Equity sensitivity theory: A test of responses to two types of under-reward situations

Equity Sensitivity (Huseman et aL, 1987) has proven to be a refinement of the original Equity Theory (Adams, 1963, 1965). However, if Equity Sensitivity (Huseman et aL, 1987) is to prove more useful than the original Equity Theory (Adams, 1963, 1965) it must be more predictive and discriminant with

The purpose of this study is to take a closer look at the efficacy of Equity Sensitivity Theory (Huseman et al., 1987). More specifically this study focuses on the ability of Equity Sensitivity to discriminate between the responses of three different classifications of individuals posited by the theory (Benevolents, Equity Sensitives and Entitleds) in response to two types of under-reward situations. Previous research has yet to examine the differences in how the three groups respond to situations in which the "under-reward" is the result of being paid the same as the "comparison-- other" for doing more work. With globalization and hyper-competition characterizing today's business environment, firms are being forced to expect ever higher levels of productivity from their work force while simultaneously maintaining cost-controlling reward systems. Thus, this new type of under-reward situation may be more applicable to the current business environment considering the downsizing, flattening and job enlargement that have occurred over the past decade.

The following section provides a brief summary of the existing literature on this topic and identifies some weaknesses in the prior research that are addressed in the study described in the remainder of this article.

Relevant Literature

Since its origins in the 1960s Equity Theory (Adams, 1963, 1965) held forth the promise of helping to explain how employees respond to situations in which they perceive they are being rewarded more or less favorably in comparison to a referent doing similar work. Shortly after its inception, Weick (1966) deemed it to be one of the most useful existing organizational behavior theories. Subsequent reviews concluded that the empirical evidence supporting Equity Theory was generally strong (Greenberg, 1982; Mowday, 1991), especially with regards to how workers respond to under-reward situations.

Equity Theory (Adams, 1963, 1965) proposed that subjects respond to under-reward situations in various ways in an attempt to bring their equity ratio back into balance. For example, subjects may choose a behave ioral response to help reduce their feelings of inequity. They may respond in such ways as reducing their inputs (i.e., not put forth as much effort) or increasing their outcomes (i.e., ask for a raise). Subjects may instead use a cognitive response to reduce feelings of inequity such as selecting another person to use as their referent. Ultimately the subject may choose to exit the situation by deciding to transfer or quit the organization.

Although previous Equity Theory research has concluded that underrewarded subjects generally respond in a manner that is consistent with classic Equity Theory, it is not easy to predict which option they will select to bring their equity ratio into balance (Greenberg, 1990). This lack of specificity regarding what responses individuals experiencing inequity are likely to have is a serious shortcoming of the original Equity Theory (Furby, 1986). As such, the original Equity Theory eventually fell out of favor (Miner, 1984; Greenberg, 1987, 1990) due in part to this inability to predict exactly how individuals would respond to an under-reward situation (e.g., lower their inputs, attempt to raise their outcomes, cognitively justifying the situation, decide to leave the organization). This lack of predictive ability of Equity Theory makes it much less useful to practitioners such as managers and human resource professionals who would greatly benefit if they could accurately predict the reactions that their employees would have to different inequitable situations.

Accordingly, research on the topic of Equity Theory moved off in another direction. Inspired by legal research, the procedural justice stream of research (Thibaut and Walker, 1975, 1978) began to focus more on the processes and procedures of how pay and recognition are determined, rather than the reactions that individuals have to them (Greenberg, 1987). Equity Theory research became less popular and eventually withered away (Greenberg, 1990).

Interest in organizational justice and equity has experienced a resurgence over the past decade (Ambrose and Kulik, 1999; Bing and Burroughs, 2001; Chan et al., 1997; George, 1994; Glass and Wood, 1996; Goodwin, 1990; Harder, 1991; Harder, 1992; Huseman and Hatfield, 1990; Lapidus and Pinkerton, 1995; Moorman, 1991; Mui, 1995; Sauley and Bedeian, 2000; Van Wijck, 1994). This rekindling of interest in equity has been spurred in part by an extension of the original Equity Theory to include individual differences. This more recent approach has been spurred by the construct of Equity Sensitivity (Hartman and Villere, 1990; Huseman et al., 1985; Huseman et aL, 1987; King et al., 1993; Patrick and Jackson, 1991). This construct posits that individuals can be categorized into three groups: Equity Sensitives, Benevolents and Entitleds. This more recent extension of original Equity Theory runs the same risk of being deemed impractical if it also is not shown to be predictive.

