ABSTRACT:In this study, we investigated whether the understanding of the particulate nature of matter by students was improved by allowing them to design and evaluate molecular animations of chemical phenomena. We developed Chemation, a learner-centered animation tool, to allow seventh-grade students to construct ßipbook-like simple animations to show molecular models and dynamic processes. Eight classes comprising 271 students were randomly assigned to three treatments in which students used Chemation to (1) design, interpret, and evaluate animations, (2) only design and interpret animations, or (3) only view and interpret teacher-made animations. We employed 2-factor analysis of covariance and calculated effect sizes to examine the impact of the three treatments on student posttest performances and on student-generated animations and interpretations during class. We used the pretest data as a covariate to reduce a potential bias related to students' prior knowledge on their learning outcomes. The results indicate that designing animations coupled with peer evaluation is effective at improving student learning with instructional animation. On the other hand, the efÞcacy of allowing students to only design animations without peer evaluation is questionable compared with allowing students to view animations.