Are Analysts' Occupational Ability Requirement Ratings Necessary?: A Look at Using Other Occupational Descriptors to Capture the Rating Policy of Analysts

Abstract

The Occupational Information Network (O*NET) contains descriptors for a number of content domains. Trained analysts have rated over 1,100 occupations on those descriptors. The present study focused on four of the O*NET descriptor domains: knowledge, skills, generalized work activities (GWAs), and abilities.

The ability domain was previously identified by a panel of experts as being more abstract and difficult to rate than other descriptor domains. This study addressed that issue by running regression analyses using factors derived from knowledge, skill, and GWA ratings to predict ratings on the ability descriptors. The predicted ability ratings were then factor analyzed and compared to factors derived from the actual ability ratings.

Although all of the resultant multiple correlations were statistically significant, they were not all of sufficient magnitude to justify replacing actual ability ratings with ability ratings estimated from the more concrete domain descriptors. It is likely that the R's for many of the abilities were attenuated by unreliability in their ratings. In general, the cognitive abilities proved to be more predictable than the motor and perceptual abilities. It would appear practical to estimate requirements for some but not all of the abilities with ratings on the other domain descriptors. Factors derived from the predicted ability ratings showed some similarity to those derived from the actual ability ratings, thus lending further support to the validity of the predicted ratings.