Replication Research Symposium

Most experimental studies in social psychology rely on dichotomous factor-level designs, testing only the direction of effects rather than their functional form. As a result, replication efforts often reproduce these limited designs, reinforcing a narrow approach to theory testing. We argue that replication efforts should move beyond dichotomous designs to systematically map the shape of effects across a broader range of manipulation intensities. This approach enhances theoretical precision, helps resolve conflicting findings, and explains variation in effect sizes—key challenges in replication research. By incorporating varied manipulation intensities and systematic manipulation checks, we can uncover functional relationships, ensuring that replications contribute to theory development rather than mere effect verification. We propose sampling strategies that support this goal: broad sampling for initial exploration, focused sampling within characteristic ranges, and strategic differentiation of competing functions. Integrating machine learning and formal modeling techniques into replication studies can further refine theoretical predictions. Emphasizing functional relationships in replication research strengthens the empirical foundation of psychological theories, moving the field toward more predictive and generalizable models.