Additional construct validity evidence for the task/contextual performance distinction

Borman and Motowidlo (1993) hypothesized two job performance domains labeled task performance and contextual performance. Previous research has provided construct validity evidence in the form of differential predictor-criterion relations for the two performance domains. This study provided additional construct validity evidence. I examined internal consistency and empirical independence of the two domains by analyzing multitrait-multirater (MTMR) matrices. Confirmatory factor analyses (CFAs) of 15 MTMR matrices showed higher correlations for performance dimensions within a domain (mean correlations were .70 and .70 for task and contextual performance, respectively) than between domains (M = .55). Evidence showed that the distinction between task and contextual performance was more clear for nonmanagerial than for managerial jobs. One aspect of contextual performance, cooperating with others, showed greater independence from task performance than did persisting with extra effort or following organizational rules. These results support the task and contextual performance model but suggest that the domains are not completely independent, and that independence varies by job type and type of contextual dimension. There was no evidence of differences in the reliability of task versus contextual performance ratings.