Study time and test performance change as a function of expecting either open or multiple-choice questions on a history test. Subjects tested in either format were led to expect the same test format on a second test. Subjects expecting open questions studied more and performed better on both test formats. (Author/CM)