Purchase this article with an account.
Aisha P. Siddiqui, Benjamin McDunn, James M. Brown; Seeking the boundary for boundary extension. Journal of Vision 2012;12(9):1074. doi: 10.1167/12.9.1074.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Boundary extension (BE) is a phenomenon in which participants’ memory about the extent of the borders in a previously viewed scene is biased towards including additional scene information. Current theories suggest this error of commission is a source monitoring error between actual scene material and participants’ representation of the scene containing actual and expected scene material. According to this view, when participants make an error they are biased to chose the representation consisting of expected scene information; in other words, they extend the boundaries of the scene with expected scene material. In an earlier study (VSS, 2011), we sought to evaluate the necessity of expectation by removing cues thought to encourage amodal continuation and hence BE (e.g., familiarity, occlusion, semantic content, texture gradients) by using abstract shapes on random dot backgrounds. Despite removing semantic content, BE was predicted only when occlusion cues were evident at the borders of our images (e.g., when an expectation of continuation was present). The overall results showed BE in all conditions regardless of expectation; however, in a few conditions we could not eliminate the possibility of boundary restriction. The current study, using a new, larger set of abstract stimuli with enlarged background dots, showed BE in all conditions and no evidence of boundary restriction or normalization. Thus according to current theories, even though BE should not have occurred when expectation of extension was absent (e.g., no occlusion at borders, no background dots) we have now consistently shown it with these type stimuli. Perhaps, BE is even more ubiquitous than initially thought and may represent a fundamental aspect of visual perception. Additional experiments will be discussed in which we parametrically vary other stimulus variables including the percentage size change between close and wide angle versions of images with and without objects.
Meeting abstract presented at VSS 2012
This PDF is available to Subscribers Only