Fifty-three high school males took a minicourse on library usage. All were randomly assigned a topic and given response sheets requiring the use of dictionaries, encyclopedias, periodicals, and books. In a discussion session, all received instructional handouts. Students were assigned either to evaluating two peers' pretests (experimental) or researching two new topics (control). New topics were randomly assigned for the posttest. Both groups mastered 50 percent more objectives on the posttest than on the pretest. No significant differences between the gains for the two groups were found. Evaluating a peer's work was as effective as performing additional tasks. (Author)