Figure 3.

Percent of trials judged causal as a function of trial number for direct and indirect trial types in Experiment 3. The positive trend line (slope = 0.014) for direct launches demonstrates that participants were slightly more likely to judge direct launching events as causal over the duration of Experiment 3. In contrast, the negative trend line (slope = −0.04) for indirect launches demonstrates that participants were less likely to judge indirect launching events as causal over the duration of Experiment 3.