The multidimensional view of commitment and the theory of reasoned action: a comparative evaluation

The multidimensional view of commitment and the theory of reasoned action: a comparative evaluation – includes appendix

Thomas E. Becker

This study examined the relative ability of the multidimensional view of commitment and the theory of reasoned action to explain employee intentions and predict work behavior. Variables within the theory of reasoned action were superior to commitment in explaining employee intentions to be punctual and to engage in altruistic acts. However, the theory of reasoned action did not explain unique variance in either volitional behavior (altruism) or in less volitional behavior (tardiness). Finally, foci and bases of employee commitment accounted for significant variance in both altruism and tardiness, and explained variance in both behaviors over and above variables contained within the theory of reasoned action. Implications of these findings for the usefulness of the approaches are discussed.

Organizational commitment has been the subject of substantial research interest for a number of years (see Mathieu & Zajac, 1990; Randall, 1990, for recent reviews). In part, the extensive and continuing interest in organizational commitment stems from its presumed link to desirable employee behaviors. For a number of years, researchers have argued that employees who value organizational membership should eschew withdrawal behaviors, such as tardiness and absenteeism (Clegg, 1983; Cotton & Turtle, 1986). More recently, organizational commitment has been linked to organizational citizenship behaviors such as altruism (behavior that is directly and intentionally aimed at helping specific persons in face-to-face situations) and conscientiousness (a more impersonal type of helping behavior that does not provide immediate aid to a particular individual, but is intended to be indirectly helpful to other people in the organization) (Brief & Motowidlo, 1986; Organ, 1988). Such behaviors are important because they, “lubricate the social machinery of the organization” (Smith, Organ & Near, 1983, pp. 653-654).

Commitment has also been linked to intentions to engage in a variety of behaviors (Steers & Mowday, 1981; Sussmann & Vechio, 1982). Rather than being content to examine simple job attitude-outcome relationships, recent researchers have shown a greater interest in investigating the complex psychological process associated with these relationships. There is a growing recognition that the influence of commitment on work behaviors may be mediated by behavioral intention (Mathieu & Zajac, 1990). This recognition is reflected in research linking commitment and turnover (Arnold & Feldman, 1982; Bluedorn, 1982; Michaels & Spector, 1982; Mobley, Griffeth, Hand & Meglino, 1979; Stumpf & Hartman, 1984) and commitment and absenteeism (Steers & Rhodes, 1978, 1984). In addition, there is evidence that commitment is closely related to affect (such as job satisfaction) and may be linked to dependent variables of interest through affective processes (Curry, Wakefield, Price & Mueller, 1986; Williams & Hazer, 1986).

Despite the general popularity of organizational commitment research, certain scholars have not been impressed with the construct’s ability to explain and predict behavior. Highly critical of traditional attitude-behavior research, Fishbein and Ajzen (1975) set forth a competing model, the theory of reasoned action, to explain and predict individuals’ actions. They rejected the assumption that there is a direct link between an attitude toward an object and any given action with respect to that object. They argued that consideration of attitudes toward objects, such as organizational commitment, does not enhance the prediction of behavior beyond that made possible by the variables contained within the theory of reasoned action. If such extraneous variables have an impact, the effect is indirect–mediated through major components of the model or the weighing of those components.

In an effort to test competing claims of these two perspectives, Hom and his colleagues (Horn & Hulin, 1981; Hom, Katerberg & Hulin, 1979) pitted organizational commitment against Fishbein and Ajzen’s (1975) theory of reasoned action. Using National Guard members as a sample, they demonstrated the comparative effectiveness of the theory of reasoned action over commitment in predicting intent to reenlist and actual reenlistment.

However, over the past decade, a number of advances have taken place involving the measurement and conceptualization of employee commitment. Indeed, in light of advances in theory and research on commitment, we argue that earlier evaluations are now inadequate tests of the comparative effectiveness of the commitment approach and the theory of reasoned action. In the present paper, building upon advances in both fields, we examine the relative power of the Fishbein and Ajzen approach and the commitment construct to explain and predict employee intentions and behavior.

Theoretical Background

The Multidimensional View of Commitment

There is growing evidence that the attitudinal commitment of employees to the workplace is multidimensional and that the foci and bases of commitment can improve the prediction of employee intentions and behaviors. Foci of commitment are the particular entities, such as individuals and groups, to whom an employee is attached (Reichers, 1985). Bases of commitment are the motives engendering attachment (O’Reilly & Chatman, 1986). It has been known for some time that employees may be committed to foci such as professions (Gouldner, 1957, 1958) and unions (Gordon, Philpot, Burt, Thompson & Spiller, 1980; Gordon, Beauvais & Ladd, 1984), as well as to organizations (Mowday, Porter & Steers, 1982). More recent research has demonstrated that employees are differentially committed to top management, supervisors, coworkers, and customers (Becket, 1992; Reichers, 1986), and a recent metaanalysis has concluded that there is substantial evidence for the existence of multiple commitments (Mathieu & Zajac, 1990). Related studies suggest that commitment is not a zero-sum game; many employees evince high degrees of commitment to multiple foci (Becket & Billings, 1993; Conlon & Gallagher, 1987).

Regarding the bases of commitment, early research suggested that there are different motivational processes underlying single attitudes. According to Kelman (1958; 1961), compliance occurs when attitudes and behaviors are adopted in order to obtain specific rewards or avoid specific punishments. Identification occurs when people adopt attitudes and behaviors in order to be associated with a satisfying, self-defining relationship with another person or group. Finally, internalization occurs when people adopt attitudes and behaviors because their content is congruent with the individuals’ value systems. More recent research has demonstrated that employee commitment, as a work-related attitude, may be based upon disparate motives (Allen & Meyer, 1990; Caldwell, Chatman & O’Reilly, 1990; O’Reilly & Chatman, 1986). O’Reilly and Chatman (1986), for example, found that compliance, identification, and internalization, viewed as bases of commitment, were differentially related to prosocial organizational behaviors, turnover, and intent to stay with an organization.

Distinctions among foci and bases of commitment are not made in the more conventional view of commitment. According to this view, organizational commitment is defined as normative, involving, “the relative strength of an individual’s identification with and involvement in a particular organization” (Mowday et al., 1982, p. 27). Consistent with this unidimensional view (Cook, Hepworth, Wall & Warr, 1981), organizational commitment has most often been measured by the Organizational Commitment Questionnaire (OCQ) developed by Porter and his colleagues (Porter, Steers, Mowday & Boulian, 1974).

However, the theory underlying the multiple commitments literature holds that an employee’s commitment to the workplace cannot be adequately explained by commitment to the organization alone because the coalitional nature of organizations leads employee commitment to be multidimensional (Reichers, 1985). Further, the theory underlying the literature on the bases of commitment suggests that the motives for attachment are an important dimension to consider in addition to the level of commitment to various foci. Thus, compared to earlier approaches, the multiple commitments perspective is much more specific with respect to the nature of an employee’s commitment.

