were written to parallel the changing other’s behavior items. These items 46 are as follows: “I would change my actions because morally it is the right thing to do”, “I would change my actions because I know I would feel better about myself afterward”, “I would change my actions because I know I would feel worse about myself if I didn’t”, “I would change my actions because I am not the kind of person who acts the way portrayed 9’ 66 in the story , 1 would change my actions because it is what a good person would do,” “1 would change my actions because people would like me better if I did,” and “I would change my actions because people wouldn’t like me if I didn’t.” Hunter’s CFA was performed on this scale, and large deviations between predicted and observed correlations led to removal of items numbered 2, 3, and 7 (see Appendix C). The remaining four items yield a unidimensional factor structure according to Hunter’s CF A, and hence were included for future analyses. The reliability (Cronbach’s or) of this four-item scale was .82. Changing other ’s behavior. The items to measure changing other’s behavior are as follows: “I would ask Chris to change his actions because morally it is the right thing to do,” “I would ask Chris to change his actions because he will feel better about himself if he does,” “I would tell Chris he will feel worse about himself if he doesn’t change his actions,” “I would tell Chris that because he is such a good person, I know he will change his actions,” “I would tell Chris that any good person would change his actions,” “I would tell Chris that people would like him better if he changed his actions,” and “I would tell Chris that if he doesn’t change his actions people will like him less.” Hunter’s CFA was also performed on this scale, and large deviations between predicted and observed correlations led to removal of items numbered 2, 6, and 7 (see Appendix C). The remaining four items yield a unidimensional factor structure according to Hunter’s 47 CF A, and hence were included for future analyses. The reliability (Cronbach’s or) of this four-item scale was .86. Manipulation check. A manipulation check was performed on the scales for realism and fairness to test consistency between the pilot study and the primary study. To test for a possible student/employed adult disparity, a 2 (type: student/employed adult) X 2 (equitable status: over/underreward) X 2 (relationship dimension: task/social) ANOVA was performed on realism and fairness. No significant differences were found. Also, to test for a difference in variables due to friendship makeup, a 2 (Chris’ gender: male/female) X 2 (participant gender: male/ female) X 2 (equitable status: over/underreward) X 2 (relationship dimension: task/social) ANOVA was performed on realism and fairness. Again, no significant differences were found. Realism. Realism of the vignettes was tested with a one-sample t-test. The comparison value was 4, or the midpoint of the scale. The test of the scale following the equitable vignette (M=6.08, SD= 0.98) was significant, t (314) = 37.47, p < .01, n2 = .82, indicating that the vignette was very realistic. The test for the scale following the inequitable vignette (M=5.29, SD=1.19) was also significant, t (312) = 19.20, p < .01, n2 = .54, indicating this vignette was also realistic. The realism of the two vignettes was compared using a paired-samples t-test. Although both scales showed that participants believed both scales to be quite realistic, the scale following the equitable vignette was significantly more realistic (M= 6.08, SD= 0.99) than the scale following the inequitable vignette (M= 5.29, SD= 1.19), t (312) = 12.58,p< .01, 122 = .34. 48 A 2 (participant gender: male/female) X 2 (equity: over/underreward) X 2 (relationship dimension: task/social) ANOVA was performed to check for gender differences in realism. There was no difference in realism for the scale following the equitable vignette, but there was a significant effect for gender differences in realism following the inequitable scale, F (l, 301) = 9.89,p < .01, 172 = .03, such that women (M = 5.47, SD = 1.48) perceived the inequitable vignette as more realistic than men (M = 5.05, SD = 1.26). Fairness. Fairness of the vignettes was also tested with a one-sample t-test for each vignette. The comparison value was again 4, or the midpoint of the scale; higher than four would indicate the vignette was fair, while lower than four would indicate it was unfair. Following the equitable vignette (M = 5.77, SD = 0.97), the scale showed scores that were significantly higher than the mean, t (313) = 32.18, p < .01, n2 = .77, indicating participants viewed this vignette as fair. Following the inequitable vignette, the underbenefit scale (M = 2.34, SD = 0.91) was significantly lower than the midpoint, t (158) = -22.98,p < .01, n2 = .77, and the overbenefit scale (M= 2.30, SD = 1.07) also showed scores that were significantly lower than the midpoint, t (152) = -l9.52, p < .01, n2 = .71, indicating that both vignettes were viewed as unfair. A 2 (gender: male/female) X 2 (equity: over/underreward) X 2 (relationship dimension: task/social) ANOVA was performed to check for gender differences in fairness. There was a significant difference for gender for the scale following the equitable vignette, F (1, 302) = 14.9, p < .01, n2 = .04, such that women (M = 5.96, SD = 0.92) believed the equitable vignette to be more fair than men (M = 5.50, SD = 0.96). 49 There was no difference for gender for the scale following the inequitable vignette, F ( l , 301) = 2.36,p = .13, n2 = .01. 50 Results Primary Study Results Overview. To test the first research question and hypotheses Hla, Hlb, H4a, and H4b, a 2 (equity level: over/underreward) x 2 (relationship dimension: task/social) x 2 (participant gender: male/ female) x 2 (behavior change: own/ partner—repeated) mixed ANOVA was performed. Hypotheses two and three were tested using a 2 (equity level: over/underreward) x 2 (relationship dimension: task/social) x 2 (participant gender: male/ female) ANOVA to check for differences in the change in the liking score from the equitable vignette to the inequitable vignette (the change in liking score was calculated by subtracting each participant’s liking for Chris following the inequitable vignette from their liking for Chris following the equitable vignette). To test for a possible student/employed adult disparity, a 2 (type: student/employed adult) X 2 (equitable status: over/underreward) X 2 (relationship dimension: task/social) AN OVA was performed on change in liking, intent to change own actions, and intent to attempt to change other’s behavior. Of those analyses, none yielded a significant difference for the type (student/employed adult) variable, so the results for those ANOVAs and the differences between students and employed adults will not be discussed further. To test for a difference in variables due to friendship makeup, a 2 (Chris’ gender: male/female) X 2 (participant gender: male/female) X 2 (equitable status: over/underreward) X 2 (relationship dimension: task/social) ANOVA was performed on change in liking, intent to change own actions, and intent to attempt to change other’s behavior.1 Some differences due simply to gender rather than friendship makeup of the participant did appear and will be discussed further in regards to each 51 variable. Descriptive statistics and correlation matrices for liking change, changing own behavior and changing other’s behavior are provided in Tables 7 and 8. Change own behavior and changing other ’s behavior. To test RQl, H1 a, Hlb, H4a, H4b, and for gender differences, a 2 (equity level: over/underreward) x 2 (relationship dimension: task/social) x 2 (participant gender: male/ female) x 2 (change: own behavior /other behavior — repeated) mixed ANOVA was performed. As no specific predictions were made regarding gender differences, they will be discussed following the other analyses. All cell means and standard deviations for this mixed ANOVA are provided in Table 9. RQl asked whether intent to change one’s own behavior or intent to attempt to change a partner’s behavior would be more likely. The ANOVA yielded a significant difference for the type of change (own: M = 4.69, SD = 1.25; partner: M = 3.66, SD = 1.31), F (1, 299) = l46.74,p < .01, n2 = .13 between the two types of change (change own actions/ attempt to change partner’s actions). There was a significant difference between the intent to change own behavior versus intent to attempt to change a partner’s behavior depending on the situation (overbenefit/ underbenefit/ task/ social) portrayed: changing own behavior was more likely overall than attempting to change a partner’s behavior. The ANOVA for relationship dimension (task: M = 4.32; social: M = 4.03) was significant, F (1, 299) = 6.71, p = .01, 112 = .02. Overall, change is more desirable in the task situation than the social situation. The interaction between type of change and relationship dimension was not significant, F (l, 299) = 0.21, p = .65, n2 = .00. 52 The ANOVA for equity level (overreward: M = 4.25; underreward: M = 4.11) was not significant, F (l , 299) = 0.72, p = .40, n2 = .00. The interaction between type of change and equity level was significant, F ( 1, 299) = 175.46, p < .01, n2 = .15. See Figure 1 for means. Post hoc analyses were conducted to check for significant differences between the two types of change for each equity level (over/underreward). In the overbenefit situation, there was a significant difference between own change (M = 5.34, SD = 1.17) and change in other’s behaviors (M= 3.15, SD = 1.33), t (152) = 15.65,p < .01, n2 = .62. In the underbenefit situation, there was no difference between own change (M= 4.06, SD = 1.00) and change in other’s behavior (M= 4.15, SD = 1.10), t (158) = - .96,p = .34, n2 = .01. Also, the interaction between type of change, equity level, and relationship dimension was significant, F (1, 299) = 13.39, p < .01, n2 = .03. See Figure 2 for means. This interaction shows results corresponding to hypotheses one (a and b) and four (a and b). T-tests were conducted to allow for specific comparisons according to the hypotheses. Hypothesis la predicted that in an underbenefit situation, changing own behavior would be more likely in the task dimension than in the social dimension. This hypothesis was tested by comparing task and social situations specifically in an underbenefit, change own behavior situation. There was not a significant difference in changing own behavior in an underbenefit situation whether it was in the task dimension (M = 3.23, SD = 1.40) or the social dimension (M = 3.06, SD = 1.26), t (150) = .76,p = .45, n2 = .00. Hypothesis 1b was tested by comparing task and social situations specifically in an underbenefit, change other’s behavior situation. Changing other’s behavior was significantly more likely in an underbenefit situation if it was in the task dimension (M = 53 4.43, SD = 1.02) than in the social dimension (M= 3.89, SD = 1.12), t (156) = 3.20,p < .05, n2 = .06. These data are consistent with the prediction made by hypothesis 1b: that changing other’s behavior in an underbenefit situation would be more likely in the task dimension than the social dimension. Hypothesis 4a was tested by comparing task and social situations specifically in an overbenefit, change own behavior situation. Changing own behavior was significantly more likely in an overbenefit situation if it was in the task dimension (M = 5.58, SD = 1.09) than in the social dimension (M = 5.06, SD = 1.20), t (151) = 2.84,p < .05, n2 = .05. The data are contrary to the prediction made in hypothesis 4a: that changing own behavior in an overbenefit situation would be more likely in the social dimension than the task dimension. Hypothesis 4b predicted that changing other’s behavior in an overbenefit situation would be more likely in the social dimension than the task dimension. This hypothesis was tested by comparing task and social situations specifically in an overbenefit, change other’s behavior situation. There was no difference between changing other’s behavior in an overbenefit situation depending on whether it was in the task dimension (M = 4.01, SD = 1.00) or the social dimension (M= 4.13, so = 1.01), z (157) = -0.73, p = 47,172 = .00. This ANOVA also revealed a significant interaction between type of change and gender, F (1, 299) = 11.55, p = .01, n2 = .01. Post hoc analyses were performed to test individual means. Men (M = 3.89, SD = 1.27) were more likely to change other’s behavior than women (M = 3.52, SD = 1.33), t (305) = 2.34, p < .05, n2 = .02. Women (M = 4.73, SD = 1.33) were not more likely to change their own behavior than men (M = 4.62, SD = 1.13), t (307) = on, p = .46, n2 = .00. See figure 3 for means. 