Test Cases: How 'Race-Norming' Works

The hot-button issue known as "race-norming" or "within-group score conversion" centers on the General Aptitude Test Battery (GATB), a government employment exam used to refer applicants to jobs in the public and private spheres. The GATB consists of 12 subtests that are intended to gauge intelligence as well as cognitive, numerical and motor skills. The vocabulary section features questions like, "Which two words have the same meaning?" Test-takers must choose from a list of words including (a) Open, (b) Happy, (c) Glad and (d) Green.

Sounds simple? (a) Yes or (b) No. If you checked "b," you've answered correctly. But what that means in terms of your overall test score depends on your race or ethnic background. In order to level the playing field for black and Hispanic job applicants, a test-taker's raw score is rated against the results of others in his or her racial or ethnic group. For example, three applicants might each earn a raw score of 300. But their converted scores would look quite different: the black test-taker would score in the top 83 percent of his group, the Hispanic would rank in the top 67 percent and the white or Asian in the top 45 percent. Most employers who receive these scores aren't aware that they are adjusted.

Test-score conversion is not a new concept. In the 1970s, the E.F. Wonderlic Personnel Test Inc., a Northfielde, Ill., testing firm, began including "ethnic conversion tables" with its test packages so that companies could adjust scores. Some 15,000 private companies now administer the Wonderlic exam to about 3 million people each year.

The controversial practice has come under attack in the past. The Reagan administration, which scorned affirmative-action programs, ordered a study of the GATB. The results didn't please critics of score conversion. In 1989, the National Research Council, a branch of the Nation Academy of Sciences, found that the unadjusted exam was not a strong indicator of job performance. The study showed that unconverted scores tended to screen out minority applicants who often prove as competent as their white or Asian counterparts. But that hasn't stopped race-norming opponents from giving the practice a low grade.