In studies of the relationship between masculinity-femininity and intelligence, data should be analyzed using correlational analysis for the complete distribution, rather than analysis of variance of extreme groups. The earlier work of Baucom and Welsh (EJ 171 822) was used as illustration. (BW)

Humphreys' criticisms (TM 503 269) of the work of Welsh and Baucom (EJ 171 822) are inappropriate since his findings are based upon a different measure of masculinity-femininity and a population which is somewhat different. Also, his insistence on correlational analysis to the exclusion of extreme group design is viewed as inappropriately…