The greater specificity of the multiple commitments approach appears to yield more detailed and relevant information on employee attachment which, in turn, is helpful in predicting specific behaviors. This line of reasoning is supported by the principle of compatibility, which suggests that a given attitude should be related to other attitudes and behaviors only to the extent that the targets (foci) of the attitudes and behaviors are similar (Ajzen, 1989; Fishbein & Ajzen, 1975, 1976). Thus, for example, a researcher interested in altruistic behavior directed toward a work group would be well advised to focus on commitment to the work group rather than on commitment to the organization.

Further, given that information on specific attitudes (such as satisfaction with supervisors and co-workers) aids in the prediction of withdrawal behaviors (Cotton & Turtle, 1986), it seems reasonable to expect that the multiple commitments approach would make a similar contribution. The logic in this case would be that norms regarding absenteeism and tardiness are often established by foci such as supervisors and work groups and, therefore, commitment to these particular foci should lead to an acceptance of withdrawal norms. This logic is supported by research tying group norms to withdrawal behaviors (George, 1990; Mathieu & Kohler, 1990).

In empirical support of the multidimensional view, Becker (1992) demonstrated that commitment to foci other than the organization, and the bases of commitment to those foci, account for variance in key dependent variables above and beyond that accounted for by the OCQ. Specifically, Becker found that commitments to top management, supervisors, and the work group were important determinants of job satisfaction, certain types of prosocial organizational behavior, and intent to quit, over and above commitment to an organization (as measured by the OCQ). Further, he found that compliance, identification, and internalization as bases of commitment were unique determinants above and beyond commitment to the foci.

As noted above, organizational citizenship behaviors, withdrawal behaviors, and intentions to perform these actions are central dependent variables in the commitment literature. In the present study, we included a subset of these intentions and behaviors: intentions to be punctual and to engage in altruism, and behavioral indicators of altruism and tardiness. These particular variables were selected in order that both volitional and nonvolitional behaviors might be represented in this investigation. The topic of volitionality is explicated in a later section. Given the above discussion, we predict the following:

H1: The multidimensional view of commitment will explain

significant variance in intentions to engage in altruistic acts,

intentions to be punctual, altruism, and tardiness.

The Theory of Reasoned Action

The theory of reasoned action proposes that a relatively small number of concepts can be used to predict, explain, and influence the behavior of individuals. Specifically the theory posits that a person’s intention to perform a behavior is the immediate determinant of the action. Behavioral intention is, in turn, a weighted sum of two determinants: attitude toward performing the behavior, and subjective norms regarding the behavior.

Attitude toward a behavior refers to the degree to which a person has a favorable or unfavorable evaluation of the given behavior. Attitude toward behavior is a function of the belief that performing the behavior will lead to certain consequences and the person’s evaluation of those consequences. The social component of Fishbein and Ajzen’s model, the subjective norm, reflects the perceived social pressure to perform or not to perform the behavior. The subjective norm is a function of the person’s beliefs about whether significant others think he or she should perform the behavior, weighted by the person’s motivation to comply with those significant others.

The theory of reasoned action has been shown to have strong predictive utility in a wide range of situations and for a wide range of behaviors (Budd, 1986; Prestholdt, Lane & Mathews, 1987; Sheppard, Hartwick & Warshaw, 1988). However, it has been noted that, “While intentions to perform behaviors under one’s volitional control should be the single best predictor of whether the person performs that behavior, intentions to perform behaviors not under one’s control … may not lead to very accurate predictions” (Fishbein & Stasson, 1990, p. 176).

Given this last point, it is likely that the theory of reasoned action will be more effective in explaining and predicting organizational citizenship behavior than it will be in explaining and predicting job withdrawal behaviors. According to Organ, “Organizational citizenship behavior represents individual behavior that is discretionary, not directly or explicitly recognized by the formal reward system” (1988, p. 4). Thus, altruism, as a form of organizational citizenship behavior, is by definition highly volitional (i.e., intentional). Withdrawal behaviors, such as tardiness and absenteeism, may also be volitional at times (Hulin, 1991; Steers & Rhodes, 1978). However, as a set, such behaviors appear to be far less volitional than organizational citizenship behavior. For example, an employee may be late to work due to heavy traffic, personal emergency, or other factors largely beyond his or her control. Thus, we suggest that altruism tends to be substantially more volitional than does tardiness.

Given the above discussion, we predict the following:

H2A: Attitude toward the behavior and subjective norms will

explain significant variance in intentions to engage in altruism and

to be punctual to work.

H2B: Intention to engage in altruism will explain significant variance

in altruism.

H2C: Intention to be punctual will not explain significant variance

in tardiness.

Commitment versus the Theory of Reasoned Action

The theory of reasoned action holds that variables such as demographic characteristics and attitudes toward individuals or groups do not help explain intentions (Ajzen, 1989; Fishbein & Ajzen, 1975). According to the theory, such variables are not directly related to intentions and their effect is completely mediated by variables contained within the theory (i.e., attitude toward the behavior and subjective norms). Hence, regardless of how commitment is conceptualized or measured, the theory of reasoned action would hold that commitment will not be important in explaining employee intentions. Therefore, we propose that the theory of reasoned action will be more successful than the multidimensional view of commitment in explaining employee intentions. Specifically,

H3A: Attitude toward behaviors and subjective norms will account

for variance in intent to be punctual and intent to perform altruistic

behaviors over and above that accounted for by the foci and bases

of commitment.

The theory of reasoned action also holds that variables such as employee commitment do not help explain or predict behavior (Ajzen, 1989; Fishbein & Ajzen, 1975). Again, according to the theory, such variables are not directly related to behavior and their effect is completely mediated by variables contained within the theory (i.e., attitude toward the behavior, subjective norms, and intention). Nevertheless, given the logic underlying H2B and H2C, we suggest that this proposition will hold only for those behaviors which are largely volitional. Therefore:

H3B: Intention to engage in altruism will account for variance in

altruism over and above that accounted for by the foci and bases of

commitment.

H3C: Intention to be punctual will not account for variance in

tardiness over and above that accounted for by the foci and bases of

commitment.

Methods

Survey Methodology and Respondents

This research was conducted in 16 restaurants, all of which are members of the same fast-service chain. The restaurants are located in Seattle, Washington, and all share the same top management (i.e., the chain is corporate managed, not franchised). To collect data on the independent variables, surveys were sent to all 306 individuals employed by the restaurants. This involved distributing survey packets to the managers of each restaurant, who then delivered them to each of his or her employees. The survey packet included a cover letter from us, a survey, and a prepaid, addressed envelope for return of the completed questionnaire. The cover letter briefly described the purpose of the study, assured the potential respondents of confidentiality, and provided instructions for the completion and return of the surveys. Completed surveys were mailed by the respondents directly to us. Following two follow-up memos, 112 usable surveys were returned for a response rate of 36.6%.

Respondents ranged in age from 16 to 64 years with a mean of 24.6 years, and ranged in education from 9 years to over 16 years completed with a mean of 12.2 years completed. The average employee had been with the organization for 25.2 months. Respondents included crew members (cooks, food preparers, cashiers, and servers), supervisors, and assistant managers. Female respondents comprised 65.8% of the sample, and 83.7% of the respondents were white. To examine the issue of nonresponse bias we gathered demographic information for nonrespondents from the restaurant managers, and then compared respondents to nonrespondents. The two groups did not significantly differ with respect to any of these variables.