54 There was also a significant interaction between the type of change, gender and equity level, F (1, 299) = 14.76,p < .01, n2 = .03. Women (M= 5.61, SD = 0.97) were more likely than men (M = 4.95, SD = 1.30) to change their own actions in an overreward situation, t (149) = -3.53, p < .05, n2 = .08. Men (M= 3.42, SD = 1.36) were more likely to change their partner’s actions in an overreward situation than women (M = 2.96, SD = 1.29), t (148) = 2.12, p < .04, n2 = .03. In an underreward situation, men (M = 4.26, SD = 0.78) were more likely to change their own actions than women (M = 3.95, SD = 1.11), t (156) = 2.10, p < .05, n2 = .03. In an underreward situation, there was no difference for changing their partner’s actions (men: M = 4.37, SD = 0.97; women: M = 4.02, SD = 1.17), t (155) = 1.91,p = .06, n2 = .03. See figure 4 for means. Change in Liking. First, a paired-samples t-test was performed to check for differences between liking for the partner in the equitable vignette versus the inequitable vignette. Participants noted significantly more liking for their partner in the equitable relationship (M = 5.15, SD = 0.90) than the inequitable relationship (M = 3.94, SD = 1.35), t (311) = 15.05, p < .01, n2 = .42. Participants clearly liked their partner more when in an equitable relationship than when in an inequitable relationship. A 2 (equity level: over/underreward) x 2 (relationship dimension: task/social) x 2 (participant gender: male/ female) ANOVA was performed to test hypotheses two and three and for gender differences simultaneously. The ANOVA for equity level was significant, F (l, 301) = 119.22, p < .01, n2 = .26, indicating participants had a larger change in their liking from the equitable to the inequitable vignette for the underreward situation (M = 1.96, SD = 1.30) than for the overreward situation (M = 0.45, SD = 1.1 1). The ANOVA for relationship dimension was not significant, F (1, 301) = 0.35,p = .55, 55 n2 = .00. Taken together, hypotheses two and three predict an interaction effect between relationship dimension and equity level. Specifically, it is predicted that liking will change more in the case of underreward in the social dimension than the task dimension, but in the case of overreward, liking will change more in the task dimension than the social dimension. The interaction between equity level and relationship dimension, was significant, F (l, 301) = 19.54, p < .01, n2 = .04. Post hoc tests were performed to test for differences between individual cells. Means are reported in Figure 5. In response to overreward, participants decreased their liking for their partner more in a social situation than in a task situation, t (151) = -3.06, p < .01, n2 = .06. In response to underreward, participants decreased their liking for their partner more in a task situation than in a social situation, t (157) = 3.73, p < .01, n2 = .08. See Figure 5 for means. These data are inconsistent with both hypotheses two and three. In fact, an interaction occurred exactly opposite the predicted directions. Liking change in response to underreward was higher in a task than a social situation and liking change in response to overreward was higher in a social than a task situation. There were no significant main effects or interaction effects for gender on change in liking. 56 Discussion Summary. The current study was designed to test the differences in responses to inequity in varying situations, including task versus social dimensions in blended friendships and underreward versus overreward inequity. Of central interest was whether participants would choose to restore equity through psychological or behavioral means and whether or not one or the other would be more desirable given the differences between task versus social dimensions and underreward versus overreward inequity. Participants were randomly assigned to either the task or social dimension and either underreward or overreward inequity. Vignettes were prepared to represent each cross of the two conditions and administered to participants along with scales inquiring how participants intended to respond to the relationship depicted. Restoration of psychological equity was operationalized as reducing liking for the partner and restoration of actual (behavioral) equity was operationalized in two ways: as intent to change one’s own behavior and as intent to attempt to change one’s partner’s behavior. It was found that as a way of restoring actual equity, people have stronger intention to change their own behavior than to try to change their partner’s behavior. However, that differed some with the different types of inequity. When people were overrewarded, people were more likely to intend to change their own behavior. In the case of underreward, there were no significant differences. There were also different responses when the two types of change were examined according to both equity level and relationship dimension. Specifically, people were more likely to intend to try to change their partner’s behavior in the task dimension if the situation was one of 57 underreward. Further, people were more likely to change their own behavior in the task dimension (as opposed to the social dimension) in the case of overreward. As for restoring psychological equity, overall, liking decreased more in an underreward situation than in an overreward situation. This is qualified by differences in liking change for task versus social and under versus overreward. Contrary to the predictions made, in an overreward situation, liking decreased more in the social dimension than the task dimension, while in an underreward situation, liking decreased more in the task dimension than in the social dimension. Relationship dimension Verbrugge (1979) suggests that “Because face-to-face contact facilitates friendship, people are likely to develop friendships with. . .coworkers. . .coworkers tend to offer certain services and emotional supports” (p. 1305). As previously stated, these friendships are inherent in many people’s lives. Further, difficulties in blended friendships may arise simply from the nature of multidimensional relationships. Bridge and Baxter ( 1992) note that people in blended friendships may not know which dimension of the friendship to enact at a given time. Studying how people respond to inequity in a multidimensional relationship is pertinent and practical when dealing with workplace interactions. Behavioral restoration. The current research shows that responses to inequity differ across relationship dimension. Previous studies have found differences due to relationship dimensions between friends and romantic partners (Winn et al., 1991), cohabiting and married couples (Kollock et al., 1994), and fiiends and strangers (Morgan & Sawyer, 1979). Because of the main effect for relationship dimension, it is pertinent to 58 ask why people respond differently in a work situation than a social situation. If behavioral resolution is a more immediate and direct form of equity restoration than psychological restoration, it may be more desirable in the work environment because of the need to maintain a current balance (equity). To return to a quote presented earlier, “The accounting system of friends and lovers is likely to be slower in its response...” (p. 341 , Kollock et al., 1994). In the social dimension of a relationship, the slower “accounting system” may allow for greater lag in returning favors, hence allowing some inequity to exist for a time without the need for action. In regards to equity and time, Cosier and Dalton (1983) point out the static nature of equity theory and suggest that perhaps a more dynamic model might clarify responses to inequity. Another explanation for task and social differences may be that social norms in the business world may be different from social norms in the “fiiend” world. For instance, a particularly cutthroat business combined with a blended friendship might make blended friendship difficult (Bridge & Baxter, 1992) and cause some differences across the two dimensions. Although there were statistically significant differences, the effect sizes of the differences between task and social dimensions overall were quite small. (Even when considered in juxtaposition with different levels of equity, the effect sizes for relationship dimension differences were still quite small.) Returning to the idea of different social norms in different relationship dimensions, the small effect sizes could be because although there are differences in how people act in the task and the social dimension, having a blended friend puts a damper on those differences. For instance, maybe two blended friends, Harold and Harvey, work in a cutthroat environment. However, they are 59 friends outside of the workplace—they and their wives go out to dinner and vacation together frequently. Clearly, these two dimensions are diametrically opposed to each other—in one, it’s a dog-eat-dog world, and in the other, exploitation is unthinkable. But these rules are established by society; although Harold and Harvey might feel the differences as dictated by society, they might choose to ignore them and support each other in their workplace as well as outside of the workplace. Cognitive dissonance theory (F estinger, 195 7) would suggest a need to be consistent with one friend across the two differing dimensions (work/outside of work); hence Harold and Harvey’s choice to support rather than undermine each other at work makes sense. Small differences such as those found in the current research might exist as a residual effect of training into a certain type of workplace followed by a decision to ignore that training. A further explanation of small effect sizes could exist in the form of “equity with the world.” A construct advanced by Austin and Walster (1975), equity with the world suggests that people have an internal gauge by which equity across all their relationships is measured. For instance, if a person is underrewarded in one relationship, that imbalance may be nullified by the presence of overreward in a different relationship. If some level of equity with the world exists, then only small amounts of equity restoration may be necessary in a given relationship, such as the one presented to participants in the current research. Equity with the world could also be applied to multidimensional relationships as if each dimension was a separate relationship. Following with that idea, some participants may have assumed some balancing action was taking place in the dimension of the relationship not discussed in the vignettes (i.e. if underreward was depicted in the social dimension, they assumed overreward in the task dimension, though 60 it was not stated), therefore negating the need for them to make a great change in the inequity presented. Psychological restoration. Despite the difference present in the area of behavior change, there were no significant differences between decreasing liking in task and social situations; people decreased their liking about the same whether it was in the task or social situation. Perhaps changing liking for the partner in an inequitable relationship happens in an unconscious way so that there is no differentiation between task and social dimensions of the relationship. In