“…Our findings on the effects of "response-cost" instructions on multiple-category reasoning are reminiscent of other findings in experiments that have tried to overcome the negative effects of cognitive heuristics in studies of judgmental anchoring (Mussweiler, Strack, & Pfeiffer, 2000), forecasting (Lovallo & Kahneman, 2003), social judgment (Lord et al, 1984), and probability judgment (Lagnado & Sloman, 2004). In each of these cases, the main focus has been on improving judgment and pre-In the present experiments, we tried to ensure that people used categories to make predictions rather than using a noncategorical approach (e.g., feature conjunction) by making it impossible to arrive at a prediction by just looking at feature pairings across exemplars.…”