Tests of General Mental Ability, or GMA, have been used for nearly a century for personnel selection. Abilities commonly measured in GMA tests include reading, spelling, vocabulary, math, reasoning and classifying. Intelligence tests often measure the same abilities.
Items are usually presented in multiple-choice format. GMA tests resemble tests given in academic environments.
Tests of GMA are probably the most widely used of all test types today. There is a good reason for this. They work. That is, people who score high on tests of GMA perform better in training and on the job. Further, they work for a wide variety of jobs, ranging from the simplest to the most complex. This has been demonstrated in hundreds, perhaps thousands, of studies and is no longer in doubt. Not only do they work, but they work better than most other types of tests. Studies consistently yield validity coefficients in the .50s. Such validity coefficients are indeed impressive. Only one other type of test, Work Sample Simulations, produces comparable results.
If tests of GMA are so effective, why is there any controversy? That is, why consider any other type of test? There are several reasons. The principal reason is the large majority-minority differences in GMA scores. European-Americans consistently achieve higher scores on GMA tests than do African-Americans and Latin-Americans. If employment decisions are based on GMA scores, the minority selection rate will often be less than 80 percent of the majority selection rate. Such employment decisions are said to be characterized by "disparate impact." Per the "Uniform Guidelines on Employee Selection Procedures," published in 1978, the U.S. Department of Labor encourages employers to search for alternatives characterized by less disparate impact. This is not to say that valid GMA tests are unlawful, just that these tests tend to attract more scrutiny than tests with less disparate impact. Also, many employers are committed to a diverse workforce. This is a compelling reason to search for alternatives.
There are other considerations as well. Tests of GMA often don't look like the job for which the applicant is applying. That is, they lack "Face Validity." Though not a recognized scientific construct, Face Validity is important. Several studies have shown that applicants react more positively to tests which have an obvious relationship with the job being sought. Applicants with favorable reactions are more likely to accept job offers, less likely to challenge the fairness of the test, and are more likely to be highly motivated to do well on the job. Finally, applicant reactions may reflect reactions of arbitrators, judges, juries and voters.
To summarize, tests of GMA work well. However considerations of disparate impact and applicant reaction encourage the search for alternatives.