The effects of behavioral objectives and/or criterion test items on the learning process were investigated. The 123 subjects were randomly assigned to either an example-only, an objective-example, a test-example, or an objective-test-example group. Objectives significantly increased the amount of time subjects spent studying the example displays. A significant ability by treatment interaction revealed that display latency had a negative relationship to reasoning ability for subjects in the test-example and objective-test-example groups, but was not related to reasoning for subjects in the example-only and objective-example groups. Differential relationships between state-anxiety and treatments were also observed. (Author/JK)