Measures

Budd (1987) pointed out that the typical ordering of measures in the literature on the theory of reasoned action may promote artificial response consistency. This is because the conventional method has presented the measures of constructs in the precise order in which they appear in the Fishbein-Ajzen (1975) model; that is, items on attitudes and subjective norms are presented first, followed by items on intentions, followed by items on behaviors. Based on Budd’s recommendations, we sought to reduce the potential effect of demand characteristics in our study by randomizing sections of the survey. Specifically, we assigned all sections a number, wrote the numbers on separate index cards, and then randomly picked the cards out of a hat. As a consequence, sections pertaining to each of the measures were randomized within the questionnaire. The order of the measures in the survey was: commitment variables, attitudes toward behaviors, intentions, subjective norms, and demographic variables.

Commitment variables. To identify meaningful foci of commitment, one-on-one interviews were conducted with 16 employees representing six restaurants; the restaurants were selected to ensure geographical diversity (i.e., inner-city versus suburban locations). Interviews included open-ended questions, such as, “If I followed you around on a typical day, whom would I see you talking to and working with?”, and standardized, closed-ended queries such as, “Could you name your supervisor?” and, “Do you know all the employees at this location?” The standardized questions were based on the work of Reichers (1985), who identified potential foci, and our understanding of the organizational structure and formal reporting relationships of the restaurant.

On the basis of the frequency with which they were mentioned in the interviews, the three following foci were selected for inclusion in the study: the organization, restaurant management, and non-managerial employees. To measure normative commitment to the organization, we used the nine-item version of the Organizational Commitment Questionnaire (OCQ) (Mowday et al., 1982). A 17-item measure developed by Booker (1992) was used to assess the bases of commitment to each of the other foci, with the only difference being theft the labels for the foci were changed to be congruent with the foci pertinent to the current investigation. Responses to both the OCQ and Becker’s measure were given on a seven-point scale ranging from “strongly disagree” to “strongly agree, with “don’t know” and “not applicable” options available.

Via a series of factor analyses, we developed a total of four scales assessing commitment to the foci. Principal axis factoring with squared multiple correlations as communality estimates was used to fit the common factor model to the data. The number of factors was selected via an examination of scree plots and differences among eigenvalues. The factors were obliquely rotated to a Harris-Kaiser orthoblique independent cluster solution. The four final scales were normative commitment to the organization, normative commitment to restaurant management, normative commitment to non-managerial employees, and overall compliance without regard to foci. These scales are consistent with prior research that has found that internalization and identification items sometimes tend to load on a single factor (Caldwell et al., 1990) and that compliance tends to be an across-foci construct (Becker, 1992). Scale scores were computed by summing across items within scales. Factor loadings and items for Becker’s measures are given in the appendix. Complete information on scale development is available upon request.

Attitudes toward behaviors. A semantic differential for each of the behaviors (altruism and tardiness) was used to assess the relevant attitude. The bipolar evaluative adjective scales were good-bad and wise-foolish. For example, our measure of attitude toward punctuality was, “My being on time to work every shift I work would be,” followed by the two semantic differentials. Responses were scored from one (extremely bad, extremely foolish) to seven (extremely good, extremely wise), and the sum across the two scales was taken as the respondent’s attitude toward the behavior. This approach is consistent with the recommended measure of attitudes proposed by Ajzen and Fishbein (1980).

Subjective norms. Our measure of subjective norms was drawn from Ajzen and Fishbein (1980). For each behavior, a single item was used. For example, our measure of subjective norms regarding punctuality was, “Most people who are important to me would probably think I should be to work on time every shift I work.” Responses were scored on a scale from one (extremely unlikely) to seven (extremely likely). It should be noted that subjective norms are not isomorphic with group norms. Subjective norms derive from the same cognitive source (the individual) as do the other survey measures; group norms do not.

Intentions. We used three items for each behavior to measure intentions. The wording for the first item was drawn from Ajzen and Fishbein (1980). For example, for punctuality, this item read, “I intend to be on time to work every shift that I work.” Responses were given on a scale from one (strongly disagree) to seven (strongly agree). The next two items recognized the distinction between behavioral self-prediction and desire (Fishbein & Stasson, 1990; Warshaw & Davis, 1985). For example, for punctuality, the self-prediction item read, “How likely is it that you will be on time to work every shift that you work?” Responses were given on a scale from one (extremely unlikely) to seven (extremely likely). For punctuality, the item measuring desire read, “I very much want to be on time to work every shift that I work.” Responses were given on a scale from one (strongly disagree) to seven (strongly agree). Factor analyses for each behavior demonstrated that the three items loaded cleanly on a single factor. The sum across the three items was taken as the respondent’s intention to engage in the behavior.

Altruism. Two months after employees completed their questionnaires, managers rated the altruism of each employee in his or her restaurant during the past two months. In assessing this behavior a shortened and revised 5-item version of the Smith et al. (1983) instrument was used. Specifically, the following five items were used to measure altruism: (1) Volunteered for things that were not required; (2) Oriented new people even though it was not required; (3) Helped others who had heavy work loads; (4) Assisted boss with his or her work; (5) Made innovative suggestions to improve the restaurant. Responses were given on a scale from one (never) to five (always). Factor analysis demonstrated the existence of one factor, and all 5 items loaded cleanly on the factor as expected. Because we substantially revised the Smith et al. scale for this study, we used exploratory rather than confirmatory factor analysis. The sum across the five items was taken as the respondent’s level of altruism.

Tardiness. During a two-month period following the completion of surveys by employees, managers of each of the 16 restaurants kept track of the punctuality of his or her employees. Using work sheets that we provided, managers recorded every incident of tardiness. Our measure of tardiness for a given employee was simply the number of times that the manager recorded the employee as being late to work. While it could be argued that punctuality is different from tardiness, in the present study the distinction is not meaningful. The number of times a given employee is on time for work is equal to the total number of shifts scheduled minus the number of shifts he or she was tardy. All employees in this study were full-time workers scheduled for the same number of shifts each week, so the number of shifts scheduled was a constant. Therefore, our results would be the same whether we labelled this dependent variable punctuality or tardiness; only the direction of the relationships would differ.

Demographic variables. The final section of the employee survey asked the respondents for demographic information. Specifically, we collected data on gender, age, education, ethnic group, and tenure with the company.

Treatment of Missing Data

It is common in survey research to have some subjects for whom data are missing on one or more independent variables. Pairwise and listwise deletion of missing values are common ways to deal with the problem. However, both of these techniques have been severely criticized because they make the dubious assumption that the data are missing randomly (Cohen & Cohen, 1983; Little & Rubin, 1987; Tabachnick & Fidell, 1983). In addition, listwise deletion often results in omitting a substantial portion of usable data; this can reduce statistical power considerably (Cohen & Cohen, 1983).

