Intentionality attribution in trolley-type moral dilemmas across the United States, Japan, and China: a preregistered cross-cultural study

Abstract Intentionality attribution plays an important role in moral cognition, but it remains unclear whether the factors shaping such judgements operate similarly across cultural contexts. This study examined the effects of intent, personal force, and actor–observer perspective on intentionality attribution in trolley-type moral dilemmas across the United States, Japan, and China. In this preregistered between-participants study, 611 participants from China, Japan, and the United States were assigned to conditions in a 2 × 2 × 2 × 3 design manipulating intent, personal force, perspective, and region. Binary responses concerning moral choice, moral acceptability, and intentionality attribution were analysed using logistic regression models. The preregistered tests examined Intent × Region and Perspective × Region interactions and the main effect of Personal Force. Post hoc simulation-based sensitivity analyses evaluated the interaction magnitudes detectable with approximately 80% power under hypothesis-consistent directional patterns. The preregistered cultural interaction hypotheses were not supported. Intentionality attribution was more likely in means than side-effect conditions and in first-person than third-person conditions in pooled analyses, whereas Personal Force did not significantly predict intentionality attribution. Baseline intentionality attribution rates differed substantially across regions. The sensitivity analyses indicated that the study was primarily capable of detecting relatively large cultural moderation effects under the assumed interaction patterns. Secondary, non-preregistered analyses suggested possible differences between the predictor patterns of moral judgement and intentionality attribution, but these findings were exploratory. The study did not detect the preregistered cultural moderation effects. The pooled results identify intent and perspective as predictors of intentionality attribution in the present samples, but the nonsignificant interactions should not be interpreted as evidence of cultural equivalence. Further preregistered research should use designs powered to detect smaller interaction effects and experimentally vary question order.