skip to main content skip to footer

Scoring a Controlled-Association Test

Hills, John R.
Publication Year:
Report Number:
ETS Research Bulletin
Document Type:
Page Count:
Subject/Key Words:
AT&T, B.F. Goodrich, Detroit Edison Company, IBM, Westinghouse Electric Corp., Association Measures, Controlled Association Test, Engineers, Job Performance, Job Placement, Scoring


Six different scores from a controlled association test have been studied, using a large sample of experienced engineers as examinees, and using criteria of job success and job placement. The test asks subjects to write, in a limited time, as many synonyms as they can to eight common words. None of the six scores from this test appears to be related to job placement, but the five of them which are based on the number of common responses given seem to be related to job success. These five scores vary along a continuum of commonness of response words, from a very stringent definition which includes all responses given. Although most of the validity coefficients computed in this study are well below correlations of .20, it is estimated that by using a revised test format a validity approaching .20 could be obtained in a similar testing situation. Although not spectacularly high, in comparison with the other variables which were studied in the Engineering Graduate Placement Test Research study such a validity coefficient for ratings of success is promising. The most promising score on the Controlled Association Test is the one based on the most stringent definition of commonness. A score based on number of uncommon responses was not significantly related to the criteria.

Read More