While there are a number of other techniques available for coping with the problem of missing data, estimating missing values by substituting means appears to be a simple and, under most conditions, justifiable approach (Cohen & Cohen, 1983). As with other approaches, only missing data for the independent variables are estimated; those subjects missing data on the dependent variable are dropped from the analyses. One potential disadvantage to this method is that, when a large number of values are missing, the variance of the variables may be significantly reduced. This, in turn, could result in overly stringent tests because the magnitude of the relationships among variables would be decreased (Tabachnick & Fidell, 1983). However, in the current study this does not appear to be a problem. The maximum percentage of subjects for whom data were missing for any given independent variable in this study was 6.42%. It is important to note that this maximum percentage involved two demographic variables (age and tenure), not the central variables of interest. Also, for each independent variable that had missing data, we tested the null hypothesis that the variance of the variable before substituting means was equal to or less than the variance after substituting means. In no instance was this hypothesis rejected. Thus, we concluded that the mean replacement approach to dealing with missing data was, in our case, a conservative and appropriate one.

Results

Table 1 contains the data set sizes, means, standard deviations, and measures of reliability for the variables. This table also shows the correlations among the variables. Because the distribution of tardiness was positively skewed and leptokurtic, tardiness data were transformed by taking the square-root of each participant’s score, This procedure produced a more normally distributed set of scores appropriate for use in the regression analyses discussed below.

Table 1. Descriptive Statistics and Intercorrelations(a,b,c)

Variable N M SD Alpha 1

Organization-Normative (1) 110 43.43 12.59 .92 .

Store Management-Normative (2) 111 61.46 14.75 .91 .57

Employee-Normative (3) 111 67.51 13.17 .89 .39

Compliance (4) 111 59.72 13.70 .79 .06

Attitude-Altruism (5) 108 12.57 1.71 .78 .24

Attitude-Punctuality (6) 108 13.14 1.80 .82 .12

Subjective Norm-Altruism (7) 109 5.97 1.21 .66 .17

Subjective Norm-Punctuality (8) 109 6.06 1.33 .66 .23

Intent-Altruism (9) 109 19.50 2.23 .83 .35

Intent-Punctuality (10) 109 20.05 1.79 .67 .46

Altruism (11) 93 14.85 6.18 .89 .28

Tardiness (12) 88 0.66 1.39 — .09

Variable 2 3 4 5 6 7

Organization-Normative (1)

Store Management-Normative (2) .

Employee-Normative (3) .35 .

Compliance (4) .10 .13 .

Attitude-Altruism (5) .24 .22 .10 .

Attitude-Punctuality (6) .09 .17 .03 .49 .

Subjective Norm-Altruism (7) .02 .15 .01 .18 .10 .

Subjective Norm-Punctuality (8) .09 .22 .04 .26 .22 .40

Intent-Altruism (9) .29 .32 .02 .38 .46 .30

Intent-Punctuality (10) .33 .24 .04 .35 .55 .15

Altruism (11) .42 .31 .02 .15 .17 .07

Tardiness (12) .02 .29 .39 .16 .07 .08

Variable 8 9 10 11 12

Organization-Normative (1)

Store Management-Normative (2)

Employee-Normative (3)

Compliance (4)

Attitude-Altruism (5)

Attitude-Punctuality (6)

Subjective Norm-Altruism (7)

Subjective Norm-Punctuality (8) .

Intent-Altruism (9) .30 .

Intent-Punctuality (10) .41 .65 .

Altruism (11) .05 .26 .30 .

Tardiness (12) .04 .13 -.04 .01 .

Notes:

(a.) Reliability for the single-item subjective norm measures was assessed via test-retest data gathered from 72 upper-level business students over a one month interval. For the multi-item scales, reliability was assessed via coefficient alpha.

(b.) Correlations greater than .20 are significant at .05. Data set sizes ranged from 85 to 111.

(c.) The number of tardinesses ranged from zero to nine. Over one-third of the respondents were tardy at least once during the two month interval.

To test our hypotheses, two sets of hierarchical regressions were run for each intention and behavior. First, a hierarchical regression was run with demographic variables entered first, variables from the theory of reasoned action entered second, anti the set of commitment variables entered last. These analyses allowed us to test H2A, H2B, and H2C, controlling for the demographic variables. In addition, these analyses allowed us to explore whether or not commitment explained variance in the dependent variables over and above that accounted for by the variables from the theory of reasoned action. Second, a hierarchical regression was run with demographic variables entered first, the set of commitment variables entered second, and variables from the theory of reasoned action entered last. These analyses allowed us to test HI, H3A, H3B, and H3C, controlling for the demographic variables. To estimate the statistical power of the tests reported in Table 2, we used the procedures described by Cohen (1988, pp. 407-520). Assuming an effect size of .20, the power for all of these tests was .81 or greater. We assumed this effect size, defined by Cohen (1988) as “medium,” because there is insufficient literature on comparisons between the multiple commitments approach and the theory of reasoned action for us to base estimates of effect size on prior research. We felt that the assumption of a large effect size would be too liberal, and that assuming a small effect size would be too conservative.

Table 2. Hierarchical Regressions: Multidimensional Commitment vs the Theory of Reasoned Action

Variables N R2 [Delta]R2 F

Intent-Altruism

Demography 109 .089 2.01(+)

TRA 109 .279 .190 13.28(**)

Commitment 109 .329 .050 1.83

Demography 109 .089 2.01(+)

TRA 109 .198 .109 3.37(*)

Commitment 109 .329 .131 9.47(**)

Intent-Punctuality

Demography 109 .121 2.84(*)

TRA 109 .496 .375 37.48(**)

Commitment 109 .545 .049 2.61(*)

Demography 109 .121 2.84(*)

TRA 109 .239 .118 3.84(**)

Commitment 109 .545 .306 32.51(**)

Altruism

Demography 93 .229 5.17(**)

TRA 93 .232 .003 0.34

Commitment 93 .331 .099 3.04(*)

Demography 93 .229 5.17(**)

TRA 93 .331 .102 3.17(*)

Commitment 93 .331 .000 0.01

Tardiness

Demography 88 .101 1.85

TRA 88 .112 .011 0.92

Commitment 88 .237 .125 3.16(*)

Demography 88 .101 1.85

TRA 88 .237 .136 3.46(*)

Commitment 88 .237 .000 0.01

Notes: For intentions, the set of variables for the theory of reasoned action (TRA) contains attitude toward the behavior and subjective norms. For behaviors, this set contains one variable: the predicted value of intentions (i.e., the value of intention estimated from the regression of intention on attitude toward the behavior and subjective norms).

(+) p <.10; (*) p <.05; (**) p <.01.

In the regressions of intentions on the independent variables, the set of variables for the theory of reasoned action contained attitude toward the behavior and subjective norms. In the regressions involving behaviors the theory of reasoned action was represented by one variable: the predicted value of intention (i.e., the value of intention estimated from the regression of intention on attitude toward the behavior and subjective norms). We used this approach because the theory of reasoned action holds that the effects of attitude toward behavior and subjective norms on behavior is completely mediated by intention; therefore, only that portion of intention that is predicted by attitudes and subjective norms should be used to predict behavior.

