This study investigated the comparative effects of adjunct questions, student self-generated questions, and note taking on learning from a multimedia database. High school students worked individually or in cooperative dyads on a computer-based multimedia unit using a study guide to answer either adjunct questions, generate self-questions, or take notes to accomplish a learning task. There were two criterion measures, a posttest and an attitude survey. Other data collected included informal observations, student interviews and review of the student study guides. The dependent variable was achievement as determined by posttest score. Analyses were conducted using both traditional Analysis of Variance (ANOVA) techniques as well as hierarchical linear modeling (HLM). Because of the dependency of scores of students working in pairs, it was determined that HLM techniques were more appropriate. Results of ANOVA analyses yielded significant differences, however, HLM analyses did not. The correlations of posttest scores of students working in pairs seemed to indicate possible levels of cooperation. Student interview responses and review of student study guides seemed to provide indications of how students utilized the questioning or note-taking strategies while working in this environment. Implications for the design of instruction are discussed. (Contains 5 tables.)