According to the latest view, one group labeled Equity Sensitives fit the classic Equity Theory propositions. Equity Sensitives prefer to be in a state of equity with regards to the outcomes they receive for the amount of inputs they expend when compared to someone doing similar work. The original propositions of Equity Theory apply to this group. If an Equity Sensitive's ratio of outcomes to inputs is out of balance with their referent other, the subject will be motivated to do things to get their ratio back into balance.

Equity Sensitivity proposes two other groups - Benevolents and Entitleds. Benevolents are more tolerant of situations in which they are being under-rewarded. While they do not seek to be under-rewarded, they are assumed to be less likely to respond (at least overtly) when they are placed in an under-reward situation. Entitleds are posited to experience less dissonance when they are over-rewarded and more dissatisfaction when under-rewarded. As such, they are assumed to be more likely than the other groups to respond overtly to an over-reward situation.

An instrument (King and Miles, 1994) to measure Equity Sensitivity has been developed and validated using five different samples (n = 2,399). The Equity Sensitivity Instrument (ESI) has been shown to be discriminantly and convergently valid. Results suggest that equity sensitivity is unique from other variables and constructs including age, gender, educational level, social desirability, self-- esteem, locus of control, Machiavellianism, pro-Protestant and non-- Protestant work ethic, input-outcome orientation, exchange ideology, altruism, organizational commitment, job satisfaction, supervisor satisfaction, propensity to turnover and perceptions of pay justice. This instrument was used in the research detailed in this article.

Some initial studies have been done to attempt to validate significant differences between Benevolents, Equity Sensitives and Entitleds. For example, Huseman et aL (1985) examined job satisfaction as a dependent variable in the context of equity sensitivity. They found that in response to an under-reward situation, Entitleds report significantly lower levels of satisfaction than Equity Sensitives or Benevolents. Conversely, Benevolents report significantly higher levels of satisfaction than the other two groups.

Miles et al. (1989) tested for differences in the three groups with respect to inputs, outcomes and preferred outcome/input ratios. With regards to inputs they found that Benevolent undergraduate students were willing to code more questionnaires or complete more interviews than Equity Sensitives or Entitleds. Interestingly, they did not find significant differences between the three groups in terms of outcomes (what they perceived as a fair wage), although they did find that Benevolents prefer lower outcome/input ratios than Equity Sensitives or Entitleds. A shortcoming regarding this research is that it did not investigate the entire gamut of responses that individuals may have to under-reward situations. It focused solely on a single outcome (pay) and single input (quantity of work). The present study addresses this shortcoming by investigating a much broader gamut of potential responses to under-reward.

King et al. (1993) found that Benevolent undergraduate students experience less distress than Entitleds when facing either under-compensation or over-compensation scenarios. They also found that Entitleds placed significantly more importance on pay and Benevolents place more importance on work characteristics when asked to distribute 100 points on the Job Descriptive Index. Furthermore, Miles et al. (1994) found that Entitleds tend to place a greater emphasis on extrinsic tangible rewards (e.g., pay), whereas Benevolents are more focused on intrinsic intangible rewards (e.g., a sense of accomplishment) when asked to rate the importance of twenty of the most common outcomes from work.

O'Neill and Mone (1998) examined equity sensitivity as a potential moderator in the relationship between self-efficacy and workplace attitudes including job satisfaction, organizational commitment, and intent to leave. They found that Benevolents exhibit a negative relationship between self-efficacy and job satisfaction and a positive relationship between self-efficacy and intent to leave. Equity sensitivity did not exhibit a moderating effect on the relationship between self-efficacy and organizational commitment.