other words, rather than a thought process such as, “Gee, this relationship seems unfair, I think I will like my partner less now,” the decrease in attraction occurs without conscious consideration of what type of relationship it is. Equity level Psychological restoration. People liked their partner more in an equitable relationship than in an inequitable relationship, as was expected. It seems obvious that people would like a partner less in an inequitable situation, and it has been empirically demonstrated that attraction is lower in inequitable situations than equitable situations (Rees & Wechler, 1984). Extending that work, this study yields a more exact picture of how liking changes in response to the differing types of inequity. Specifically, people dislike those who take advantage of them more than those of whom they take advantage. Equity theory notes that although participants will feel distress in response to both types of inequity, there may be differences in how the distress is felt. For example, Hegtvedt (1990) noted that subjects may feel guilt in response to overreward while they may feel anger in response to underreward. It is possible that the different affective responses expected in response to the two types of inequity may be the root of the 61 decrease in liking. If a person is underrewarded, and hence angry, it seems quite likely that one result would be an increase in dislike or a decrease in liking for the partner. By the same token, feeling guilty due to overreward may prompt a person to decrease the importance of the relationship by liking their partner less. Behavioral restoration. Paralleling this difference in affective response to the types of inequity is the difference in how people may respond to the two types of inequity. Greenberg (1990) points out the fuzziness of equity theory on the point of how people will respond to inequity in terms of behavior in differing situations. Walster et al. (1973) give specific information regarding responses to inequity in exploitative relationships and helping relationships, but do not provide specific predictions as to how individuals in intimate relationships such as friendship will respond to inequity. In regards to using equity as a governing principle for intimate relationships, Walster et a]. simply mention the matching hypothesis: that people will generally find intimate others who are similar to them in social worth. Although it is not delineated by equity theory, based on the results from this study, it seems clear that there are differences in how people will respond to overreward versus underreward. It is possible that responsibility in the relationship comes into play in this instance. Specifically, when people are overrewarded, they have more responsibility in a given relationship. Since social norms dictate that friendships should be equitable, something about the inequitable relationship needs to change. The overrewarded person is the one with the responsibility to do the changing; therefore, changing one’s own behavior is a natural response to overreward. 62 In the underbenefit situation, there was no strong preference for changing own behavior or attempting to change a partner’s behavior. The lack of differences here could simply indicate that both methods of restoring actual equity were viewed as equally feasible; therefore people would be just as likely to do either one in efforts to restore equity. Relationship dimension and equity level have been discussed here as separate influences on choice of equity restoration methods. Although they both have effects alone, it is also important to test for effects of the combination of the two influences. The combined effects will be considered next. Relationship dimension and equity level Behavioral restoration. In a task relationship, underreward garners strong responses of intent to change other’s behavior (as opposed to intending to change one’s own behavior). Going back to the responsibility explanation, being underrewarded in a task relationship would mean being the one with less responsibility in the relationship. According to the previous logic, the one with less responsibility expects the one with the more responsibility in the relationship to right the wrong (inequity) in a relationship which society says should be equitable. The results found here are consistent with this logic. Being overrewarded in a task versus a social situation did not make a difference in attempts to change a partner’s behavior and being underrewarded in the social versus the task dimension of a relationship did not make a difference in the intent to change one’s own behavior. Both could be due to the lack of an equity orientation. Scores on underreward for both social and task were below the mean, indicating that people might 63 be likely not to do anything in response to inequity. This could indicate a communal orientation rather than an equity orientation. Essentially, a communal orientation governs by suggesting that benefits should be given only in response to needs; rather than the give and take of reciprocal orientations such as equity, a communal orientation dictates that if one partner is always needy, the other will always be giving, no matter how inequitable it is. The possibility of communal orientation will be discussed further later, in terms of directions for future studies. Psychological restoration. People who are underrewarded are being exploited, or at the very least, taken advantage of; this explains why they would decrease their liking more when being underrewarded than overrewarded. Being friends with someone who is “weak” or easy to take advantage of is not very desirable. Consequently, decreasing liking for a partner of whom you are taking advantage socially makes sense. Differences in overreward responses due to social versus task situations indicate that people will restore equity psychologically more in the social dimension. This could be due to equity sensitivity, a construct proposed by Huseman et a1. (1987). This construct postulates three types of people: benevolents, equity sensitives, and entitleds. Of interest here are the entitleds, who don’t mind (and even enjoy) being overrewarded. Given that these hypothesized entitleds will psychologically restore equity, why might they do it more often in the social than task dimension? Perhaps in the social element, it is more pleasant to think that their friend is a very giving person and likes being that way (benevolent) rather than that they themselves are exploiting the friend by being overrewarded. 64 On the other hand, if these people are not entitleds, rather just equity sensitives or benevolents, who are quite uncomfortable with being overrewarded, cognitive dissonance (Festinger, 1957) can offer an explanation. If, based on social norms (or internal standards of equity), people believe they should not take advantage of their friends, then the idea that they are overrewarded is in disagreement with their perception of what they should be (fair friends). Resolving this disagreement psychologically would allow their idea of self to be congruent with their observations of themselves. Since according to the current research, people restore equity psychologically more in the social setting, it follows that they restore equity psychologically less in the task setting. It is feasible that the social norm of not taking advantage of friends gets thrown out the window, so to speak, in the workplace. In a business setting, efficiency and effectiveness rather than fiiendliness are valued; business dealings are usually rather impersonal. Therefore, being overrewarded would not cause cognitive dissonance and the need for psychological restoration of equity because overreward inequity may be sanctioned in the work setting. In fact, overreward might be considered a prestigious position in a work setting, even something to brag about. Clearly if this is the case, there is no need to resolve cognitive dissonance caused by inequity in the friendship. However, in response to underreward, psychologically restoring equity is more likely in the task dimension than in the social dimension. Cosier and Dalton (1983) suggest that a possible moderator between inequity and taking action could be the costs involved in taking action to restore equity in the workplace. Specifically, restoring actual equity in the workplace might involve confrontation with a superior (Cosier & Dalton), which could engender negative feelings. Further, even if that risk is small enough to take, 65 there is the added possibility that resolution of the inequity will not result from the action taken (Cosier & Dalton). This might explain why people would be more likely to psychologically restore equity in a work situation than in a social situation. Why it might occur in an underrewarding situation could be explained by the need to save face. Workers might not want to appear as if they are complaining about unfair treatment, especially if they are complaining to the person doling out the treatment (the manager). Therefore, a much safer route would be to restore equity psychologically. As an added note regarding relationships between actual and psychological equity restoration methods, zero-order correlations for the three dependent variables (decrease in liking, intent to change own behavior, intent to attempt to change partner’s behavior) are shown in table 8. Of interest are the significant relationships that appear between these three variables in the various combinations of relationship dimension and equity level. Although there are differences in choices of method in different situations (such as choosing psychological restoration in a task, underbenefit situation), clearly the methods of restoring equity may oftentimes work in concert with one another rather than operating separately and independently of each other. Gender and friendship composition. Carrell and Dittrich (1978) cite multiple studies which suggest that gender is important as a motivator of differences in perceptions of partner inputs (Taynor & Deaux, 1973; Taynor & Deaux, 1975) and equity perceptions (Wahba, 1971). Canary and Stafford (1992) found gender differences in relational maintenance strategies employed in the face of inequity. In the current research, when considering the combined effects of the type of change and gender, differences emerged. Overall, men were more likely than 66 women to attempt to change their partner’s behavior. Both men and women were equally likely to change their own behavior. According to Lane and Messé’s (1971) research, women are less self-interested than men. The current research corroborates that finding, because the more self-interested men try to make their partners conform to their needs rather than changing themselves, while the less self-interested women try to make their partners happy by not asking them to change and changing their own behavior. This differs across equity level, however. For a combination of the type of change, gender, and equity level, more differences were apparent. In an overreward situation, women were more likely than men to change their own actions. This is consistent with previous research (Sprecher, 1992), which suggested that women believe they will be more concerned with equity in relationships and will work more than men to restore equity. In fact, Sprecher’s evidence revealed that when predicting their own behavior, women believed more strongly than men that they would change their own actions in an overreward situation. Also in an overreward situation, men were more likely than women to attempt to change their partner’s behavior. Lane and Messé’s (1971) findings also offer support to this finding: when being overrewarded, women, being less self-interested, change their own actions to restore equity, while men, being more self-interested, do not change themselves, but try to change their partners’ actions. Although this finding is not statistically demonstrated in Sprecher’s (1992) findings, it is logically consistent: Sprecher’s finding that women will change their own behavior in an overreward situation leads to the conclusion that men are more likely to do just the opposite (change other’s behavior) in response to the same situation. 