This research examines the beneficial effects of student-generated diagrams versus studentgenerated summaries on conceptual understanding in the domain of plate tectonics. Fifty-eight Grade 5 students read a brief expository text about plate tectonics. During their reading of the text, students were asked to either draw diagrams, produce written summaries, or simply read the text (control). Conceptual understanding was measured by the diagrams and summaries which were generated during students' reading of the text, as well as by a posttest which assessed students' understanding of both spatial/static and causal/dynamic knowledge of the domain. Results indicated that the summaries generated during the reading of the text contained more domain-related information than the diagrams which were generated during the reading of the text. However, on the posttest measures, the diagram group outperformed both the summary and text only groups in terms of understanding both the spatial/static as well as causal/dynamic aspects of the domain. Results are discussed with regard to the differential effects that generating diagrams as compared to generating summaries or simply reading has on both on-line comprehension during reading and resulting conceptual understanding of the domain. © 1999 John Wiley & Sons, Inc. J Res Sci Teach 36: 39-53, 1999 Diagrams are widely used in virtually all science domains at all levels of education since they are thought of as an effective way of clarifying subject matter (Lowe, 1989). In school settings as well as in research, they are primarily presented as adjunct aids to text (Dean & Enemoh, 1983;Joseph & Dwyer, 1984;Mayer, 1989) emphasis on interactionist and constructivist methods of learning, diagrams are often given to students with the assumption that the presence of the diagram alone should facilitate learning. Unfortunately, there are several problems which arise when text and diagrams are used for learning, both in school settings and in research. First, many of the diagrams used in science have merely an illustrative rather than explanatory role. Explanatory diagrams are contrasted to illustrative diagrams such that explanatory diagrams seek to explain some aspect of the subject matter, whereas illustrative diagrams serve only to depict certain characteristics (Lowe, 1991). Second, contrary to the belief of science teachers and researchers (Kindfield, 1993), students often do not know how to use diagrams effectively (Gobert, 1994;Lowe, 1989), nor do they know what features are important in diagrams (Anzai, 1991). Finally, learning from a text and diagram may increase memory load (Sweller et al., 1990) or may require additional cognitive effort to integrate the two media (Hegarty, 1992;Yee, Hunt, & Pellegrino, 1991). The result of these problems is that the intended goal, which is for students to develop deeper understanding, is not met (Lowe, 1989).In brief, the existing research on visual displays as adjunct aids to text has demonstrated that they facilitate both recall a...