The results are shown in Table 2. As can be seen by examining the first set of hierarchical regressions, variables from the theory of reasoned action (attitude toward behavior and subjective norms) account for significant variance in intentions to engage in altruism and to be punctual, controlling for the demographic variables. This supports H2A. However, predicted values of intentions to engage in altruism do not account for significant variance in altruism; this contravenes H2B. As predicted by H2C, predicted values of intentions to be punctual do not explain significant variance in punctuality.

The first set of hierarchical regressions for the intent to engage in altruism shows that the set of commitment variables (normative commitment vis-a-vis the organization, store management, and co-workers) does not explain variance in intentions to engage in altruism over and above variables within the theory of reasoned action. However, as the first set of hierarchical regressions for the other dependent variables demonstrates, the commitment variables do account for variance in intent to be punctual, altruism, and tardiness over and above variables from the theory of reasoned action.

The second set of hierarchical regressions for each dependent variable shows that the commitment variables explain significant variance in each of the intentions and behaviors, controlling for the demographic variables. This supports H1. These analyses also indicate that attitude toward behavior and subjective norms account for variance in intentions above and beyond that accounted for by the commitment variables. This supports H3A. Contrary to H3B, intentions do not account for variance in altruism over and above the commitment variables. Finally, in accordance with H3C, intentions do not account for variance in tardiness above and beyond that accounted for by the commitment variables.

To examine the relationship between each independent variable and the dependent variables (controlling for the other independent variables), we conducted four simultaneous regressions. Table 3 contains the results. As can be seen, attitude toward behavior and subjective norms each account for unique variance in intentions. In addition, ethnic group accounts for unique variance in intent to be altruistic, and age, ethnic group, and normative commitment to the organization account for unique variance in intent to be punctual. With respect to behaviors, tenure, normative commitment to store management, and normative commitment to employees explain unique variance in altruism. Normative commitment to employees and compliance account for unique variance in tardiness.

Table 3. Simultaneous Regressions: Multidimensional Commitment vs the Theory of Reasoned Action

Variables N Beta SE t

Intent-Altruism

Gender 109 -.04 .10 -0.41

Age 109 .11 .10 1.14

Education 109 .06 .09 0.72

Ethnic Group 109 -.18 .09 -2.05(*)

Tenure 109 .07 .10 0.69

Organization-Normative 109 .10 .11 0.87

Store Management-Normative 109 .02 .11 0.20

Employees-Normative 109 .18 .10 1.85+

Compliance 109 -.02 .09 -0.26

Attitude 109 .27 .09 3.00(**)

Subjective Norm 109 .24 .09 2.75(**)

Intent-Punctuality

Gender 109 -.03 .08 -0.36

Age 109 .17 .08 2.14(*)

Education 109 .02 .07 0.29

Ethnic Group 109 -.15 .07 -2.08(*)

Tenure 109 .09 .08 1.11

Organization- Normative 109 .20 .09 2.22(*)

Store Management-Normative 109 .06 .09 0.71

Employees-Normative 109 -.01 .08 -0.11

Compliance 109 .03 .07 0.47

Attitude 109 .45 .07 6.34(**)

Subjective Norm 109 .27 .07 3.62(**)

Altruism

Gender 93 .03 .11 0.29

Age 93 .06 .11 0.56

Education 93 .10 .09 1.08

Ethnic Group 93 .03 .10 0.33

Tenure 93 .31 .11 2.93(**)

Organization-Normative 93 -.08 .12 -0.68

Store Management-Normative 93 .26 .12 2.13(*)

Employees-Normative 93 .22 .10 2.13(*)

Compliance 93 -.04 .10 -0.43

Predicted Intentions 93 .00 .10 0.01

Tardiness

Gender 88 -.18 .11 -1.54

Age 88 -.13 .12 -1.12

Education 88 .04 .10 0.42

Ethnic Group 88 -.03 .11 -0.30

Tenure 88 -.05 .11 -0.48

Organization-Normative 88 .05 .13 0.35

Store Management-Normative 88 -.08 .13 -0.62

Employees-Normative 88 .29 .11 2.51(*)

Compliance 88 .21 .11 1.97(*)

Predicted Intentions 88 .00 .10 0.06

Notes: Gender was coded 0 = male, 1 = female. Ethnic group was coded 0 = nonwhite, 1 = white, + p <. 10;

(*) p < .05; (**) p < .01.

Discussion

The current paper contributes to the past literature on commitment and the theory of reasoned action in several ways. It extends recent research on the foci and bases of commitment by including an examination of commitment to foci not included in previous investigations and dependent variables not previously examined. More importantly, this represents the first attempt to compare the multiple commitment approach and the theory of reasoned action. In contrast to prior research comparing the theory of reasoned action and organizational commitment (Horn & Hulin, 1981; Horn et al., 1979), this work includes (1) a sample from the private sector, (2) dependent variables which, we would argue, are more relevant to business organizations, and (3) current, state-of-the-art conceptualizations and measures. While Horn and his colleagues compared a unidimensional view of organizational commitment with the theory of reasoned action, our comparisons involved a multi-dimensional perspective consistent with more recent conceptualizations of attachment.

Our primary aim was to compare the relative ability of the multidimensional view of commitment and the theory of reasoned action to explain employee intentions and to predict work behavior. In explaining intentions to perform work behaviors, the picture was clear cut. As predicted, attitude toward behavior and subjective norms were uniformly superior to the multidimensional view of commitment. The commitment approach failed to account for variance in intention to be altruistic over and above that accounted for by variables from the theory of reasoned action. Although commitment explained unique variance in intent to be punctual, the amount of explained variance was relatively small.

While findings regarding employee intentions are of interest, the prediction of employee behaviors may have more practical implications. In predicting the work behaviors included in this study, the theory of reasoned action did not outperform the multidimensional view of commitment. In fact, the commitment approach performed better than the theory of reasoned action in the prediction of both altruism and tardiness. Therefore, the proposition that the theory of reasoned action has greater predictive power than the commitment approach for work behaviors cannot be supported.

We hypothesized that the theory of reasoned action would not be particularly useful in predicting tardiness because the theory is meant to be most relevant to behaviors which are largely volitional; tardiness, while partly under the control of employees, would also appear to have a substantial nonvolitional component. Our data supported this line of reasoning. Intent to be punctual was unrelated to tardiness (r = -.04, ns), and the theory of reasoned action did not account for variance in tardiness beyond that explained by demographic and commitment variables. We also hypothesized that the theory of reasoned action would explain significant variance in altruism because this behavior is, by definition, largely volitional. The correlational analysis appeared to support this hypothesis; intent to engage in altruistic acts was significantly related to altruism (r = .26, p [is less than] .05). However, we also found that variables from the theory of reasoned action did not account for variance in altruism over and above variance accounted for by demographic or commitment variables.

There are several possible reasons for why the theory of reasoned action did not explain unique variance in altruism. First, we controlled for a number of demographic variables that are not often controlled for in research on the theory of reasoned action. Some advocates of the theory might argue that demographic effects should not be controlled because variables within the theory are presumed to mediate such effects. Our view is that if the theory cannot explain variance in key dependent variables beyond that accounted for by demographic variables, then the theory is of questionable utility. Second, in predicting behavior, we used predicted intentions (estimated from the regression of intention on attitude toward the behavior and subjective norms) rather than measured intentions. While past researchers have used measured intentions in their analyses (e.g., Hom et al., 1979; Horn & Hulin, 1981), we assert that our approach was more consistent with the theory. Third, as explained below, altruism may be partly a function of affective processes. Because the theory of reasoned action does not appear to assess affect, it may not be particularly useful in predicting behaviors that have a substantial affective component.