67 In an underreward situation, however, men were more likely than women to change their own actions. Again, this is congruent with Lane and Messé’s (1971) finding: men, who are more self-interested, stand up for themselves when they are being taken advantage of, while women are less likely to do so. Also in an underreward situation, men and women were equally likely to try to change their partner’s actions. So, when being underrewarded, men take action themselves to remedy inequity, while both men and women try to change their partner’s actions to restore equity. These findings extend Sprecher’s (1992) work to include more information on how men and women will likely respond to underreward. As to whether the composition of the friendship results in differing types of equity restoration, only one significant difference was found. In a social situation, females who perceived the friend to be female decreased their liking more than females who perceived the friend to be male. This could be because women are thought to be more emotional; hence, logically, decreasing liking would be a more effective method of equity restoration to use on a female friend, whereas a male fiiend might not even notice a difference. This would also explain why the difference was found in the social dimension, which allows for emotional interactions, while the task dimension requires businesslike transactions. Limitations. One potential limitation of the study could have been using only a student sample. Gordon, Slade, and Schmitt (1986) offer a pseudo-meta-analysis of a number of studies to determine whether using students versus real-world subjects might cause a problem for generalization. They concluded that indeed, using students would be problematic based on the studies they examined. 68 However, in realizing the built-in nature of student samples and the difficulty of getting nonstudent samples, Gordon et al. (1986) offered suggestions for how to improve the generalizability of data from student samples. One of their suggestions was employed in this research: to use subjects with demographic or interest profiles similar to the nonstudent population of interest. The problem of low generalizability due to a student sample was pre-empted by limiting the student sample to upper-level students (e. g. juniors and seniors), who would likely be working currently or have some work experience, and therefore be similar to working adults on the variable of interest (working friendships). As another measure against this potential problem, a sample of employed adults was collected. When testing for differences between the student and employed adult sample, no disparity was apparent. Another possible explanation for the lack of differences comes from other researchers interested in the student- nonstudent debate. Farber (1952) and Kruglanski (1975) suggest that using student samples will not be problematic if the research examines relationships between general behavioral constructs. Although this research could be considered to fall into this category (general study of distributive justice), it also involves specific situational confines, and therefore, taking measures against a biased student sample was pertinent. Given that about half of the student sample indicated that they worked 20 hours a week or more at a current job, the samples seem to be rather homogenous, as intended, hence causing the lack of differences between students and employed adults. Given that there were no significant differences between the two, generalizability is increased to include a working population. 69 A second limitation lies in the assumption that people are able to and do differentiate between task and social dimensions of their friendships, and further, that they separate the two entirely such that they would respond to inequity in the task dimension by attempting to restore equity in the task dimension and they would respond to inequity in the social dimension by attempting to restore equity in the social dimension. It is possible that people either choose not to separate the task and social dimensions of their blended friendships, or that they simply do not possess the ability to separate the two. This may account for the small amount of the differences due specifically to the task versus social distinction. It is also possible that since the vignettes only presented specific information regarding one dimension of the friendship that it might not be apparent to participants that both dimensions existed, and they may have been biased to respond as if the relationship had only a task dimension or a social dimension. A third weakness is the use of vignettes rather than actual friendships (behavioral observation versus self-report). Obviously, self-report based on a stimulus such as the vignettes used in this study cannot compare to the richness of a blended friendship that exists in real life. Using real people with real blended friendships may yield different results. Seventy-nine percent of the participants indicated that they have a coworker who is also a friend, so it seems likely that most of this sample would be able to conjure up a realistic image of what a blended friendship is like in real life. Using participants who work or have worked should allow for a reasonably realistic portrayal of how real employed people might react to a situation such as the one described. 70 A related problem is the issue of intention. The current research measured only intent to take various actions. This might be viewed as a limitation, in that an actual behavioral study may show people acting differently than they might intend to. However, the theory of reasoned action (Ajzen & Fishbein, 1980) suggests that behavioral intentions lead to behavior consistent with the intentions. Kim and Hunter (1993) did a meta-analysis on the relationships between attitudes, behavioral intentions, and behavior itself. In this meta-analysis, they found a mean correlation of .82 between behavioral intention and behavior, indicating that the two are strongly related. Pertaining to the current research, it would be interesting to see if the intention to enact the various types of equity restoration would lead to actual behaviors. In the workplace, some factors that could deter people from following their behavioral intentions could involve such things as locus of control, hierarchical status in the workplace, and other existing relationships that conflict with the examined relationship (i.e. your partner is friends with the boss), all of which would be good factors to examine in a study of behavioral responses to inequity. A final point of interest, which could be a potential problem, is that of cumulative equity. Cosier and Dalton (1983) suggest including past instances of inequity as part of analysis of an inequitable relationship because of the potential effects of past inequity. The method used in the current research presented a situation in which inequity had been occurring in small increments over a length of time. Despite the incremental nature of the inequity (i.e. inequity had taken place in the same form multiple times), it was considered to be one instance of inequity. Carrell and Dittrich (1978) also point out that the perception of inequity is something that may ebb over time. It is assumed that since participants read the vignettes and immediately answered questions about them, that the 71 situation was quite salient and therefore that they responded accordingly strongly, and without the effects of past relational inequities. Future studies could look at the cumulative effect of inequity as a variable of interest. Future directions. Previous research applying equity theory to interpersonal relationships has focused one dimension at a time, rather than considering that two dimensions can co-exist in a relationship. Others have studied multiple types of relationships and compared the types (e. g., Kollock et al., 1994; Winn et al., 1991), but not looked at multidimensional relationships. By examining blended friendship, a relationship that encompasses multiple dimensions, the current research focuses attention on the dialectical tensions that may exist in multidimensional relationships. Since some differences due to the relationship dimension seem to exist, though somewhat weak, studying other multidimensional relationships may be of interest. For example, some organizations do not allow romantic relationships between coworkers. Since there seems to be an interaction between different segments of multidimensional relationships and the different types of inequity, extension of study to multidimensional romantic/coworker relationships could shed light on the risks taken by an organization that allows romantic relationships. Finding out how these individuals are likely to respond to different types of inequity could also relieve organizations’ worries about bad romance backfiring on their workplace. For example, if individuals are more prone to resolve workplace inequity psychologically or only to resolve inequity in the romantic dimension outside of work rather than allowing it to spill over into the work dimension could be quite practical knowledge for organizations with dating employees. 72 Although the current research focused on equal-status partners (i.e. no supervisor- subordinate relationships), and assumed that these partners could resolve inequity with one another, the possibility exists that a third party may necessarily be involved. Given that a third party in the form of a boss may exist, the strong deference subordinates have to superiors (Peabody, 1962) may affect their responses to inequity. Specifically, this may come into play in the form of locus of control. Whether a person thinks they have the ability to control the inequity in a work relationship might depend on whether they think it is within their control or it is out of their hands and in someone else’s, such as the boss or the partner. People with external loci of control will likely not respond to inequity in the workplace in the same way that those with internal loci of control will. Hence, locus of control could have a mediating effect between inequity and attempts to restore actual inequity behaviorally. Looking at unequal status partners or simply measuring locus of control in participants would be useful in future studies of this type of relationship. Another individual difference that may be of interest is equity sensitivity. It has been postulated that a construct known as equity sensitivity exists in all people, and encompasses three types: benevolent, equity sensitive, and entitled (Huseman, Hatfield, & Miles, 1987). An ‘equity sensitive’ behaves according to the predictions of equity theory—this person prefers to have a balance, or equity, in their relationships. In contrast, a ‘benevolent’ is more comfortable with underbenefit in a relationship, prefening to be giving in his/her relationships. Huseman et a1. characterize ‘benevolents,’ as people whose “contentment derives from perceptions that their outcome/input ratios are less than the comparison other’s” (p. 225). On the other hand, an ‘entitled’ is quite comfortable 73 being overbenefited in a relationship, believing him/herself to be deserving of more benefit. Huseman et al. point out that for ‘entitleds,’ “Distress would occur if they were not ‘getting a better deal’ than their comparison other” (p. 225). Huseman, Hatfield, and Miles (1987) tested their contention that equity sensitivity is a construct, and did indeed find differences for different types of equity sensitivity in responses to varied conditions of inequity. Patrick and Jackson (1991) also found differences in responses to inequity that indicate the existence of a personality variable, equity sensitivity, which moderates individual’s reactions