Abstract
The linkage process of job analysis requires subject matter experts (SMEs) to rate an overwhelming number of linkages between tasks and KSAOs. Archival data was utilized to explore the capacity of multiple matrix sampling (MMS) theory to reduce SME rater workload while maintaining the desired accuracy and reliability of obtained ratings. SMEs (N = 28) were divided into pairs and each pair was assigned to rate subsets of linkages. MMS helped reduce the total amount of ratings required from each individual. A generalizability study supported the use of MMS such that the variance in the ratings generally reflected true differences in the linkages themselves. The SME raters, their classifications, and the various pairs contributed relatively little to the linkage variance. This study provides a solid basis from which future studies can continue to explore the feasibility of MMS in improving the linkage process.