Measuring General Job Satisfaction: Which Is More Construct Valid—Global Scales or Facet-Composite Scales?

Abstract

In order to assess general job satisfaction, researchers often use composites formed by summing scores from multiple satisfaction facets. The appropriateness of composites, however, is a matter of contention: some researchers have argued that composites provide an effective means of assessing general job satisfaction (e.g., Spector, American Journal of Community Psychology, 13, 693-713. 1985), whereas others have argued that composites are inappropriate and should be replaced with global job satisfaction scales (e.g., Ironson et al., Journal of Applied Psychology, 74, 193-200. 1989). To address this debate, we compared the construct validities of composite job satisfaction scales with those of global job satisfaction scales. We first created a nomological network that specifies a hypothetical pattern of relationships that exists between general job satisfaction and several external variables. Using this network as a guide, we conducted two studies (total N = 676) that examined the construct validities of four global scales and five composite scales. Within both studies, we found that composite satisfaction scales and global satisfaction scales generally yielded similar relationships with the external variables included in the nomological network. This suggests that the two types of scales are equally effective at assessing general job satisfaction. Researchers, therefore, should not summarily disregard previous studies that have used composite measures. These findings also suggest that it is appropriate to use composite measures in instances where global satisfaction data are unavailable, a situation that can occur when working with archival or organizational datasets.