The impact of raters' cognition on judgment accuracy: An extension to the job analysis domain

Eighty-six incumbents of three different jobs produced job-analytic ratings using either a decomposed (task-based) or a holistic (job-based) rating strategy. Approximately half of them received rater training in making inferential decisions. When tasks were less complex than the job as a whole, rating decomposition generally had positive effects on ratings' quality. Similarly, when the number of tasks rated was low to moderate, rater training was effective. A contingency approach, where limitations concerning the use of rating decomposition and inferential training were outlined, should serve to inform future uses and theories of rating aids in job analysis.We would like to acknowledge David Dorsey for his significant contribution to the rater training program. This article is a summary of the doctoral dissertation of Juan Sanchez, which was conducted under the supervision of Edward Levine.