to inequity. Equity sensitivity has most often been related to other constructs such as job satisfaction (e. g. Huseman et al., 1987), job performance (6. g. Bing & Burroughs, 2001), and even business ethics (Mudrack, Mason, & Stepanski, 1999). Clearly, equity sensitivity could help explain why certain people respond to inequity differently than others. It would be a valuable addition to future studies. Another area to examine in the multidimensional relationship realm presented here is whether the predictions (Walster et al., 1973) of feelings of guilt in response to overbenefit and anger in response to underbenefit will hold true in these types of relationships. If people do indeed feel anger in response to underbenefit, a greater understanding of these processes could be helpful in organizations where members have historically been violent. Organizations such as the US. Postal service or schools may benefit from recognizing that inequity exists in working relationships (even group class projects) and that people will have feelings in response to inequity. How people translate those feelings into action is another question yet to be answered. Information regarding this might further benefit organizations such as those 74 mentioned above. If one feels guilty about being overrewarded, what resolution can be had? Similarly, if one feels angry about being underrewarded, what action might one take to resolve that inequity? Study of the relationships between type of inequity, affect induced by inequity, and response following from said inequity and affect seems practically valuable in the working world. Another interesting factor to be considered is severity of inequity or severity of dilemma. A finding by Greenberg (1988) indicated that the amount of increase or decrease in outputs of employees whose office status was changed was indeed proportional to the amount of increase or decrease of status. So, if an employee was moved to the office of someone two levels above him in the organization, his outputs increased more than those of an employee moved to the office of someone only one level above him. The strength of the inequity in a relationship dictates the strength of the reaction to it. Since the current study attempted to monitor reactions to inequity, this could have implications. For instance, perhaps in the case of stronger inequity, a member of a blended friendship would be more likely to resolve inequity by taking an action than to resolve inequity by rationalizing it to himself. Perhaps if the inequity is of a weaker nature, the blended friend would be more likely to psychologically resolve the inequity rather than confronting his friend. The strength of the inequity, although not varied here, could be an interesting factor to examine in future studies. In the case of underreward in this study, whether the inequity was in the task or social dimension, participants’ intent to change their own behavior was below the midpoint (which indicated neutrality). This could indicate the possibility that people might not do anything if underrewarded in their blended friendships. Further, although 75 men were more likely to intend to change their partner’s behavior than women, both men and women’s scores were below the midpoint, again suggesting that they might be likely not to do anything to remedy the inequity. It has been suggested that people may not respond to inequity as Walster et a1. (1973) have suggested. Consider that it is being assumed that people will do something to change the inequity. What if this was not the case? What if people did not follow social norms dictating the necessity for equity in relationships? Sprecher (1992) suggests that people might do nothing in response to inequity in their relationships. It is possible that doing nothing may indicate the lack of a governing principle involving social exchange of any kind; rather, people who do nothing may actually have a communal orientation rather than an equity orientation to their social relationships. In a relationship with a communal orientation, the members expect to give a benefit in response to a need for the benefit rather than give in response to receiving. This is inherently different from an equity orientation, which suggests giving/receiving ratios should be equal for both members of the relationship. Van Yperen and Buunk (1991) found that in Americans, equity theory predictions held true for participants low in communal orientation, while participants high in communal orientation, contrary to equity theory predictions, were more satisfied with overrewarding relationships than they were with equitable relationships. Communal orientation lends support to the possibility of responses contrary to equity theory predictions, including the “do nothing” response to inequity. Returning to the example of Heidi and Hannah, if Heidi and Hannah’s relationship was governed by communal principles, Hannah being overrewarded would 76 not be perceived as such. Following a communal orientation, the extra effort Heidi was putting into the relationship was simply in response to Hannah’s need for help. In this instance, there would be no need for repayment. In fact, according to Clark and Mills (1979), if there was a communal expectation in the relationship, and a person such as Heidi was to ask for something in return for her extra effort (reciprocation), the effect would be decreased attraction between the two. Clark and Mills assume this is because the exchange of benefits is inappropriate in a communal relationship. Communal orientation may be a subject of interest for further studies. Conclusion. As would be predicted from equity theory, people responded differently to underreward than they did to overreward. Further, as postulated in this research, people also responded differently to inequity in the task dimension of their relationships than to the social dimension of their relationships. The current research is a step toward clearer, more specific predictions of responses to the two types of inequity in multidimensional relationships and in specific situations. How people respond to inequity in multidimensional relationships, particularly blended friendships, is of practical value to organizations dealing with this very issue. The continued study of this phenomenon should aid organizations in understanding the inner workings of multidimensional relationships, which is crucial as these relationships may have a devastating or enriching effect on the organizations themselves. 77 I The ANOVA for participant gender and Chris’ gender on liking change was not significant, F (1, 285) = 2.09, p = .15, n2 = .00, indicating the fiiendship makeup did not make a significant difference in the amount of liking decrease in the face of inequity. The interaction between participant gender, Chris’ gender, and relationship dimension (task/social) on liking change was significant, F (1, 285) = 4.48,p < .05, n2 = .01. Post- hoc t-tests were performed to pinpoint specific significant differences between means. One significant difference was found, in the social dimension, between participants who were female and believed Chris to be male (M = 1.05, SD = 1.38) and participants who were female and believed Chris also to be female (M = 1.65, SD = 1.20), t (95) = -2.17, p < .05, n2 = .05. The interaction between participant gender, Chris’ gender, and equity level (overreward/underreward) did not have a significant effect on liking change, F (1, 285) = 3.44, p = .07, 172 = .01. The interaction between participant gender, Chris’ gender, relationship dimension (task/social) and equity level (overreward/underreward) also did not have a significant effect on liking change, F (l, 285) = .01, p = .94, n2 = .00. All means for this ANOVA are provided in table 10. A 2 (change own actions/change partner) repeated subjects x 2 (participant gender: male/female) x 2 (Chris’ gender: male/female) x 2 (equity level: over/underreward) x 2 (relationship dimension: task/social) AN OVA was performed to test for possible effects of friendship makeup (cross—sex versus same-sex). The interactions between the two gender variables (participant and Chris) and other variables show how much difference in the scores on type of change for the participants is because of the friendship makeup. 78 The interaction between type of change, gender, and Chris’ gender was not significant, F (l , 284) = 1.3 l , p = .25, n2 = .00, indicating that gender did not make a significant difference in the type of change participants indicated they would attempt to instigate. The interaction between type of change, participant gender, Chris’ gender, and relationship dimension (task/social) was also not significant, F (l, 284) = .09, p = .76, 172 = .00. The interaction between type of change, participant gender, Chris’ gender, and equity level (over/underreward) was not significant, F (1, 284) = 3.07, p = .08, n2 = .00. The interaction between type of change, participant gender, Chris’ gender, relationship dimension (task/social) and equity level (over/underreward) was also not significant, F (1, 284) = .21,p = .65, n2 = .00. The interaction between participant gender and Chris’ gender did not make a significant difference in scores on taking action to change inequity, F (1, 284) = .09, p = .76, n2 = .00. Neither did the interaction between participant gender, Chris’ gender, and relationship dimension (task/social), F (l, 284) = 1.01, p = .32, n2 = .00. Similarly, the interaction between participant gender, Chris’ gender, and equity level (over/underreward) was not significant, F ( 1, 284) = .18, p = .67, n2 = .00. Finally, the four-way interaction between participant gender, Chris’ gender, relationship dimension (task/social) and equity level (over/underreward) was not significant, F (l, 284) = 1.71, p = .19, n2 = .00. All means and standard deviations for these interactions are in tables 1 l and 12. 79 APPENDIX A Coworker/Social/Equitable You and Sam are coworkers who are also friends. Not only are you friends, but you also work on similar projects and hold identical positions in the same department of the same company. You have been friends ever since you met at work. Outside of work, you and Sam mountain bike together. You often spend weekends biking trails together, and you also do other things together, like going out for a drink after work. You take turns driving to and from the trails. Since you both have trucks, it is just as easy to transport your bikes no matter who drives. You and Sam take turns bringing snacks and water for the day of biking. Coworker/Social/Overbenefit You and Sam are coworkers who are also friends. Not only are you friends, but you also work on similar projects and hold identical positions in the same department of the same company. You have been fiiends ever since you met at work. Outside of work, you and Sam mountain bike together. You and Sam often spend weekends biking trails together and you also do other things together, like going out for a drink after work. You often ask Sam to drive to the trails when biking, even though your truck is better for transporting your bikes. Not only that, but you usually depends on Sam to bring water and snacks along for the day also. Coworker/Social/Underbenefit You and Sam are coworkers who are also friends. Not only are you friends, but you also work on similar projects and hold identical positions in the same department of the same company. You have been friends ever since you met at work. Outside of work, you and Sam mountain bike together. You often spend weekends biking trails together and you also do other things together, like going out for a drink after work. Sam often asks you to drive to the trails when biking, even though Sam’s truck is better for transporting your bikes. Not only that, but Sam usually depends on you to bring water and snacks along for the day also. Coworker/Task/Equitable You and Sam are coworkers who are also friends. Not only are you friends, but you also work on similar projects and hold identical