What Causes What: Causal Directionality Moderates Evaluative Conditioning Effects

Past research showed that Evaluative Conditioning (EC) effects are shaped by the framing of the relationship between conditioned stimuli (CS) and unconditioned stimuli (US). We tested if causal directionality—whether the CS causes the US or the US causes the CS—moderates EC effects. Across two preregistered experiments (N = 680), participants viewed food brands (CSs) paired with facial expressions (USs). In Experiment 1, the EC effect was stronger in the CS-causes-US than in the US-causes-CS condition, but this emerged only in self-reports. Experiment 2 extended these findings to a health-related context, with expressions framed as physical comfort/discomfort. Stronger EC effects emerged in the CS-causes-US condition on both self-report and Implicit Association Test (IAT). Aggregated analyses suggest that the superiority of the CS-causes-US (vs. US-causes-CS) condition emerged on both measures. These results refine a propositional view of EC, showing that pairing may function as a symbolic cue conveying more than co-occurrence.