Patrick and Jackson (1991) surveyed groups of undergraduate students and fast-food workers to see how the reactions of Benevolents, Equity Sensitives, and Entitleds differed after reading scenarios that put them in inequitable situations. Their resuits indicate that Benevolents and Equity Sensitives are significantly more likely to alter their inputs and their outcomes than Entitleds when facing an over-reward situation. Conversely, Equity Sensitives and Entitleds are more likely than Benevolents to act on their comparison other by telling them to put forth less effort in response to an over-reward situation.

Patrick and Jackson (1991) found even less distinction between the three groups in how they responded to under-reward situations. They tested how subjects would respond to under-reward situations in which someone else was being paid more for doing the same amount of work. In fact, no statistically significant differences existed between the three groups in two of three under-reward scenarios they tested. In just one scenario and in only one specific type of response was a statistically significant difference exhibited - the Benevolent and Equity Sensitive groups were significantly more likely to change their comparison other than were the Entitleds in response to an under-reward, salary scenario. The results of this study are troubling with regards to the ability of Equity Sensitivity to discriminate between the three groups and their reactions to a condition of under-reward. No clear distinction between the three groups was evident.

Patrick and Jackson (1991) also failed to consider another potentially important type of under-reward situation - a situation in which someone is being paid the same amount as their referent but doing more work. In this age of corporate re-engineering, downsizing and organizational flattening this type of under-reward situation is likely to be just as prevalent as the situation in which the referent is being paid more for doing the same amount of work. We address this deficiency in our study by testing this type of under-reward situation.

Considering the importance of the issue of differentiation between the three groups' responses to under-reward situations and the implications this could have on practitioners we find it disturbing that Patrick and Jackson (1991) did not find stronger effects of Equity Sensitivity. Patrick and Jackson (1991) called for further research into this potential lack of predictive capability of Equity Sensitivity Theory. Likewise, as Ambrose and Kulik concluded in their recent review of motivation research from the past decade, ". . . to date there has been no research demonstrating that members of the three groups (Benevolents, Equity Sensitives, and Entitleds) respond differently to inequality situations in terms of motivation or behavior" (1999: 243). As such, the construct of Equity Sensitivity runs the risk of falling out of favor, just as original Equity Theory did in the 1980s. The current study attempts to address this shortcoming.

Another important conclusion regarding Equity Theory is that research has shown that subjects who perceive they are being over-paid are not very likely to respond with behaviors that attempt to bring their equity ratio into balance (e.g., work harder, give some of their pay back) (Goodman and Friedman, 1971). Over-rewarded individuals are much more likely to mentally justify their situation or just not even perceive that they are being over-rewarded. Based on these findings we decided to focus solely on under-reward situations in the current research. This decision had additional benefits regarding size of the research instrument and length of time to administer, and served to increase the focus of the study.

In summary, if Equity Sensitivity is to be considered more valid and useful than original Equity Theory, empirical research must demonstrate significant differences in the way the three groups (Benevolents, Equity Sensitives, and Entitleds) intend to respond to inequitable situations. While previous research (Patrick and Jackson, 1991) has found significant differences in the responses of the three groups to over-reward situations, the significant differences to an under-reward situation were much less pronounced. Only one of the six potential reactions to an under-reward condition resulted in a significant difference - Benevolents and Equity Sensitives were more likely to change their comparison others than were the Entitleds. Considering that the reactions that employees may have to under-payment are generally negative for the organization (e.g., reduced productivity, increased turnover), the lack of differentiation exhibited thus far between the groups by Equity Sensitivity has been disappointing. If more significant differences exist between the groups it could have significant implications for organizations. Thus, further research on the issue of differences between the three groups when facing under-reward situations is warranted.

Furthermore, previous research has investigated only one type of under-reward situation - less pay for the same amount of work as the comparison other. Another, and potentially more significant type of underreward, same pay for doing more work than a comparison other, has yet to be explored. The study presented here also addresses this important deficiency in the Equity Sensitivity research.

