Abstract
Previously we assessed whether category expectations (e.g. object-color) that strongly-matched (e.g. yellow banana), weakly-mismatched (orangish-yellow banana), or strongly-mismatched (purple banana) study information boosted recall accuracy over information with no-expected match (yellow shirt). We found a boost in color recall accuracy for strong-matches suggesting prior expectations helped boost memory accuracy scores relative to mismatches or no-expectations. However, this past work was unable to shed insight on two hypotheses: (1) Error caused by the mismatch was offset by a surprisal memory boost, and (2) A simplified “least bits” coding mechanism (e.g. “off-yellow banana”) was used for matches and weak mis-matches, allowing for easier encoding over strong-mismatches and no-matches. Here we assessed whether the limited boost of having object-color expectations on recall accuracy extends to other features of the object (e.g. object-shape). If in the prior study participants were using a “least bit” coding strategy when prior expectations exist, then shape memory may not suffer as much as no-expectation or mis-match. However, if color mismatches caused greater attention to overall object properties including shape, then color mis-matches should see improved shape memory relative to both no-expectation and expectation cases. Results revealed no significant differences between strong-match and either mismatches, with worse performance for no-expected matches. These results lend support that participants rely on prior expectations to simplify encoding, which are not available when prior knowledge cannot be leveraged. Interestingly, these results also support the surprisal boost hypothesis, suggesting that (at least in the context of color-object pairings) surprisal response in seeing a color-object mismatch triggers an increase in overall memory for object-shape matches over no-expectation conditions, but is not strong enough to out-perform the boost gained by expectation matches. Thus, category expectations and surprise may play a role in how features of objects are bound together in visual memory.