Two experiments tested whether a peak-shifted generalization gradient could be explained by the averaging of distinct gradients displayed in subgroups reporting different generalization rules. Across experiments using a causal judgment task (Experiment 1) and ... ...
