Using Similarities and Differences: A Meta-Analysis of Its Effects and Emergent Patterns

The purpose of the study was to update previous meta‐analytic findings on the effectiveness of using similarities and differences as an instructional strategy. The strategy includes facilitating student comparison, classification, use of analogies, and use of metaphors. Previously, Marzano, Pickering, and Pollock reported a mean effect size of 1.61. For the present meta‐analysis, literature was searched to locate experimental studies meeting the following inclusion criteria: published between 1998 and 2008; examined effects of facilitating student comparison, classification, use of analogies, and use of metaphors; measured student academic achievement as an outcome; involved students in kindergarten through grade 12; and provided data necessary to compute effect sizes. Based on the eligible research, the overall mean effect size (Hedges' g) was .65, confirming earlier results indicating that using similarities and differences positively influences student achievement. Type of control group, however, moderated the effect. Larger effect sizes were associated with control groups receiving textbook‐guided instruction versus those receiving interactive instruction. Emergent patterns were observed for the positive influence of long‐term instruction, systematic instruction, supportive cuing, and opportunity for reflection and discussion. Results support recommendations to guide students through analogical reasoning about, and classification of, important concepts and relationships in content‐area instruction.