Purchase this article with an account.
Stephen Killingsworth, Alex Franklin, Daniel Levin; Change Detection is Better Specifically for Object Properties that Change More Frequently in the Real World. Journal of Vision 2011;11(11):153. doi: https://doi.org/10.1167/11.11.153.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Previous work has demonstrated that participants more frequently detect changes to objects that are likely to change in the real world (Beck, Angelone, & Levin, 2004). Furthermore, it has been shown that across blocks, participants can learn which novel objects are more likely to change and subsequently increase fixations to and detect changes faster for these objects (Droll, Gigone, & Hayhoe, 2007). We used a flicker paradigm to test whether detection was faster for frequently changing objects and whether this advantage was specific to changes that were similar to those observed for the objects in the real world. Participants detected orientation and luminance changes to object arrays displayed in naturalistic scene settings or in hexagonal array. A separate group of judges rated the degree to which each of the objects was likely to be physically moved during typical interactions. We observed faster change detection for objects rated as typically moved in the real world, but only for orientation change detection, and not for luminance change detection. This suggests that the detection benefit of knowledge about probabilities is specific to detecting types of changes that match those typically observed for an object. Thus, in addition to learned probabilities influencing where attention is directed (Droll et al., 2007), object-specific knowledge seems to enhance change detection along relevant stimulus dimensions.
This PDF is available to Subscribers Only