Armed Forces Qualification Test

The Armed Forces Qualification Test (AFQT) is the name given to a series of tests constantly in use from 1950 to the present. The current AFQT is a composite [Page 45]of subtests from the Armed Services Vocational Aptitude Battery (ASVAB). Constituent members of the composite are verbal and quantitative components, weighted to be of equal variance contribution. The AFQT is used by the U.S. military as a measure of quality and trainability. Although AFQT scores are reported in percentiles, it also has five categories that are used for legal and classification purposes. By law, the American military is not permitted to enlist individuals who score in the lowest category, that is, at the 10th percentile or less, except by direction or legal exception.

The AFQT is the successor to the World War II Army General Classification Test and the Navy General Classification Test. Although these two tests had similar names, their content was not identical. The first AFQT (1950) was equated to a mixed Army-Navy sample based on percentile equivalent scores from these two tests. The American military calls this procedure “calibrating” and used it for many years. All forms of the AFQT were scored on the basis of a normative sample of men in uniform as of 1944.

The 1950 AFQT had four content areas; verbal, arithmetic, spatial, and spatial visualization. In 1953, the content areas of the AFQT were changed to verbal, arithmetic, spatial, mechanical ability, and tool knowledge. The content remained more or less unchanged through the first seven versions of the test. When the ASVAB was adopted (1976), the AFQT became a part of it. In ASVAB versions 5, 6, and 7, the AFQT strongly resembled the numbered AFQT forms. Scores were reported on the 1944 normative sample.

With the implementation of ASVAB forms 8, 9, and 10, the AFQT portion was reworked to add a timed subtest, Numerical Operations. With the implementation of the 1980 nationally representative normative sample in 1984, a problem was encountered with the timed subtests of the ASVAB. A small mistiming of these subtests could and did have large impacts on the resultant scores. The AFQT composite was then changed to be equally verbal and arithmetic in importance. Scores were still reported on the metric of the 1980 normative sample.

In 1997, another nationally representative normative sample was collected, and the ASVAB subtests, including those contributing to the AFQT, were placed on this new metric as of July 1, 2004. The current AFQT still consists of verbal and arithmetic components. It is basically a measure of general cognitive ability.

Although the AFQT is not available for commercial use, it has been used in nonmilitary research, most notably in Herrnstein and Murray's The Bell Curve (1994) and in numerous econometric and occupational studies. It has been offered as an ability measure and related to training and job performance, earnings, educational attainment, and interregional migration.