Why did the multidimensional approach to commitment outperform the theory of reasoned action in predicting behavior? In retrospect, we suspect that one major reason is that the multidimensional commitment perspective is linked to behavior through both cognitive and affective processes. For example, with respect to cognitions, more committed individuals tend to have different expectations, intentions, and attributions than do less committed people (Mowday et al., 1982). In addition, regarding affect, more committed employees tend to be more motivated and satisfied than their less committed co-workers (Mathieu & Zajac, 1990). Thus, even after cognitive variables (such as intentions) are partialled out, the affective links between commitment and behavior remain. The theory of reasoned action, on the other hand, is strictly a cognitive approach (Fishbein & Ajzen, 1975, 1976). To the extent that behavior is influenced by both cognition and affect, a purely cognitive approach would be at a disadvantage compared to the multiple commitments approach. This line of reasoning is elucidated below.

The simultaneous regressions suggested that commitments to the supervisor and workgroup were the key forms of attachment accounting for unique variance in altruism. Given that our measure of altruism involved cooperative behaviors directed towards the supervisor and workgroup (e.g., helping co-workers who had heavy workloads, assisting the boss with his or her work), this finding is not surprising. As discussed in an earlier section, such a finding is consistent with the notion that commitment should be most related to behavior when commitment and behavior have the same foci. Because the relationship between commitment and altruism remained after controlling for intentions, we suggest that the affective link mentioned above is relevant here. Commitment to supervisors and co-workers may generate positive affect because working with others who share one’s values and with whom one identifies is intrinsically rewarding; simply put, it feels good to be a member of such a group. Positive affect, in turn, can lead directly to altruism (sec Krebs & Miller, 1985, for further discussion of this link), including prosocial organizational behaviors (Isen & Baron, 1991).

The simultaneous regressions also showed that commitment to employees and overall compliance were the only variables accounting for unique variance in tardiness. The fact that compliance was positively related to tardiness is consistent with the assertion that “compliance has negative implications, apparently because this form of attachment is fleeting and does not involve acceptance of norms and values beneficial to an organization” (Becker, 1992: 234). In fact, Becker (1992) found that compliance was positively related to intent to quit and negatively related to prosocial behaviors; consistent with our proposed affective link, he also found a negative relationship between compliance and satisfaction.

More surprising is the finding that normative commitment to employees was also positively related to tardiness. However, a potential explanation for this result becomes clear when one recalls that normative commitment to employees involves the internalization of norms and values of the workgroup. To the extent that workgroup norms are accepting of tardiness, normative commitment to the workgroup would be expected to be positively related to tardiness. One might argue that this relationship should have been reflected in the theory of reasoned action via the subjective norms construct. However, like behavior, acceptance of norms and values may be determined by both cognitions and affect (Van Maanen & Schein, 1979). It is possible that subjective norms, as operationalized by the theory of reasoned action, only capture the cognitive component of norm acceptance. Consistent with our on-going argument, the multiple commitments approach is likely to capture both cognitive and affective components. In addition, the typical measure of subjective norms (including that used in our study) simply asks respondents about the perceived norms of “most people who are important to me.” It is possible that most people (such as supervisors, family members, and so on) could value punctuality while a minority of individuals (such as the other employees in a respondent’s restaurant) could value the freedom to come to work late occasionally. If so, we would argue that local norms established by the workgroup would be more psychologically potent due to the salience and influence of this group in the immediate work setting.

Based on this study, we can make the following recommendations. First, researchers and practitioners interested in explaining and predicting intentions would be well-advised to utilize variables from the theory of reasoned action (attitude toward behavior and subjective norms) rather than foci and bases of employee commitment. Second, those interested in predicting organizational citizenship behavior and employee withdrawal should probably prefer the multidimensional approach to commitment over the theory of reasoned action.

In considering these suggestions, several caveats deserve mention. While we have tried to be impartial in our comparisons, it is difficult to determine whether or not our analyses are entirely fair. As Cooper and Richardson (1986) pointed out, comparative evaluations of different theories can be hindered by one theory being more strongly operationalized, manipulated, or measured than another theory. However, we believe that by including state-of-the-art measures for both approaches and selecting a sample with sufficient variance in the variables, we have taken reasonable steps to conduct a fair comparative evaluation.

Further, in assessing the generalizability of our findings, it should be noted that the sample was selected from a relatively low-skilled population in the fast-service industry, characterized by high levels of employee withdrawal and the opportunity to engage in a wide-range of organizational citizenship behaviors. In fact, we selected this sample because of these characteristics, the logic being that commitment and variables within the theory of reasoned action should be most relevant in such a setting. At the same time, we can think of no convincing reason why these features of the sample should have influenced the relative ability of the theory of reasoned action and the commitment approach to explain intentions or to predict behavior. Still, additional research with samples from other populations would help to address the issue of the generalizability of our findings.

Finally, two methodological points deserve mention. First, consistent with the recommendations of Budd (1987), and in an attempt to reduce concerns for artificial response consistency, we randomized sections of the survey. All subjects were then presented with the same random order. In retrospect, we can imagine how this approach could by chance produce an order of sections that would promote consistency effects. We recommend that future researchers randomize the ordering of sections separately for individual subjects (or at least for small groups of subjects). This would make it less likely that a randomized order accidentally encourages response bias for all subjects. Alternatively, the researcher could use his or her own judgment to present sections to all subjects in an order that seems least apt to promote a consistency bias.

Second, while our sample was large enough to allow the detection of moderate effects, a larger sample would have clearly created greater statistical power to detect small effect sizes. Also, while our response rate (about 37%) was not unusually low for survey research, neither was it impressively high. Thus, we recommend that later research replicate our findings with a larger sample, using methods (e.g., additional follow-up mailings) likely to produce higher response rates.

In conclusion, the results of this investigation provide support for the theory of reasoned action in explaining variability in employee intentions and the usefulness of the multidimensional view of commitment for the prediction of certain work behaviors. However, future work should continue to advance the theory underlying these two different approaches. For instance, more attention to how organizational commitment is translated into behavior is needed. Future researchers should also investigate conditions (such as time lag between attitude and behavior) which may influence the relative superiority of the approaches.

Note

(1.) We have used the term “normative commitment” in the same sense as did Becker (1992) and Caldwell et al, (1990), to wit: in referring to a form of commitment characterized by the internalization of values and identification with certain individuals or groups. Note that this is somewhat different than the way Allen and Meyer (1990) used the term.

References

Allen, N.J. & Meyer, J.P. (1990). The measurement and antecedents of affective, continuance and normative commitment to the organization. Journal of Occupational Psychology, 63: 1-18.

Ajzen, I. (1989). Attitude structure and behavior. Pp. 241-274 in A. Pratkanis, S. Breckler & A. Greenwald (Eds.). Attitude structure and function. Hillsdale, NJ: Erlbaum.

