Abstract
The change probability effect occurs when probable changes are detected more accurately than improbable changes (Beck, Angelone, & Levin, 2004). This effect occurs for familiar everyday objects and can be the result of incidentally learned change probability information for novel objects. The change probability effect has been attributed to biases at the retrieval and comparison stages (Beck et al., 2007) and the decision stage (Yange, Chang, & Wu, 2013). The current study examined whether participants could learn that changes to either high spatial frequency (HSF) or low spatial frequency (LSF) items were more probable and then use this information to bias attention during encoding. Participants were shown arrays of six Gabor patches, three HSF and three LSF, oriented 45째 to the left or the right. After an 800ms presentation of the study array, a test array was presented in which one of the patches changed orientation. Participants indicated which of the six items had changed. For half of the participants, HSF items changed on 80% of the trials (HSF probable condition) and for the other half, the LSF items changed on 80% of the trials (LSF probable condition). When possible, the HSF item that changed was an item fixated during study, and the LSF item that changed was an item that was not fixated during study. Participants completed 3 blocks of 40 trials. In the HSF probable condition, performance improved across blocks for HSF changes, but not for LSF changes. In the LSF probable condition, performance improved across blocks for LSF changes, but not for HSF changes. In addition, the amount of time participants spent looking directly at items decreased across blocks for the LSF probable condition, but not for the HSF probable condition, demonstrating a shift in the allocation of attention across blocks in the LSF probable condition.
Meeting abstract presented at VSS 2014