Repeated exposure to tests composed of memory items was compared to repeated exposure to tests including higher-level cognitive items and to tests composed of both item types. Other design factors were sex, ability level, subject matter sequence (science and social studies), and learning session. (A sample of 2008 eleventh-grade students participated in phase one of the study and a total of 288 participated in the final study). Although significant main effects for the test conditions did not occur, treatment by sex interactions on both memory and higher-level tests indicated that male performance was enhanced by a variety of item types, female performance by memory items. (Author/MLP)