Ajzen, I. & Fishbein, M. (1980). Understanding attitudes and predicting social behavior, Englewood Cliffs, NJ: Prentice-Hall.

Arnold, H.J. & Feldman, D.J. (1982). A multivariate analysis of the determinants of job turnover. Journal of Applied Psychology, 67: 350-360.

Becker, T.E. (1992). Foci and bases of commitment: Are they distinctions worth making? Academy of Management Journal, 35: 232-244.

Becker, T.E. & Billings, R.S. (1993). Profiles of commitment: An empirical test. Journal of Organizational Behavior, 14: 177-190.

Bluedorn, A.C. (1982). A unified model of turnover from organizations. Human Relation, 35: 135-153.

Brief, A.P. & Motowidlo, S.J. (1986). Prosocial organizational behaviors. Academy of Management Review, 11: 710-725.

Budd, R.J. (1986). Predicting cigarette use: The need to incorporate measures of salience in the theory of reasoned action. Journal of Applied Social Psychology, 16: 663-685.

–. (1987). Response bias and the theory of reasoned action. Social Cognitions, 5: 95-107.

Caldwell, D.F., Chatman, J.A. & O’Reilly, C.A. III. (1990). Building organizational commitment: A multi-firm study. Journal of Occupational Psychology, 63: 245-261.

Clegg, C.W. (1983). Psychology of employee lateness, absence, and turnover: A methodological critique and an empirical study. Journal of Applied Psychology, 68: 88-101.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences, 2nd ed. Hillsdale, NJ: Erlbaum.

Cohen, J. & Cohen, P. (1983). Applied multiple regression/correlation analysis for the behavioral sciences, 2nd ed. Hillsdale, NJ: Erlbaum.

Conlon, E.J. & Gallagher, D.G. (1987). Commitment to employer and union: Effects of membership status. Academy of Management Journal, 30: 151-162.

Cook, J.D., Hepworth, S.J., Wall, T.D. & Warr, P.B. (1981). The experience of work: A compendium and review of 249 measures and their use. New York: Academic Press.

Cooper, W.H. & Richardson, A.J. (1986). Unfair comparisons. Journal of Applied Psychology, 71: 179-184.

Cotton, J.L. & Tuttle, J.M. (1986). Employee turnover: A meta-analysis and review with implications for research. Academy of Management Review, 11: 55-70.

Curry, J.P., Wakefield, D.S., Price, J.L. & Mueller, C.W. (1986). On the causal ordering of jab satisfaction and organizational commitment. Academy of Management Journal, 29: 847-858.

Fishbein, M. & Ajzen, I. (1975). Belief attitude, intention and behavior: An introduction to theory and research. Reading, MA: Addison-Wesley.

–. (1976). Misconceptions about the Fishbein model: Reflections on a study by Songer-Nocks. Journal of Experimental Social Psychology, 12: 579-584.

Fishbein, M. & Stasson, M. (1990). The role of desires, self-predictions, and perceived control in the prediction of training session attendance. Journal of Applied Social Psychology, 20: 173-198.

George, J.M. (1990). Personality, affect, and behavior in groups. Journal of Applied Psychology. 75: 107-116.

Gordon, M.E., Beauvais, L.L. & Ladd, R.T. (1984). The job satisfaction and union commitment of unionized engineers. Industrial and Labor Relations Review, 37: 359-370.

Gordon, M.E., Philpot, J.W., Burt, R.E., Thompson, C.A. & Spiller, W.E. (1980). Commitment to the union: Development of a measure and examination of its correlates. Journal of Applied Psychology, 65: 479-499.

Gouldner, A.W. (1957). Cosmopolitans and locals: Toward an analysis of latent social roles–I. Administrative Science Quarterly, 2: 281-306.

–. (1958). Cosmopolitans and locals: Toward an analysis of latent social roles–II. Administrative Science Quarterly, 2: 444-480.

Hom, P.W. & Hulin, C.L. (1981). A competitive tot of the prediction of reenlistment by several models. Journal of Applied Psychology, 66: 23-29.

Hom, P.W., Katerberg, R. Jr. & Hulin, C.L. (1979). Comparative examination of three approaches to the prediction of turnover. Journal of Applied Psychology, 64: 280-290.

Hulin, C.L. (1991). Adaptation, persistence, and commitment in organizations. Pp. 445-505 in M.D. Dunnette & L.M. Hough (Eds.), Handbook of industrial and organizational psychology, Vol. 2, 2nd ed. Palo Alto, CA: Consulting Psychologists Press, Inc.

Isen, A.M. & Baron, R.A. (1991). Positive affect as a factor in organizational behavior. Pp. 1-53 in L.L. Cummings & B.M. Staw (Eds.), Research in organizational behavior, Vol. 13. Greenwich, CT: JAI Press.

Kelman, H.C. (1958). Compliance, identification, and internalization: Three processes of attitude change. Journal of Conflict Resolution, 2:51-60.

–. (1961). Processes of opinion change. Public Opinion Quarterly, 25: 57-78.

Krebs, D.L. & Miller, D.T. (1985). Altruism and aggression. Pp. 1-71 in G. Lindzey & E. Aronson (Eds.), The handbook of social psychology, Vol. 2, 3rd ed.

Little, R.J.A. & Rubin, D.B. (1987). Statistical analysis with missing data. New York: Wiley.

Mathieu, J.E. & Kohler, S.S. (1990). A cross-level examination of group absence influences on individual absence. Journal of Applied Psychology, 75: 217-220.

Mathieu, J.E. & Zajac, D.M. (1990). A review and meta-analysis of the antecedents, correlates, and consequences of organizational commitment. Psychological Bulletin, 108: 171-194.

Michaels, C.E. & Spector, P.E. (1982). Causes of employee turnover: A test of the Mobley, Griffeth, Hand, and Meglino model. Journal of Applied Psychology, 67: 53-59.

Mobley, W., Griffeth, R., Hand, H. & Meglino, R. (1979). Review and conceptual analysis of the employee turnover process. Psychological Bulletin, 86: 493-522.

Mowday, R.T., Porter, L.W. & Steers, R.M. (1982). Employee-organization linkages: The psychology of commitment, absenteetism, and turnover. New York: Academic Press.

O’Reilly, C.A. III & Chatman, J. (1986). Organizational commitment and psychological attachment: The effects of compliance, identification, and internalization on prosocial behavior. Journal of Applied Psychology, 71: 492-499.

Organ, D.W. (1988). Organizational citizenship behavior. Lexington, MA: Lexington Books.

Porter, L., Steers, R., Mowday, R. & Boulin, P. (1974). Organizational commitment, job satisfaction, and turnover among psychiatric technicians. Journal of Applied Psychology, 59: 603-609.

Prestholdt, P.H., Lane, I.M. & Mathews, R.C. (1987). Nurse turnover as reasoned action: Development of a process model. Journal of Applied Psychology, 72: 221-227.

Randall, D.M. (1990). The consequences of organizational commitment: Methodological investigation. Journal of Organizational Behavior, 11: 361-378.