Hypotheses

Based on original Equity Theory (Adams, 1963, 1965), a state of underreward inequity exists when subjects' perceived outcome/input ratio is less than the ratio of a referent to which they compare themselves. There are a number of possible responses that subjects can have to attempt to rectify such an inequitable situation and bring their ratio into a state of balance with their referent (Adams, 1965; Carrell and Dittrich, 1978; Miner, 1980; Mowday, 1991). The subjects may attempt to reduce their inputs. For example, they may decide to reduce their effort and not work as hard. Alternately the subjects could attempt to increase their outcomes. They may ask their supervisor for a raise. The subjects might attempt to increase the inputs of their referent. For instance, subjects may try to make the other person work harder by not cooperating with them when they need help. Subjects could attempt to reduce the outcomes of their referent. They might tell the supervisor something about the referent that could result in getting the other person's pay reduced. Subjects may attempt to mentally justify why they are being under-rewarded. For example, subjects may begin to question their own effort or qualifications. Conversely, subjects might attempt to mentally justify why the referent is being over-rewarded. Perhaps the pay structure is such that everyone is paid the same amount no matter how little they produce. Alternatively, subjects could actually change their referent and begin comparing themselves with someone else. Another option is for some subjects to transfer to another part of the organization. Finally, some subjects may decide to resign from the organization.

According to Equity Sensitivity (Huseman et al., 1987), Benevolents are proposed to be more tolerant of under-reward situations than either Equity Sensitives or Entitleds (King et aL, 1993). Therefore, Benevolents should logically be less likely to take actions that attempt to bring their outcome/input ratio into balance with their referent others.

This study set forth to empirically examine this proposition by testing the following hypotheses regarding under-reward situations. When faced with an inequitable situation, there should be significant differences in the responses of the three groups (Entitled, Equity Sensitive, and Benevolents). More specifically, when facing an under-reward situation, we hypothesize that Entitleds will be more likely than equity Sensitives and equity Sensitives will be more likely than Benevolents to:

* Hypothesis 1: choose to reduce their inputs.

* Hypothesis 2: attempt to increase their outcomes (pay).

* Hypothesis 3: attempt to increase the inputs of their referent other.

* Hypothesis 4: choose to attempt to decrease the outcomes (pay) of their referent other.

* Hypothesis 5: mentally justify why they are being under-rewarded.

* Hypothesis 6: mentally justify why their referent other is being over-rewarded.

* Hypothesis 7: change their referent other.

* Hypothesis 8: choose to transfer to another part of the organization.

* Hypothesis 9: to leave the organization (quit).

Method

Participants and Procedure

A survey was administered to a convenience sample of 240 business students at an urban public university. Participation was voluntary and anonymity was promised. Two hundred and three students participated for a response rate of 84.6%. The sample was split almost evenly between males (48.3%) and females (51.7%). Of this sample, 78.3% of the participants were juniors, while the remaining 21.7% were seniors. The participants' age and work experience were greater than is typical of traditional college undergraduates. Their ages ranged from 19.5 to 51.2 years, with a mean age of 23.7 years and a standard deviation of 5.4 years. Sixty-four and one-half percent of the respondents had at least 6 months of fulltime work experience (with a mean of 3.2 years). Eighty-six and one tenth percent had at least 6 months of parttime work experience (with a mean of 3.3 years).

Measures

Equity Sensitivity. Each participant's equity sensitivity was measured using the Equity Sensitivity Instrument (ESI) (King and Miles, 1994). Previous research studies using the ESI have reported coefficient alphas ranging from .77 to .88 (King and Miles, 1994; Patrick and Jackson, 1991) and a test-retest reliability of .80 (Miles et al., 1989). For this study, the Cronbach's alpha was 0.79, which is consistent with the values found in the other studies cited above.

Previous research has established the convention that the sample should be split into 3 groups (Benevolents, Equity Sensitives, Entitleds) using the decision rule of plus and minus one-half standard deviation from mean ESI score (King and Miles, 1994; King et al., 1993; Miles et aL, 1994; Miles et al., 1989). This decision rule is used because the unique characteristics of any particular sample can influence responses to the ESI. For example, differing levels of social desirability, varied organizational contexts, age homogeneity, and other variables may influence responses (King and Miles, 1994). We adhered to this decision rule to trichotomize our sample into the three groups. The mean equity sensitivity score of our sample was 29.14, with a standard deviation of 6.35 and a range of 10 to 48. Using the plus and minus one-half standard deviation decision rule, our sample was split into 55 Benevolents, 92 Equity Sensitives and 55 Entitleds.

