ABSTRACT
Item discrimination refers to "the degree to which an item differentiates correctly among examinees in the behavior that the test is designed to measure" (Anastasi, 1988, p. 21 0). The point biserial correlation technique, in which each item is correlated with the total test score, was used to determine item discrimination (sometimes called discriminating power). Unfortunately, the literature gives little guidance regarding the magnitude of acceptable discriminating powers. This being the case, we applied Anastasi's (1988) suggestions regarding the interpretation of validity coefficients to the interpretation of discriminating powers. She suggests that statistically significant coefficients of .2 or .3 can be considered acceptable.
