Study 1: Scale Development

The goal of this study was to develop a reliable measure of motives for engaging in OCB. On the basis of extensive literature reviews and professional consultations, an initial set of 110 items concerned with the motives that might underlie OCB was created. ¹ These items were used to create a scale that measured OCB motives, and then the factor structure and the psychometric properties of the scale were examined.

Method Participants

There were two groups of participants. The first group contained 616 undergraduates (424 women, 172 men, and 20 participants who did not provide their gender). About 70% of the participants self-identified as European American, 10% as African American, and 10% as Hispanic; the remainder checked "other" or did not respond to this question. Seventy-five percent of participants were between the ages of 18 and 25 years, 77% were employed full-time or part-time, and 82% of these individuals worked more than 15 hr per week. The second group contained 178 undergraduates (131 women, 41 men, and 6 participants who did not provide their gender). Their demographic and employment characteristics were virtually identical to those of the first group. The jobs held by the participants ranged from low-level service positions to middle-level supervisory ones.

Procedure

The first group filled out the 110-item version of the Citizenship Motives Scale (CMS). The items were preceded by an explanation or a definition of OCB and several behavioral examples of OCBs. Then, using a 6-point scale, the participants rated how important each motive statement would be in their decision to engage in OCB at a job (1 = not at all important , 6 = extremely important ).

Results

Four of the 110 items were eliminated because of extremely skewed distributions (45% of the respondents chose the same response). The remaining 106 items were subjected to an iterated principal-axis factor analysis ( Rummel, 1970 ). An examination of the scree plot of the eigenvalues for the factors led to the decision to extract three factors. These items were then subjected to an oblique rotation. Items that loaded .50 or higher on one factor and no more than .29 on any other factor were retained for further analysis. These decision criteria yielded 79 items. However, because the scale was intended for usage in field settings, we further reduced the number of items it contained. Two criteria were used in the selection of the items for the final version of the scale. The first was factor loadings; the second was the actual content of the items. This process yielded the 30-item CMS presented in Table 1 .

Items loading on the first factor involved a desire by the participants for the company to do well and a desire for the participants to show pride in and commitment to the organization. Thus, this factor was labeled Organizational Concern (OC) motives. The second factor was labeled Prosocial Values (PV) motives because the items with the highest loadings on this factor involved a need to be helpful and a desire to build positive relationships with others. The final factor was labeled Impression Management (IM) motives because the items with the highest loadings on it were concerned with a desire to avoid looking bad to coworkers and supervisors and to obtain rewards. The three factors correlated significantly with one another: OC—PV, r = .53; OC—IM, r = .36; and PV—IM, r = .30.

The 30-item CMS was administered to the second group of participants, and the same principal-axis factor analysis was carried out. All items loaded on the same factors, and the intercorrelations between the pairs of factors were comparable to those found in the first sample. To further evaluate the similarity between the solutions from the two groups, we computed coefficients of congruence or agreement ( Harman, 1976 ) between the factors obtained for each group. The coefficients of congruence were .98 for OC, .97 for PV, and .96 for IM. (Perfect agreement would be 1.00.)

Psychometric Properties

Subscale scores were calculated by giving the 10 items that represented each factor a unit weight of 1 for that subscale, and the remaining items were given a weight of 0. Among the first group of participants, the coefficient alphas for the subscales ranged from .94 (OC) to .91 (PV and IM); average interitem correlations ranged from .59 (OC) to .49 (IM). Among the second group of participants, the coefficient alphas ranged from .92 (OC) to .84 (IM); the average interitem correlations ranged from .54 (OC) to .35 (IM). Finally, 104 members of the second group of participants completed the CMS twice, 2 to 3 weeks apart. The test—retest reliabilities for the three subscales ranged from .82 (IM) to .71 (PV). The factor structure and the psychometric properties of the three subscales were essentially invariant across gender and work status (i.e., full-time vs. part-time). ²

Date: 2015-01-29; view: 882