Under-reward Scenarios. A hypothetical under-reward scenario was developed and presented to the respondents patterned after those used by Austin et aL (1980), King et al. (1993), and Patrick and Jackson (1991). This scenario put the respondents in a hypothetical under-reward situation in which they were student workers assigned to code questionnaires for a professor. The student was doing the same amount of work (coding of questionnaires per hour) as their referent (another student worker), but discovered they were being paid less per hour.

A second type of under-reward scenario that has not been examined in previous equity sensitivity research was also included. In this scenario student workers were being paid the same hourly rate as their referent, but they were coding more questionnaires per hour. In other words, they were being under-rewarded by working harder than their referent but receiving the same pay.

Responses to the Under-reward Situations. After reading each scenario, the respondents were asked to complete nine Likert-type scales regarding how they would be most likely respond to each type of under-reward situation. These scales covered all nine of the hypothesized responses that people may use in an attempt to rebalance their equity ratios discussed previously in the hypothesis section.

Results

One-way analyses of variances were computed and are summarized for both scenarios and all nine hypotheses in Tables 1 and 2. For those analyses that resulted in significant findings, a further analysis using the Scheffe Post Hoc test was used and the results are summarized in Table 3.

Hypothesis 1, regarding intent to work less hard, was partially supported. The analysis of variance was significant (p < 0.001) for the first scenario (less pay for the same amount of work) and also significant (p < 0.05) for the second scenario (same pay for more work). Post hoc tests (Scheffe) indicated significant differences between the average response levels of the Entitled and Sensitive groups and Benevolent and Entitled groups for the first scenario. Only the Entitled and Benevolent groups were significantly different in the second scenario.

Hypothesis 2 was not supported. There were no significant differences in either scenario between the responses of the different groups rearding their efforts to attempt to increase their pay when faced with an inequitable situation regardless of their equity orientation.

Hypothesis 3 was partially supported. The analysis of variance for scenario 2 (more work for less pay) was significant (p < 0.05) with a post hoc test (Scheffe), indicating significant differences between the average response levels of Entitled and Benevolents. Benevolents were less likely to attempt to make the other person (referent) work harder when faced with an inequitable situation regardless of their equity orientation than were Entitleds.

Hypothesis 4 was partially supported. The analysis of variance for scenario 2 was significant (p < 0.05) with a post hoc test (Scheffe), indicating significant differences between the response levels of Benevolents and Equity Sensitives. Benevolents were less likely to respond to an inequitable scenario with a response of attempting to do something to reduce their referent's pay than were those subjects measured to be Sensitives.

Hypothesis 5 was not supported. There were no significant differences between the responses of the subjects regarding their efforts to attempt to think about reasons that justify why they were getting paid the same for doing more work (i.e., mentally justify their under-payment situation) when faced with an inequitable situation regardless of their equity orientation to either scenario.

Hypothesis 6 was not supported. There were no significant differences between the responses of the subjects regarding their efforts to attempt to think about reasons that justify why the other person was getting paid the same for doing less work (i.e., mentally justify their referent's over-payment situation), regardless of their equity orientation to either scenario.

Hypothesis 7 was not supported. There were no significant differences between the responses of the subjects regarding their likelihood to compare themselves to someone else other than the original referent when faced with an under-reward situation, regardless of their equity orientation.

Hypothesis 8 was partially supported. The analysis of variance was significant (p < 0.001) for under-reward scenario 1 (same work for less pay) and also significant (p < 0.01) for scenario 2 (more work for same pay). A post hoc test (Scheffe) for scenario 1 indicated significant differences between Entitleds and Sensitives and Entitleds and Benevolents, but no significant difference between Sensitives and Benevolents. For scenario 2 the post hoc test indicated significant differences between the average response levels of Entitled and Benevolents.