Reichers, A.E. (1985). A review and reconceptualization of organizational commitment. Academy of Management Review, 10: 465-476.

–. (1986). Conflict and organizational commitments. Journal of Applied Psychology, 71: 508-514.

Sheppard, B.H., Hartwick, J. & Warshaw, P.R. (1988). The theory of reasoned action: A meta-analysis of past research with recommendations for modifications and future research. Journal of Consumer Research 15: 325-343.

Smith, C.A., Organ, D.W. & Near, J.P. (1983). Organizational citizenship behavior: Its nature and antecedents. Journal of Applied Psychology, 68: 653-663.

Steers, R.M. & Mowday, R.T. (1981). Employee turnover and post-decision accommodation processes Pp. 235-281 in B. Staw & L.L. Cummings (Eds.), Research in organizational behavior, Vol. 3. Greenwich, CT: JAI Press.

Steers, R.M. & Rhodes, S.R. (1978). Major influences on employee attendance: A process model. Journal of Applied Psychology, 63: 391-407.

–. (1984). Knowledge and speculation about absenteeism In P.S. Goodman & R.S. Atkin (Eds.), Absenteeism: New approaches to understanding, measuring, and managing employee absence. San Francisco: Jossey-Bass.

Stumpf, S.A. & Hartman, K. (1984). Individual exploration to organizational commitment or withdrawal. Academy of Management Journal, 27: 308-329.

Sussmann, M. & Vechio, R.P. (1982). A social influence interpretation of worker motivation. Academy of Management Review, 7: 177-186.

Tabachnick, B.G. & Fidell, L.S. (1983). Using multivariate statistics. New York: Harper & Row.

Van Maanen, J. & Schein, E.H. (1979). Toward a theory of organizational socialization. In L.L. Cummings & B.M. Staw (Eds.), Research in organizational behavior, Vol. I. Greenwich, CT: JAI Press.

Warshaw, P.R. & Davis, F.D. (1985). Disentangling behavioral intention and behavioral expectation. Journal of Experimental Social Psychology, 21: 213-228.

Williams, L.J. & Hazer, J.T. (1986). Antecedents and consequences of satisfaction and commitment in turnover models: A reanalysis using latent variable structural equation methods. Journal of Applied Psychology, 71: 219-231.

APPENDIX

The following are the rotated factor patterns for commitment to non-managerial employees and restaurant management.(a)

Normative

Item Commitment

Non-managerial Employees

If the values of the employees were different, I

would not be as attached to the employees. .51

Unless the employees reward me for it in some way,

I see no reason to expend extra effort on behalf of

the employees. -.04

When someone criticizes the employees, it feels

like a personal insult. .58

In order for me to get rewarded by the employees,

it is necessary to express the right attitude. .48

I am very interested in what others think about the

employees. .70

My attachment to the employees is primarily based

on the similarity of my values and those

represented by the employees. .63

When I talk about the employees, I usually say “we”

rather than “they.” .61

The employees’ successes are my successes. .71

Since starting this job, my personal values and

those of the employees have become more similar. .78

When someone praises the employees, it feels like a

personal compliment. .66

What the employees stand for is important to me. .67

I talk up the employees to my friends as a great

group to work with. .65

How hard I work for the employees is directly

linked to how much I am rewarded. .35

My private views about the employees are different

than those I express publicly. -.11

I am proud of the employees. .70

The reason I prefer the employees to other groups

is because of what they stand for, that is, their

values. .84

I feel a sense of “ownership” for the employees. .67

Restaurant Management

If the values of store management were different, I

would not be as attached to store management. .37

Unless store management reward me for it in some

way, I see no reason to expend extra effort on

behalf of store management. -.06

When someone criticizes store management, it feels

like a personal insult. .65

In order for me to get rewarded by store

management, it is necessary to express the right

attitude. .39

I am very interested in what others think about

store management. .78

My attachment to store management is primarily

based on the similarity of my values and those

represented by store management. .66

When I talk about store management, I usually say

“we” rather than “they.” .70

Store management’s successes are my successes. .72

Since starting this job, my personal values and

those of store management have become more similar. .72

When someone praises store management, it feels

like a personal compliment. .70

What store management stands for is important to

me. .75

I talk up store management to my friends as a great

group to work for or with. 62

Flow hard I work for store management is directly

linked to how much I am rewarded. -.03

My private views about store management are

different than those I express publicly. -.12

I am proud of store management. .66

The reason I prefer store management to other

groups is because of what they stand for, that is,

their values. .83

I feel a sense of “ownership” for store management. .76

Item Compliance(b)

Non-managerial Employees

If the values of the employees were different, I

would not be as attached to the employees. .03

Unless the employees reward me for it in some way,

I see no reason to expend extra effort on behalf of

the employees. .62

When someone criticizes the employees, it feels

like a personal insult. -.11

In order for me to get rewarded by the employees,

it is necessary to express the right attitude. .11

I am very interested in what others think about the

employees. .09

My attachment to the employees is primarily based

on the similarity of my values and those

represented by the employees. .05

When I talk about the employees, I usually say “we”

rather than “they.” -.23

The employees’ successes are my successes. -.06

Since starting this job, my personal values and

those of the employees have become more similar. .12

When someone praises the employees, it feels like a

personal compliment. -.35

What the employees stand for is important to me. -.23

I talk up the employees to my friends as a great

group to work with. -.03

How hard I work for the employees is directly

linked to how much I am rewarded. .65

My private views about the employees are different

than those I express publicly. .54

I am proud of the employees. -.10

The reason I prefer the employees to other groups

is because of what they stand for, that is, their

values. .13

I feel a sense of “ownership” for the employees. .23

Restaurant Management

If the values of store management were different, I

would not be as attached to store management. .56

Unless store management reward me for it in some

way, I see no reason to expend extra effort on

behalf of store management. .55

When someone criticizes store management, it feels

like a personal insult. .06

In order for me to get rewarded by store

management, it is necessary to express the right

attitude. .27

I am very interested in what others think about

store management. -.14

My attachment to store management is primarily

based on the similarity of my values and those

represented by store management. .16

When I talk about store management, I usually say

“we” rather than “they.” -.30

Store management’s successes are my successes. -.06

Since starting this job, my personal values and

those of store management have become more similar. -.03

When someone praises store management, it feels

like a personal compliment. -.13

What store management stands for is important to

me. -.03

I talk up store management to my friends as a great

group to work for or with. -.03

Flow hard I work for store management is directly

linked to how much I am rewarded. .27

My private views about store management are

different than those I express publicly. .38

I am proud of store management. .00

The reason I prefer store management to other

groups is because of what they stand for, that is,

their values. .16

I feel a sense of “ownership” for store management. .03

Notes:

(a.) Retained items arc underlined under the appropriate factor. Non-underlined items were dropped prior to the creation of composite Kales.

(b.) Compliance was treated as an across-foci construct. Underlined items were included in the scale assessing overall compliance without regard to foci.

Direct all correspondence to: Thomas E. Becker, University of Delaware, Department of Business Administration, 306 Purnell Hall, Newark, DE 19716-2710.

COPYRIGHT 1995 JAI Press, Inc.

COPYRIGHT 2004 Gale Group