INVESTIGATION OF COMPONENTS OF JOB SATISFACTION IN A LARGE SOUTHERN UTILITY SYSTEM
Data from an attitude survey administered to 1,150 employees in a southern utility system were examined to discover causes for inconsistencies among reported studies. Two basic conceptual issues were addressed in this study. The first set of analyses concerned technical aspects of scoring and interpreting attitude survey data, and the second set of analyses concerned the examination of specific groups (age, sex, length of service, and organizational-level) to determine whether, despite any quantified difference in level of satisfaction, any differences in the pattern of satisfaction on specific scales was characteristic of these groups. The first set of analyses involved the computation and comparison of unit and regression weights, both within and across scales, to the entire sample and to two independent random subsamples. Three distinctly different scoring strategies were used: (1) the "combined" method, which simply sums all items determined to comprise a scale; (2) the "overall" method, which relies on the use of a "summary item, typically appearing as the final item in a scale; and (3) the "averaged" method, where a score is computed by averaging all items in a scale. This analysis showed that interpretations of the data would be similar either within or across survey scales, regardless of the weighting strategy or scoring method used. The second set of analyses indicated that level of satisfaction on specific scales and global job satisfaction varied characteristically across survey scales for the various work groups (length of service, age, sex, and organizational-level), and that the groups also differed in how well satisfaction on specific scales predicted global satisfaction. In general, however, the analyses showed that these work groups did not exhibit significantly different patterns of job satisfaction on the survey scales. Factor analysis results cast some doubt on the construction of the items or the placement of specific items in the various scales. The results showed that, regardless of the scoring method used, only one factor emerged. This finding is inconsistent with a wide body of other research, which has typically identified a large general factor along with the presence of several auxiliary factors.
CONNOLLY, PAUL MARTIN, "INVESTIGATION OF COMPONENTS OF JOB SATISFACTION IN A LARGE SOUTHERN UTILITY SYSTEM" (1983). ETD Collection for Fordham University. AAI8323518.