Hypothesis 9 was partially supported. The analysis of variance for scenario 2 was significant (p < 0.01) with a post hoc test (Scheffe) indicating significant differences between the average response levels of Benevolents and all others. That is, Benevolents were less likely to respond to an inequitable situation by looking for a new job than were Entitleds and Sensitives.

Discussion and Conclusions

A number of important conclusions can be drawn from the results of this study. First, this research validates the dimensions of Equity Sensitivity consistent with the theoretical work of Huseman et aL (1987). We tested nine hypotheses looking for differences in responses of Benevolents, Sensitives, and Entitleds when placed in two hypothetical under-reward situations. We were able to find support for five of the nine hypotheses whereas previous research (Patrick and Jackson, 1991) found support for just one. Thus, in over half of the nine different responses to under-reward situations presented to the subjects, there were significant differences in responses between the Benevolents, Sensitives, and Entitleds. It appears that significant differences do exist between the groups in terms of how they respond to two types of under-reward situations.

It also appears that the type of under-reward situation has an important impact on the differentiation between the two groups. In the first type of scenario the participants were put into a situation in which they received less pay for doing the same amount of work as their referent. An outcome (pay), not an input (amount of work), was the key-determining factor in making them experience under-reward. In this type of scenario the Entitleds were differentiated from all others. They were significantly more likely to reduce their effort or transfer.

In the second type of scenario, the participants were put into a situation that has previously not been tested in which they received the same pay for doing a greater amount of work than their referent. An input (amount of work), not an outcome (pay), was the key-determining factor in making them experience under-reward. This situation resulted in a much greater number of significant differences than the less pay for same work situation (five versus two).

Furthermore, the Benevolents were the group that stood out as responding differently to this type of an under-reward. They were significantly less likely to respond to this sort of under-reward by reducing their effort, trying to make their referent work harder, attempting to reduce their referent's pay, attempting to transfer or quit.

While previous research (King et al., 1993; Miles et al., 1994) has concluded that Entitleds tend to place more importance on external tangible outcomes (e. g., pay), whereas Benevolents place more importance on intrinsic outcomes (e. g., a sense of accomplishment), these studies failed to investigate whether there were significant differences in how the groups would intend to react to inequitable situations based on these preferences. It is worth noting that a conclusion that can be drawn from the present study is that the Entitleds and Benevolents are indeed significantly more likely to respond in a manner that is in line with their respective value orientations when faced with intrinsically- versus extrinsically-based, under-reward situations.

In summary, it appears "more work for same pay" situations produce stronger reactions than "less pay for same work" situations. It also appears that Entitleds are more likely to react to pay inequities. Conversely, Benevolents seem to be less likely to react to workload inequities. They exhibited a significantly higher threshold for workload inequities. Entitleds, on the other hand, seem to be more sensitive to differences in extrinsic rewards (pay). They exhibited a significantly lower threshold for experiencing inequity when pay differences were present.

These differences may have important practical applications. For example, managers of Entitleds should be cognizant that these employees will be more likely to reduce their effort or attempt to transfer if faced with a situation where they are being paid less than their referent for doing a similar amount of work. Conversely, Benevolents will be less likely to react to under-reward situations in which they are expected to do more work than their referent for the same pay. The old adage that effective managers must know their people seems to be particularly relevant.

Another conclusion that can be drawn from the results concerns how subjects will respond to resolve the dissonance when experiencing an inequitable situation. Equity Sensitivity suggests that Benevolents will behave differently than Sensitives who will behave differently than Entitleds (Huseman et al., 1987). Equity Sensitivity was most successful at differentiating between the two groups when responses required overt behavioral actions on the part of the subjects such as transferring, quitting, reducing one's effort, attempting to make the other person work harder or do something to reduce the referent's pay. There was no differentiation between equity orientation when choices were cognitive such as changing who was used as the referent or mentally justifying the inequitable situation.

This finding is particularly relevant to practicing managers because it implies their employees are more differentiated with regard to what they will do as a response to inequity; that is, Benevolents, Sensitives, and Entitleds will not equally "suffer in silence." When faced with an inequitable situation in which the employee is being paid less for doing the same work, the Entitled worker is more inclined to be less productive or attempt to transfer. On the other hand, when experiencing a situation where the inequity is due to receiving the same pay for doing more work, the Benevolent group is more differentiated. Benevolents are less likely to respond overtly to such an inequitable situation than the Entitled or Sensitive group.

