Abstract

ABSTRACTTwo studies are reported. The first examined the factors that predict differences in item responses of two populations (Black and matched White examinees) to GRE analogies. Three factors were found to independently predict differential item responses of these two groups to 234 GRE analogy items: (1) item difficulty, (2) analogy stems with a part/whole relationship and (3) and analogy stems with an “attribute” relationship. An independent sample of 220 SAT analogy items was found to yield very similar results, especially for item difficulty and part/whole relationships. For both GRE and SAT analogies, Black examinees were found to do differentially better than matched White examinees on the hard analogy items.Since item difficulty was found to be the most important predictor variable in Study I, Study II explored the possible importance of item difficulty as a predictor of differential item responses of Black versus matched White examinees for three other verbal item types of both the GRE and SAT tests (these types were antonyms, sentence completions, and reading comprehension items). Item difficulty was found to be an important predictor of the observed differences. An overview of the results suggests that the amount of verbal context might be an important determinant of the magnitude of the relationship between item difficulty and differential performance of Black versus matched White examinees. That is, analogies and antonyms (which have minimal verbal context) produced a stronger correlation of difficulty with differential ethnic performance than did sentence completion and reading comprehension items. This was found for both the GRE and SAT verbal item types. Several hypotheses are advanced to account for some of the observed ethnic differences.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call