Purchase this article with an account.
Sarah Cook, Diana Perez, Mary Peterson; An Investigation of the Effect of Prediction on Object Perception. Journal of Vision 2017;17(10):1241. doi: 10.1167/17.10.1241.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Can prediction affect the perceived sharpness of an object's edges? We tested whether a masked word prime affected the point at which two stimuli with borders varying in blur were perceived as equally blurry (the Point of Subjective Equality). Participants viewed 180-ms exposures of two small, enclosed black silhouette objects on a gray background. The Test object varied in blur, from low to high levels across trials, while the Standard object remained at a constant medium blur value. In Experiment 1, one object was familiar; the other (PR) was novel, created by rearranging the parts of the familiar object. Both objects were Test and Standard objects equally often. Before the object display, a prime word (related or unrelated to the Test) was presented outside of awareness. Participants reported which object appeared blurrier. We hypothesized that the familiar Test would appear sharper when preceded by a related vs an unrelated prime. Experiment 2 was identical to Experiment 1, with small changes in masking procedures. Neither experiment revealed an effect of prime. However, in both experiments, the familiar object was perceived as sharper than the PR object (Exp 1, p = < .001, n = 17; Exp 2, p = < .001, n = 13). These results may be due to the activation of object memories by the familiar Test, and subsequent modulation of the mismatch between predicted sharp edges and blurry Test edges. Alternatively, they could be due to response bias, with participants biased to choose the PR object as blurrier when there were small differences in blur. Experiment 3, currently underway, uses a same/different judgment to eliminate response bias. If we still find that familiar objects are perceived as sharper, these results will support the claim that percepts of familiar objects are not veridical representations of stimulus information, but are biased by object memories.
Meeting abstract presented at VSS 2017
This PDF is available to Subscribers Only