Limitations and Areas for Future Research

The results of this study are exploratory and merit further research. Due to the structural design of the study certain limitations add caution to the interpretation of the results.

Foremost among these limitations is the fact that the situations presented were hypothetical rather than real. While it would certainly be preferable to conduct this research in a real-world setting with actual working adults, this was infeasible. A proper organizational setting which would allow access to a large sample of working adults and the ability to manipulate such sensitive variables as pay and workload was simply unavailable.

Considering this limitation, the scenario approach which has a history of use and acceptance in the Equity Theory literature (Austin et al., 1980, King et al., 1993) was utilized. While this methodology provides an excellent vehicle for standardizing situations and reducing within-subject variation, it may be criticized as unrealistic or, worse, far removed from the subjects' experience. The scenarios used in this study placed the subjects in the role of a student providing assistantship work to a professor (i.e., coding questionnaires). This type of hypothetical work situation was not a large mental jump for the subjects to make even if they did not have any previous work experience. All of the subjects were currently students; therefore, they were easily able to relate to the hypothetical work scenarios. Furthermore, a large percentage of the sample (86%) had at least prior part-time work experience. The ability of this predominately work-experienced group of subjects to cognitively relate to a part-time work situation seems logically sound. As such, no subjects expressed concern or confusion about being able to relate to the hypothetical scenarios or to answer any of the survey questions. Thus, the scenarios proved to be relevant to both inexperienced undergraduates and those students with prior work experience.

The study is also limited by its exclusive use of undergraduate students as subjects. At issue is the external validity of the study. Do the results apply only to traditional undergraduate students and have little or no relevance to working adults? While we do not claim that the results are generalizable to the entire working population, we do think that the results are representative of a segment of the population that is older and more experienced than the traditional undergraduate subject. The facts that the average age of the subjects in this study was 23.7 years and that over sixty-four percent had full-time work experience exhibit that this sample was not entirely representative of traditional undergraduate students. As such, these older, more experienced subjects could better relate to a work situation (albeit a part-time, student job) and provide more thoughtful answers to the questions than a sample comprised primarily of inexperienced undergraduates.

Future research of the scenario type may be strengthened by gaining entre to a real world organization and by providing workplace-specific survey scenarios to a sample comprised entirely of full-time working adults. This sort of research design may be less likely to suffer from the aforementioned weaknesses and still be feasible from a practicality point of view.

Finally, the study is limited by its use of common method variance; that is, the independent and dependent variables were collected simultaneously. Method effects may be generated when the structure of the instruments in use generate covariant error rather than measurement variance. For example, prior questions may influence a respondent's interpretation of an item, memory, judge ment, or selection of a response (Tourangeau and Rasinski, 1988). Contextual factors such as the presentation of the order of questions gives rise to concern when some questions are more relevant to the research than others (Schwarz and Bless, 1992). This study used all questions (except demographic) equally to scale responses. Another contextual factor Swartz has warned of is the use of numerical anchors (Swartz, 1996) or frequency questions such as "always," "frequently," or "never" (Swartz, 1999). None of these concerns were relevant to this study.

Of utmost concern is the validation of construct validity. Statistical validation of the instruments measuring equity sensitivity has been extensive and is reported earlier in this article. The instrument used to measure reactions to an inequitable situation (dependent variable) had no item overlap with equity sensitivity and was not worded in a fashion that implied an appropriate or desired response. Although common method variance issues cannot be summarily dismissed, the demonstrated validity of the equity sensitivity instrument, the lack of contextual issues used in constructing the instruments and the independence of construction of the dependent measures should moderate common method effects.

We call for future research to resolve some of these issues. The use of field experimental designs with working adults as subjects would address the major limitations of this study. Unfortunately, it is difficult to convince practicing managers to allow academicians to manipulate workers' senses of equity on the job. More likely, longitudinal designs may add to the understanding of this research by demonstrating the stability of the findings over time.