positions in the same department of the same company. You have been fiiends ever since you met at work. At work, you both keep up with your projects without any trouble. Neither of you has any difficulty completing projects on time. You discuss the projects and share ideas back and forth freely. You lend each other a hand if needed. You and Sam are supportive of each other, but you both work on your own projects alone. C oworker/Task/Overbenefit You and Sam are coworkers who are also friends. Not only are you friends, but you also work on similar projects and hold identical positions in the same department of the same company. You have been friends ever since you met at work. 80 When you are at work, it seems like you can never catch up enough to be on time with your projects. You were working on a project the other day and Sam was the one with all the ideas. Sam is always there to lend a hand when you need help. Sam never complains about helping you, but simply does whatever you need and moves on. C oworker/Task/Underbenefit You and Sam are coworkers who are also friends. Not only are you friends, but you also work on similar projects and hold identical positions in the same department of the same company. You have been fiiends ever since you met at work. When you are at work, it seems like Sam can never catch up enough to be on time with Sam’s projects. Sam was working on a project the other day and you were the one with all the ideas. You are always there to lend a hand when Sam needs help. You never complain about helping Sam, but simply do whatever Sam needs and moves on. F riend/Social/Equitable You and Sam are friends who work together. You work in identical positions in the same department of the same company. Not only do you work together, but you are also friends outside of the workplace. You were friends before you started working together. Outside of work, you and Sam mountain bike together. You often spend weekends biking trails together, and you also do other things together, like going out for a drink after work. You take turns driving to and from the trails. Since you both have trucks, it is just as easy to transport your bikes no matter who drives. You and Sam take turns bringing snacks and water for the day of biking. Friend/Social/Overbenefit You and Sam are fiiends who work together. You work in identical positions in the same department of the same company. Not only do you work together, but you are also friends outside of the workplace. You were friends before you started working together. Outside of work, you and Sam mountain bike together. You and Sam often spend weekends biking trails together and you also do other things together, like going out for a drink after work. You often ask Sam to drive to the trails when biking, even though your truck is better for transporting your bikes. Not only that, but you usually depends on Sam to bring water and snacks along for the day also. Friend/SociaI/Underbenefit You and Sam are fiiends who work together. You work in identical positions in the same department of the same company. Not only do you work together, but you are also friends outside of the workplace. You were friends before you started working together. Outside of work, you and Sam mountain bike together. You often spend weekends biking trails together and you also do other things together, like going out for a drink after work. Sam often asks you to drive to the trails when biking, even though Sam’s truck is better for transporting your bikes. Not only that, but Sam usually depends on you to bring water and snacks along for the day also. 81 Friend/Task/Equitable You and Sam are friends who work together. You work in identical positions in the same department of the same company. Not only do you work together, but you are also fiiends outside of the workplace. You were friends before you started working together. At work, you both keep up with your projects without any trouble. Neither of you has any difficulty completing projects on time. You discuss the projects and share ideas back and forth freely. You lend each other a hand if needed. You and Sam are supportive of each other, but you both work on your own projects alone. F riend/Task/Overbenefit You and Sam are friends who work together. You work in identical positions in the same department of the same company. Not only do you work together, but you are also friends outside of the workplace. You were friends before you started working together. When you are at work, it seems like you can never catch up enough to be on time with your projects. You were working on a project the other day and Sam was the one with all the ideas. Sam is always there to lend a hand when you need help. Sam never complains about helping you, but simply does whatever you need and moves on. Friend/Task/Underbenefit You and Sam are friends who work together. You work in identical positions in the same department of the same company. Not only do you work together, but you are also fiiends outside of the workplace. You were friends before you started working together. When you are at work, it seems like Sam can never catch up enough to be on time with Sam’s projects. Sam was working on a project the other day and you were the one with all the ideas. You are always there to lend a hand when Sam needs help. You never complain about helping Sam, but simply do whatever Sam needs and moves on. Modified Main Study Vignettes Social/ Equitable You and Chris are both friends and coworkers. You work in identical positions in the same department of the same company. Not only do you work together, you are also friends outside of the workplace. Outside of work, you and Chris rollerblade together. You often spend time on the weekends rollerblading just for fun and you also do other things together, like going out for a drink after work. You take turns picking each other up to go to the park, which is halfway between your houses. By the same token, you both take turns bringing water and snacks for after your workout. Social/ Overbenefit You and Chris are both friends and coworkers. You work in identical positions in the same department of the same company. Not only do you work together, you are also friends outside of the workplace. 82 Outside of work, you and Chris rollerblade together. You often spend time on the weekends rollerblading just for fun and you also do other things together, like going out for a drink after work. You often ask Chris to pick you up and drive to a park that is closer to your house, even though it’s quite far out of Chris’s way and Chris has to drive through a lot of traffic to get there. Not only that, but you usually depend on Chris to bring water and snacks for after your workout. Social/ Underbenefit You and Chris are both friends and coworkers. You work in identical positions in the same department of the same company. Not only do you work together, you are also fiiends outside of the workplace. Outside of work, you and Chris rollerblade together. You often spend time on the weekends rollerblading just for fun and you also do other things together, like going out for a drink after work. Chris often asks you to drive and go to a park that is near Chris’s house, even though it’s quite far out of your way and you have to drive through a lot of traffic to get there. Not only that, but Chris usually depends on you to bring water and snacks for after the workout. Task/ Equitable You and Chris are both friends and coworkers. You work in identical positions in the same department of the same company. Not only do you work together, you are also friends outside of the workplace. At work, you and Chris must jointly come up with a monthly budget report for your company’s supply closet. You are supposed to do an inventory of everything in the supply closet and Chris is supposed to do the math to figure out what needs to be replaced and how much the replacements will cost. This arrangement seems to work out fine. You do your part and Chris does Chris’s part. Task/ Underbenefit You and Chris are both friends and coworkers. You work in identical positions in the same department of the same company. Not only do you work together, you are also friends outside of the workplace. At work, you and Chris must jointly come up with a monthly budget report for your company’s supply closet. You are supposed to do an inventory of everything in the supply closet and Chris is supposed to do the math to figure out what needs to be replaced and how much the replacements will cost. However, sometimes Chris can’t seem to get his part done and you end up doing the entire report. You never complain about helping Chris, but simply do the report and move on. Task/ Overbenefit You and Chris are both friends and coworkers. You work in identical positions in the same department of the same company. Not only do you work together, you are also friends outside of the workplace. At work, you and Chris must jointly come up with a monthly budget report for your company’s supply closet. You are supposed to do an inventory of everything in the 83 supply closet and Chris is supposed to do the math to figure out what needs to be replaced and how much the replacements will cost. However, sometimes you can’t seem to get your part done and Chris ends up doing the entire report. Chris never complains about helping you, but simply does the report and moves on. 84 APPENDIX B Prestudy Scales with Items Realism l A friendship like this could develop in real life. 2. It is possible to have a coworker like Sam. 3. You can imagine being in a friendship like this one. 4. This type of relationship with Sam in the story can happen in real life. Fairness 1. Sam is taking advantage of you. 2. You are getting more benefits from the relationship. 3. This friendship seems fair. 4. This work relationship seems fair. 5. You and Sam both get the similar amount of benefits from the relationship. 6. You are taking advantage of Sam. 7. Sam is getting more benefits from the relationship. 85 APPENDIX C Main Study Items Realism 1. A friendship like this could develop in real life. 2. It is possible to have a coworker like Chris. 3. I can imagine being in a friendship like this one. 4. The relationship with Chris could happen in real life.* Fairness 1. Chris is taking advantage of me.* 2. I am getting more benefits from the relationship.* 3. Chris and I both get a similar amount of benefits from the relationship.* 4. I am taking advantage of Chris.* 5. Chris is getting more benefits from the relationship. Liking (Attraction) I like Chris a lot. It seems to me that it is very easy for Chris to gain admiration. T Chris is the sort of person whom I myself would like to be. T I think that Chris and I are quite similar to each other. T I would vote for Chris in a class or group election. T I would highly recommend Chris for a responsible job. T I find Chris a desirable coworker. I find Chris a desirable friend. I care about Chris. .‘OWNP‘MPP’NT‘ Behavior change Change in own behavior 1. I would change my actions because morally it is the right thing to do. 2. I would change my actions because I know I would feel better about myself afterward. 3. I would change my actions because I know I would feel worse about myself if I didn’t. 4. I would change my actions because I am not the kind of person who acts the way portrayed in the story. 5. I would change my actions because it is what a good person would do. 6. I would change my actions because people would like me better if I did. 7. I would change my actions because people wouldn’t like me if I didn’t. Intent to gain compliance 1. I would ask Chris to change his actions because morally it is the right thing to do. 2. I would ask Chris to change his actions because he will feel better about himself if he does. 3. I would tell Chris he will feel worse about himself if he doesn’t change his actions. 86 4. I would tell Chris that because he is such a good person, I know he will change his actions. 5. I would tell Chris that any good person would change his actions. 