The Terman-Miles M-F test: An examination of Exercises 1, 2, and 3 forty years later

Abstract

This study tested the reliability (test—retest and internal) and validity (ability to differentiate the sexes) of three subtests of a well-known and historically influential masculinity—femininity instrument, the Terman-Miles Attitude-Interest Analysis Test. Exercises 1 (word association), 2 (inkblot association), and 3 (information) of Form A of the Terman-Miles test were administered to 352 subjects, who included both student and older, nonstudent samples. While all three subtests successfully differentiated the sexes, Exercise 3 was the most powerful differentiator. Test—retest and internal reliabilities were variable, with Exercise 3 having the best reliabilities and Exercise 2 having the worst reliabilities. The fakability of the subtests were investigated with an additional sample of 173 subjects. Exercises 2 and 3 were found to be unfakable. Overall, Exercise 3 emerged as the most promising subtest. It is interesting and important to users of the Terman-Miles test that the measure's validity was reconfirmed in a different subculture some 40 years after its original development.