LAUSR

Understanding what leads people to reverse their choices is important in many domains. We introduce a contrast paradigm for studying reversals in choices—here between pairs of abstract paintings—implemented in both within-subject (Experiment 1; N = 320) and between-subject (Experiment 2; N = 384) designs. On each trial, participants chose between a pair of paintings. A critical pair of average-beauty paintings was presented before and after either a reversal or control block. In the reversal block, we made efforts to bias preference away from the chosen average-beauty painting (by pairing it with more-beautiful paintings) and toward the non-chosen average-beauty painting (by pairing it with less-beautiful paintings). Meta-analysis revealed more reversals after reversal blocks than after control blocks, though only when the biasing manipulations succeeded. A second meta-analysis revealed that reversals were generally more likely for participants who later misidentified their initial choice, demonstrating that memory for initial choices influences later choices. Thus, the contrast paradigm has utility both for inducing choice reversals and identifying their causes.