6. I would tell Chris that people would like him better if he changed his actions. 7. I would tell Chris that if he doesn’t change his actions people will like him less. Do nothing 1. I would not change anything about this relationship. 2. I would let the relationship ride as is for a while. 3. I would wait and see what happens. 4. I would leave it alone. 5. Confrontation is not worth the risk. Termination 1. I would end the friendship. 2. I would stop working with Chris. 3. I would quit myjob. 4. I would ask for a transfer. 5. I would stop spending time with Chris outside of work. Satisfaction 1. Chris meets my needs well. I 2. I am very satisfied with this relationship. T 3. This relationship is good compared to most. I 4. Given the situation described, I would wish I hadn’t gotten in this relationship. I 5. This relationship meets my expectations for a fiiendship. I 6. I care for Chris very much. T 7. There are no problems in this relationship. it * modified from the prestudy T taken from Rubin (1970) and slightly modified i modified from Hendrick (1988) 87 APPENDIX D Table 1 Means and Standard Deviations for Realism and Relationship foundation/ Relationship Dimension/ Equity Level ANO VA. Mean Std. Deviation Equitable Friend Task 6.16 0.89 Social 6.20 1.02 Total 6.18 0.95 C oworker Task 6.29 0.81 Social 6.26 0.83 Total 6.27 0.82 Total Task 6.22 0.85 Social 6.22 0.93 Total 6.22 0.89 Underreward Friend Task 5.58 0.94 Social 5.78 1.12 Total 5.67 1.03 C oworker Task 5.70 1.28 Social 5.74 0.96 Total 5.72 1.14 Total Task 5.64 1.1 1 Social 5.76 1.04 Total 5.69 1.08 88 Table l (cont’d). Mean Std. Deviation Overreward Friend Task 5 .66 1.08 Social 5.61 0.98 Total 5.64 1.03 Coworker Task 5.57 1.21 Social 5.70 0.88 Total 5.63 1.07 Total Task 5.62 1.14 Social 5.65 0.93 Total 5.63 1.04 89 Table 2 Means and Standard Deviations for Realism and Gender ANOVA. Mean Std. Deviation Equitable Male 6.25 0.84 Female 6.20 0.92 Underreward Male 5.60 1.16 Female 5.75 1.03 Overreward Male 5 .49 1.19 Female 5 .72 0.94 90 Table 3 Means and Standard Deviations for Realism and Order ANO VA. U stands for underreward, 0 stands for overreward, and E stands for equitable. Mean Standard Deviation Equitable UOE 6.33 0.81 UEO 6.50 0.54 EUO 5.98 1.45 EOU 6.21 0.82 OUE 6.19 0.52 OEU 6.1 1 0.93 Underreward UOE 5.53 1.27 UEO 6.00 0.91 EUO 5.89 0.71 EOU 6.00 0.84 OUE 5.43 1.14 OEU 5.38 1.27 Overreward UOE 5.58 1.08 UEO 5.64 1.27 EUO 5.55 1.07 EOU 5.79 0.92 OUE 5.51 0.95 OEU 5.72 1.00 91 Table 4 Means and Standard Deviations for Fairness and Relationship foundation/ Relationship Dimension/ Equity Level ANO VA. Mean Std. Deviation Equitable Friend Task 5.85 0.99 Social 6.08 1.11 Total 5.96 1.05 C oworker Task 6.01 1.07 Social 6.23 0.89 Total 6.1 l 0.99 Total Task 5 .92 1.03 Social 6.15 1.01 Total 6.03 1.02 Underreward Friend Task 2.93 l . l 3 Social 3.08 1.11 Total 3 .00 1.12 Coworker Task 3.07 0.97 Social 2.90 0.90 Total 2.99 0.94 Total Task 3.00 1.06 Social 3.00 1.02 Total 3.00 1.03 92 Table 4 (cont’d). Mean Std. Deviation Overreward Friend Task 3 .02 1 . l 2 Social 3.12 0.93 Total 3.06 1.03 Coworker Task 2.95 0.90 Social 2.89 0.74 Total 2.92 0.82 Total Task 2.99 l .02 Social 3.01 0.85 Total 3 .00 0.94 93 Table 5 Means and Standard Deviations for Fairness and Gender ANO VA. Mean Std. Deviation Equitable Male 5.98 1.10 Female 6.06 0.98 Underreward Male 3.05 1.07 Female 2.97 1.02 Overreward Male 3.09 0.94 Female 2.94 0.94 94 Table 6 Means and Standard Deviations for Fairness and Order ANO VA. U stands for underreward, 0 stands for overreward, and E stands for equitable. Mean Standard Deviation Equitable UOE 5.97 1.08 UEO 6.24 1.00 EUO 5.94 0.98 EOU 5.90 1.03 OUE 6.20 0.82 OEU 5.92 1.19 Underreward UOE 3.22 1.08 UEO 3.1 1 1.15 EUO 2.61 0.99 EOU 3.09 1.02 OUE 3.05 0.88 OEU 2.89 1.04 Overreward UOE 3.32 0.75 UEO 2.65 1.01 EUO 2.64 1.01 EOU 2.98 0.85 OUE 3.12 0.79 OEU 3.22 1.03 95 Table 7 Means and standard deviations for change own behavior, attempt to change other '5 behavior, and liking change (categorized by equity level and relationship dimension). Mean Std. Deviation Overbenefit Task Change Own 5.58 1.09 Behavior Changing other’s 3.23 1.40 behavior Liking Change .19 1.00 Overbenefit Social Change Own 5.06 1.20 Behavior Changing other’s 3.06 1.26 behavior Liking Change .72 1.17 Underbenefit Task Change Own 4.01 1.00 Behavior Changing other’s 4.43 1.02 behavior Liking Change 2.33 1.26 Underbenefit Social Change Own 4.13 1.01 Behavior Changing other’s 3.89 1.12 behavior Liking Change 1.59 1.25 96 Table 8 Zero—order correlations for change own behavior, attempt to change other ’5 behavior, and liking change (categorized by equity level and relationship dimension). Change Own Changing Liking Behavior other’s behavior Change Over Task Change Own benefit Behavior Changing other’s .12 behavior Liking Change -.11 .30** Over Social Change Own benefit Behavior Changing other’s -.05 behavior Liking Change .14 -.24* Under Task Change Own benefit Behavior Changing other’s .31 ** behavior Liking Change -.03 .34** Under Social Change Own benefit Behavior Changing other’s .52** behavior Liking Change .24* .24* * Correlation is significant at the .05 level. ** Correlation is significant at the .01 level. (Dfs ranging from 72 — 79.) 97 Table 9 Means and Standard Deviations for Type of C hange/ Gender/ Equity level/ Relationship dimension ANO VA. Type/ Change Gender Dimension Equity Mean Std. Deviation Change own Male Social Overreward 4.51 1.21 behavior Underreward 4.3 l .81 Total 4.42 1.04 Task Overreward 5.3 1 1 .28 Underreward 4.23 .76 Total 4.78 1.18 Total Overreward 4.95 1 .30 Underreward 4.26 .78 Total 4.61 1.13 Female Social Overreward 5.41 1.04 Underreward 4.04 1 .09 Total 4.65 1.27 Task Overreward 5.84 .83 Underreward 3.81 l . l 3 Total 4.83 1.42 Total Overreward 5.62 .96 Underreward 3 .94 l .1 1 Total 4.73 1.34 Total Social Overreward 5 .06 l .20 Underreward 4.13 l .01 Total 4.57 1.19 Task Overreward 5.60 1 .08 Underreward 4.00 l .00 Total 4.80 1.31 Total Overreward 5.34 1.16 Underreward 4.06 l .01 Total 4.69 1.26 98 Table 9 (cont’d). Type/ Change Gender Dimension Equity Mean Std. Deviation Partner Male Social Overreward 3 .46 1 .30 change Underreward 4.05 .87 Total 3.74 l .14 Task Overreward 3 .39 1 .42 Underreward 4.60 .98 Total 3 .99 l .36 Total Overreward 3.42 1 .36 Underreward 4.37 .97 Total 3 .88 1 .27 Female Social Overreward 2.83 1.20 Underreward 3.81 1 .23 Total 3 .38 1 .3 1 Task Overreward 3 .09 1 .3 8 Underreward 4.30 1 .04 Total 3.70 1 .35 Total Overreward 2.96 l .29 Underreward 4.02 1 .17 Total 3 .52 1 .33 Total Social Overreward 3.06 1 .26 Underreward 3.89 1 . 12 Total 3.49 1 .25 Task Overreward 3 .23 1 .40 Underreward 4.43 1 .02 Total 3.83 1 .36 Total Overreward 3.15 1.33 Underreward 4. 15 1 . 10 Total 3.66 l .32 99 Table 10 Means and Standard Deviations for participant gender/ Chris 'gender/ relationship dimension/ equity level ANO VA on Liking. Participant Chris’ Task/ Over/ Under Mean Standard Gender Gender Social Reward Deviation Male Male Social Overreward .60 .98 Underreward l .88 1 . 14 Total 1.21 1.23 Task Overreward .03 .80 Underreward 2.07 .76 Total 1.09 1.28 Total Overreward .29 .92 Underreward l .99 .92 Total 1.14 1.26 Female Social Overreward .54 .94 Underreward .65 1 .34 Total .58 1.04 Task Overreward .58 .92 Underreward 1 .69 1 .60 Total .98 1.27 Total Overreward .56 .90 Underreward 1 .23 1 .50 Total .80 1.17 Total Social Overreward .58 .96 Underreward 1 .66 1 .24 Total 1.07 1.21 Task Overreward . 17 .84 Underreward 2.01 .91 Total 1.07 1.27 Total Overreward .36 .91 Underreward 1 .87 1 .06 Total 1.07 1.24 Female Male Social Overreward .66 1.37 Underreward 1.40 1 .32 Total 1.04 1.38 Task Overreward .27 1 .24 Underreward 2.59 1 .39 Total 1.48 1.75 Total Overreward .46 1.30 Underreward 2.01 1 .47 Total 1.27 1.59 100 Table 10 (cont’d). Participant Chris’ Task/ Over/ Under Mean Standard Gender Gender Social Reward Deviation Female Female Social Overreward 1.20 1.12 Underreward 1 .96 l . 19 Total 1.65 1.20 Female Task Overreward -.06 .79 Underreward 2.64 1 .63 Total 1.29 1.86 Total Overreward .70 l . 17 Underreward 2. 1 8 l .35 Total 1.52 1.47 Total Social Overreward .84 1.30 Underreward l .62 1.29 Total 1.27 1.34 Task Overreward . 19 1 . l4 Underreward 2.60 l .42 Total 1.44 1.77 Total Overreward .53 1.26 Underreward 2.07 1 .43 Total 1.35 1.56 Total Male Social Overreward .63 1.21 Underreward 1 .58 1.26 Total 1.11 1.32 Task Overreward . l 6 1 .06 Underreward 2.36 l . 1 7 Total 1.31 1.56 Total Overreward .38 1.15 Underreward 2.01 1 .27 Total 1.22 1.46 Female Social Overreward 1.00 1.09 Underreward 1 .75 l .28 Total 1.40 1.24 Task Overreward .24 .89 Underreward 2.32 1 .63 Total 1.16 1.63 Total Overreward .64 1.06 Underreward l .97 1.43 Total 1.30 1.41 101 Table 10 (cont’d). Participant Chris’ Task/ Over/ Under Mean Standard Gender Gender Social Reward Deviation Total Total Social Overreward .74 1 . 1 8 Underreward 1 .64 1 .26 Total 1.20 1.30 Task Overreward . 1 8 1 .01 Underreward 2.35 1 .26 Total 1.27 1.57 Total Overreward .45 l . 13 Underreward 1.99 1 .3 1 Total 1.24 1.44 102 Table 11 Means and Standard Deviations for participant gender/ Chris 'gender/ relationship dimension/ equity level ANO VA on changing own behavior. Participant Chris’ Task/ Over/ Under Mean Standard Gender Gender Social Reward Deviation Male Male Social Overreward 4.60 1.34 Underreward 4.20 .83 Total 4.41 l . l 3 Task Overreward 5 .43 1. 1 8 Underreward 4.21 .82 Total 4.80 1 . 1 7 Total Overreward 5.06 1.3 l Underreward 4.21 .81 Total 4.63 1 . 1 7 Female Social Overreward 4.25 .71 Underreward 4.69 .75 Total 4.41 .72 Task Overreward 4.94 1 .55 Underreward 4.30 .48 Total 4.71 1 .28 Total Overreward 4.64 1.26 Underreward 4.47 .61 Total 4.5 8 l .06 Total Social Overreward 4.51 1.21 Underreward 4.28 .82 Total 4.41 1 .05 Task Overreward 5.3 1 1 .28 Underreward 4.23 .77 Total 4.78 1 . 1 9 Total Overreward 4.95 1.30 Underreward 4.25 .79 Total 4.62 l . 14 Female Male Social Overreward 5.33 1.00 Underreward 3.98 l . 1 8 Total 4.62 l .28 Task Overreward 5.82 .89 Underreward 3 .67 1.18 Total 4.71 l .50 Total Overreward 5.58 .97 Underreward 3.83 1 . 1 8 Total 4.67 1 .50 Female Social Overreward 5.57 1.14 Underreward 4. 10 l .01 Total 4. 71 1 .29 103 Table 1 1 (cont’d). Participant Chris’ Task/ Over/ Under Mean Standard Gender Gender Social Reward Deviation Female Female Task Overreward 5.83 .69 Underreward 4.28 .82 Total 5.05 1.08 Total Overreward 5.67 .98 Underreward 4. l 5 .95 Total 4.83 1.22 Total Social Overreward 5 .41 1.04 Underreward 4.03 1 .1 1 Total 4.65 1.28 Task Overreward 5.82 .84 Underreward 3 .8 l 1 . 13 Total 4.79 1.42 Total Overreward 5.61 .97 Underreward 3 .93 1 . l 2 Total 4.72 1.34 Total Male Social Overreward 5.02 1.20 Underreward 4.06 1 .06 Total 4.54 1.22 Task Overreward 5 .64 1 .04 Underreward 3.92 1 .06 Total 4.75 1.36 Total Overreward 5.35 1 . l 6 Underreward 3.99 1 .06 Total 4.65 1.30 Female Social Overreward 5.15 1.19 Underreward 4. l 9 .99 Total 4.64 1.18 Task Overreward 5.41 1.22 Underreward 4.28 .7 1 Total 4.91 1.16 Total Overreward 5.27 1 .20 Underreward 4.23 .88 Total 4.75 1.17 Total Social Overreward 5 .06 1 .19 Underreward 4.1 l 1 .03 Total 4.57 1.21 Task Overreward 5 .59 l .09 Underreward 3 .99 l .01 Total 4.79 1.32 Total Overreward 5 .33 1 . 16 Underreward 4.05 1 .02 Total 4.68 1.26 104 Table 12 Means and Standard Deviations for participant gender/ Chris 'gender/ relationship dirnension/ equity level ANO VA on changing one ’s partner '5 behavior. Participant Chris’ Task/ Over/ Under Mean Standard Gender Gender Social Reward Deviation Male Male Social Overreward 3 .39 1.39 Underreward 4.05 .92 Total 3.71 1.22 Task Overreward 3. 19 1 .42 Underreward 4.8 1 .83 Total 4.03 1.40 Total Overreward 3.28 1.39 Underreward 4.5 1 .94 Total 3.89 1.33 Female Social Overreward 3.68 1.03 Underreward 4. 13 .97 Total 3.84 .98 Task Overreward 3.97 1 .34 Underreward 3.60 1 .26 Total 3.84 1.27 Total Overreward 3.84 1.1 8 Underreward 3.83 1.10 Total 3.84 1.13 Total Social Overreward 3.46 1.30 Underreward 4.07 .91 Total 3.74 1.17 Task Overreward 3.39 l .42 Underreward 4.63 .99 Total 3.99 1.37 Total Overreward 3.42 l .36 Underreward 4.40 .99 Total 3.88 1.29 Female Male Social Overreward 2.91 1.33 Underreward 3.94 l .29 Total 3.45 1.40 Task Overreward 3. 