NotesFAQContact Us
Search Tips
ERIC Number: ED310173
Record Type: Non-Journal
Publication Date: 1988
Pages: 83
Abstractor: N/A
Reference Count: N/A
Cross Validation of Job Families Using an Expanded Data Set. USES Test Research Report No. 53.
Swarthout, David
The analyses of J. E. Hunter (1983) were replicated with an expanded data set. The Hunter study, the basis of the Validity Generalization system used by the United States Employment Service, contained 515 General Aptitude Test Battery validation studies. The data set in this study included these and additional studies to bring the data set to 755 studies representing 74,187 individuals. Analyses were conducted for the total sample, the original sample, and the new or additional sample. The beta weights developed by Hunter were compared with the beta weights developed for the total sample. Results indicate that the beta weights developed by Hunter in 1983 and those developed from the total sample yield virtually identical validity coefficients. Consequently, there is no need to change the beta weights currently being used. A secondary conclusion is that the validities for the additional sample are lower than those for the original sample, particularly for the perceptual or psychomotor aptitudes. There is no easy explanation for the decline in validities, but several suggestions are advanced. Eight tables present study information. An appendix contains tabulated data on the 755 studies that constitute the total sample. (SLD)
Publication Type: Reports - Research
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: Employment and Training Administration (DOL), Washington, DC.
Identifiers - Assessments and Surveys: General Aptitude Test Battery