Three conditions for administering creativity tests by Torrance and by Wallach and Kogan were compared: (1) untimed, gamelike; (2) conventional testlike; and (3) administration of measures under testlike conditions on two adjacent days, using the second testing as the predictor. The conventional testlike condition seems optimal. (Author/CP)

The relationship between test anxiety and performance on oral comprehensive examinations was studied, with special attention to the repressor phenomenon. Repressors are persons who report low anxiety, but are actually more stressed than are low anxious subjects, as evidenced on measures not relying on self-report. Subjects were 42 female graduate…

Test anxiety scores from a 15-item internally consistent instrument were correlated with performance on an aggregate science test administered to 171 fourth- and 187 eight-grade students subdivided by sex and race. The interpretation of data possibly suggest that test anxiety is operating differentially for black and white students. (Author/PN)