1 8 l .36 Underreward 4.33 1 .03 Total 3.77 1.33 Total Overreward 3.05 l .34 Underreward 4.14 1.17 Total 3.61 1.37 Female Social Overreward 2.70 .95 Underreward 3.57 l . 1 8 Total 3.21 1.16 105 Table 12 (cont’d). Participant Chris’ Task/ Over/ Under Mean Standard Gender Gender Social Reward Deviation Female Female Task Overreward 3.25 1.29 Underreward 4.1 8 1 .09 Total 3.71 1.25 Total Overreward 2.92 1.10 Underreward 3 .77 1.17 Total 3.39 1.21 Total Social Overreward 2.84 1 .20 Underreward 3.79 l .25 Total 3.36 1.31 Task Overreward 3.20 l .33 Underreward 4.30 1 .04 Total 3.76 1.30 Total Overreward 3.01 l .27 Underreward 4.02 1 . l 8 Total 3.54 1.32 Total Male Social Overreward 3.1 1 1.36 Underreward 3.98 1 .16 Total 3.55 1.33 Task Overreward 3.18 1 .38 Underreward 4.55 .97 Total 3.89 1.36 Total Overreward 3. 15 1 .36 Underreward 4.29 1 .09 Total 3.73 1.36 Female Social Overreward 3.01 1.06 Underreward 3.66 1 .l 5 Total 3.36 1.14 Task Overreward 3.59 1 .33 Underreward 3.98 l . 14 Total 3.76 1.24 Total Overreward 3 .28 1 .21 Underreward 3.78 1 . 14 Total 3.53 1.20 Total Social Overreward 3 .08 1.27 Underreward 3.88 1.16 Total 3.49 1.27 Task Overreward 3.29 1.37 Underreward 4.44 l .02 Total 3.86 1.33 Total Overreward 3.19 l .32 Underreward 4.16 1.12 Total 3.68 1.32 106 l 6. l . .1 . I 4 , 4'06 4 5 Iowr 1 Cl under} 3 l i 2 . | T 1 - own change partner change Figure 1 Interaction between type of change and equity level. 107 ’ l i e 5.60 ' 5 - 4.43 _ 4 4004'” Dtask 3.23 3.06 I social 3 a 2 a 1 4.. l I over under over under I i own change 3 partner change I Figure 2 Means for interaction between equity level, relationship dimension, and type of change. 108 . 6 q 5 ~ 4.62 4-73 3.89 :1 Male 4 ‘ 3.52 I Female 3 “t 2 _ 1 4 -L- . Own Change Partner Change L____ Figure 3 Interaction between type of change and gender. 109 4.95 4.26 3.42 2.96 I female. overreward underreward overreward underreward own change other change Figure 4 Interaction between type of change, gender, and equity level. 110 3 [3 Task 2-33 I Social i. 1.59 1 I 0.72 f i 0.19 - l 0 r " I Overreward Underreward Figure 5 Interaction between task/social and over/underreward for change in liking. 111 REFERENCES Adams, J. S. (1965). Inequity in social exchange. In Berkowitz, L. (Ed.), Advances in Experimental Social Psychology: Vol. 2. New York: Academic Press. Austin, W., & Walster, E. (1975). Equity with the world: The trans-relational effects of equity and inequity. Sociometry, 38, 474-496. Azj en, I. & F ishbein, M. (1980). Understanding attitudes and predicting social behavior. Englewood-Cliffs, NJ: Prentice-Hall. Baxter, L. A. (1988). A dialectical perspective on communication strategies in relationship development. In S. W. Duck (Ed.), Handbook of personal relationships. New York: John Wiley & Sons. Bing, M. N., & Burroughs, S. M. (2001). The predictive and interactive effects of equity sensitivity in teamwork-oriented organizations. Journal of Organizational Behavior, 22, 271-290. Bridge, K., & Baxter, L. A. (1992). Blended relationships: Friends as work associates. Western Journal of Communication, 56, 200-225. Brockner, J. & Swap, W. C. (1976). Effects of repeated exposure and attitudinal similarity on self-disclosure and interpersonal attraction. Journal of Personality and Social Psychology, 33, 531-540. Burgess, R.L., & Nielsen, J.M. (1974). An experimental analysis of some structural determinants of equitable and inequitable exchange relations. American Sociological Review, 39, 427-443. Buunk, B. P., & Bakker, A. B. (1997). Responses to unprotected extradyadic sex by one’s partner: Testing predictions from interdependence and equity theory. The Journal of Sex Research, 34, 378-397. Canary, D. J ., & Stafford, L. (1992). Relational maintenance strategies and equity in marriage. Communication Monographs, 59, 243-267. Carrell, M. R., & Dittrich, J. E. (1978). Equity theory: The recent literature, methodological considerations, and new directions. Academy of Management Review, 3, 202-210. Clark, M. S., & Mills, J. (1979). Interpersonal attraction in exchange and communal relationships. Journal of Personality and Social Psychology, 3 7, 12-24. C osier, R. A., & Dalton, D. R. ( 1983). Equity theory and time: A reformulation. The Academy of Management Review, 8, 311-319. 112 Davis, K. E. (1985). Near and dear: Friendship and love compared. Psychology Today, 19, 22-30. Davidson, B., Balswick, J., & Halverson, C. (1983). Affective self-disclosure and marital adjustment: A test of equity theory. Journal of Marriage and the Family, 45, 93- 102. Farber, M. L. (1952). The college student as laboratory animal. American Psychologist, 7, 102. Festinger, L. (1957). A theory of cognitive dissonance. Stanford, CA: Stanford University Press. Fine, G. A. (1986). Friendships in the workplace. In V. J. Derlega & B. A. Winstead (Eds.) Friendship and Social Interaction. New York: Springer-Verlag. Fizel, J ., Krautmann, A. C., & Hadley, L. (2002). Equity and arbitration in major league baseball. Managerial and Decision Economics, 23, 427-435. Gordon, M. E., Slade, L. A., & Schmitt, N. (1986). The “science of the sophomore” revisited: From conjecture to empiricism. Academy of Management Review, I 1, 191-207. Greenberg, J. (1988). Equity and workplace status: A field experiment. Journal of Applied Psychology, 73, 606-613. Greenberg, J. (1990). Organizational justice: Yesterday, today and tomorrow. Journal of Management, 16, 399-432. Hatfield, E., Traupmann, J., Sprecher, S., Utne, M., & Hay, J. (1985). Equity and intimate relations: Recent research. In Ickes, W. (ed.), Compatible and Incompatible Relationships. Springer-Verlag, New York. Hatfield, E. Utne, M. K., & Traupmann, J. (1979). Equity theory and intimate relationships. In Burgess, R. L., & Huston, T. L. (eds), Social Exchange in Developing Relationships. Academic Press, New York. Hegtvedt, K. A. (1990). The Effects of Relationship Structure on Emotional Responses to Inequity. Social Psychology Quarterly, 53, 214-228. Hendrick, S. S. (1988). A generic measure of relationship satisfaction. Journal of Marriage and the Family, 50: 93-98. 113 Huseman, R. C., Hatfield, J. D., & Miles, E. W. (1987). A new perspective on equity theory: The equity sensitivity construct. Academy of Management Review, 12, 222-234. Katz, D., & Kahn, R. L. (1978). The social psychology of organizations. (2“. Ed.) New York: Wiley. Kim, M.-S. & Hunter, J. E. (1993). Relationships among attitudes, behavioral intentions, and behavior: A meta-analysis of past research, part 2. Communication Research, 20, 331-364. Kollock, P., Blumstein, P., & Schwartz, P. (1994). The judgment of equity in intimate relationships. Social Psychology Quarterly, 5 7, 340-351. Kram, K. E., & Isabella, L. A. (1985). Mentoring alternative: The role of peer relationships in career development. Academy of Management Journal, 28, 1 10- 132. Kruglanski, A. W. (1975). The human subject in the psychology experiment: Fact and artifact. In L. Berkowitz (ed.), Advances in Experimental Social Psychology. New York: Academic Press. Lane, I. M., & Messé, L. A. (1971). Equity and the distribution of rewards. Journal of Personality and Social Psychology, 20, 1-17. Leventhal, G. S., & Anderson, D. (1970). Self-interest and the maintenance of equity. Journal of Personality and Social Psychology, 15, 57—62. Leventhal, G. S., & Bergman, J. T. (1969). Self-depriving behavior as a response to unprofitable inequity. Journal of Experimental Social Psychology, 5, 153-171. Leventhal, G.S., Weiss, T., & Long, G. (1969). Equity, reciprocity, and reallocating rewards in the dyad. Journal of Personality and Social Psychology, 13, 300-305. Michaels, J. W., Edwards, J. N., & Acock, A. C. (1984). Satisfaction in intimate relationships as a function of inequality, inequity, and outcomes. Social Psychology Quarterly, 47, 347-357. Miles, E. W., Hatfield, J. D., & Huseman, R. C. (1989). The equity sensitivity construct: Potential implications for worker performance. Journal of Management, 15, 581- 588. Morgan, W. R., & Sawyer, J. (1979). Equality, equity, and procedural justice in social exchange. Social Psychology Quarterly, 42, 71-75. 114 Mudrack, P. E., Mason, E. S., & Stepanski, K. M. (1999). Equity sensitivity and business ethics (fairness in employee selection). Journal of Occupational and Organization, 72, 539-557. O’Neill, B. S., & Mone, M. A. (1998). Investigating equity sensitivity as a moderator of relations between self-efficacy and workplace attitudes. Journal of Applied Psychology, 83, 805-816. Patrick, S.L., & Jackson, J. J. (1991). Further examination of the equity sensitivity construct. Perceptual and Motor Skills, 73, 1091-1106. Peabody, R. L. (1962). Perceptions of organizational authority: A comparative analysis. Administrative Science Quarterly, 6, 463-482. Pritchard, R. D., Dunnette, M. D., & Jorgenson, D. O. (1972). Effects of perceptions of equity and inequity on worker performance and satisfaction. Journal of Applied Psychology Monograph, 56, 75-94. Rees, C. R. & Wechsler Segal, M. (1984). Intragroup competition, equity, and interpersonal attraction. Social Psychology Quarterly, 47, 328-336. Roloff, M. E. (1981). Interpersonal communication: The social exchange approach. Beverly Hills, CA: Sage. Ronen, S. (1986). Equity perception in multiple comparisons: A field study. Human Relations, 39, 333-346. Rubin, Z. (1970) Measurement of romantic love. Journal of Personality and Social Psychology, 16: 265-273. Saegert, S., Swap, W., & Zajonc, R. B. (1973). Exposure, context, and interpersonal attraction. Journal of Personality and Social Psychology, 25, 234-242. Shapiro, E. G. (1975). Effect of expectations of future interaction on reward allocations in dyads: equity or equality. Journal of Personality and Social Psychology, 31, 873-880. Sias, P. M., & Cahill, D. J. (1998). From coworkers to friends: the development of peer friendship in the workplace. Western Journal of Communication, 62, 273-299. Sprecher, S. (1992). How men and women expect to feel and behave in response to inequity in close relationships. Social Psychology Quarterly, 55, 57-69. Taynor, J. & Deaux, K. (1973). When women are more deserving than men: Equity, attribution, and perceived sex differences. Journal of Personality and Social Psychology, 28, 360-367. 115 Taynor, J. & Deaux, K. (1975). Equity and perceived sex differences: Role behavior as defined by the task, the mode, and the actor. Journal of Personality and Social Psychology, 32, 381-390. Van Yperen, N. W., & Buunk, B. P. (1991). Equity theory and exchange and communal orientation from a cross-national perspective. The Journal of Social Psychology, 131, 5-20. Verbrugge, L. M. (1979). Multiplexity in adult fiiendships. Social Forces, 57, 1286- 1309. Vogl-Bauer, S., Kalbfleisch, P. J ., & Beatty, M. J. (1998). Perceived equity, satisfaction, and relational maintenance strategies in parent-adolescent dyads. Journal of Youth and Adolescence, 28, 27-49. Wahba, M. A. (1971). Equity theory as a predictor of payoff apportionment among partners in coalition formations. Psychonomic Science, 24, 177-180. Walster, E., Berscheid, E., & Walster, G.W. (1973). New directions in equity research. Journal of Personality and Social Psychology, 25 , 151-176. Walster, E., Walster, G. W., & Berscheid, E. ( 1978). Equity: Theory and research. Allyn and Bacon, Boston, MA. Winn, K. 1., Crawford, D. W., & Fischer, J. (1991). Equity and commitment in romance versus friendship. Journal of Social Behavior and Personality, 6, 301-314. Winstead, B. A., Derlega, V. J ., & Montgomery, M. J. (1995). The quality of fiiendships at work and job satisfaction. Journal of Social and Personal Relationships, 12, 199-215. Wright, P. H. (1985). The acquaintance description form. In S. W. Duck & D. Perlman (Eds), Sage Series in Personal Relationships, 1, 39-62. London: Sage. 116 llitigiiijiiiiii