Generalizability of General Education Development ratings of jobs in the United States

Abstract

Examined the reliability of the US Department of Labor's rating of educational requirements of occupations. On 2 occasions, 71 field analysts at 11 field centers used the General Education Development (GED) scale to evaluate the reasoning, mathematics, and language requirements of 27 occupations. Jobs were classified by the length of the job description. The major source of variability proved to be Ss' ratings, whereas the occasion of measurement and the location of the center contributed little to the variation among ratings. Over all jobs, generalizability analyses showed that generalizable GED ratings were obtained using the average ratings of 4 raters on 1 occasion. Similar results were obtained from the generalizability analyses for each length of job description. Further, ANOVAs revealed that ratings did not differ according to the length of the job description. Thus, the questionable validity of ratings and the instability of estimates of variance components are noted. (7 ref) (PsycInfo Database Record (c) 2025 APA, all rights reserved)