The present study compares the performance of volunteers and nonvolunteers in a 10-hour integrated series of management simulation exercises. It was hypothesized that there would be no performance differences in these two groups. Subjects were business students. Dependent measures were global ratings of management level and potential, as well as a behavioral checklist score. The results partially support the hypothesis. Volunteers performed significantly better on the global measures but not on the behavioral checklist. Differences on global measures were attributed to demand characteristics, whereas the behavioral checklist was not susceptible to this artifact. This finding raises doubts as to the usefulness of global measures and the validity of research based on them. Studies designed to upgrade and validate the behavioral checklist are needed.