Over the past 20 years, there has been much research on how people learn from case comparisons. This work has implemented comparison activities in a variety of different ways across a wide range of laboratory and classroom contexts. In an effort to assess the overall effectiveness of case comparisons across this diversity of implementation and contexts and to explore what variables might moderate learning outcomes, we conducted a meta-analysis of 57 experiments with 336 tests. Random effects analyses of the 336 tests revealed that case comparison activities commonly led to greater learning outcomes than other forms of case study including sequential, single case, and nonanalogous, as well as traditional instruction and control (d = .50), 95% CI [.44, .56]. Of 15 potential moderators, four variables were found to reliably moderate the effectiveness of case comparisons: the objective of the comparison, the presentation of a principle, the content, and the lag between the comparison and testing. Asking learners to find similarities between cases, providing principles after the comparisons, using perceptual content, and testing learners immediately are all associated with greater learning. We conclude with a discussion of the theoretical and practical implications of these results for cognitive theory and classroom practice.
Stored only when both papers are in the corpus — 322 citations exist upstream.