Purchase this article with an account.
Michael Wenger, Douglas Bryant, James Townsend, Ru Zhang, Yanjun Liu; Detecting mean shift integrality using the Hering illusion: initial results using general recognition theory and systems factorial theory. Journal of Vision 2018;18(10):797. doi: https://doi.org/10.1167/18.10.797.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
General recognition theory (GRT; Ashby & Townsend, 1986) and its reaction time (RT) based extensions (Townsend, Houpt, & Silbert, 2012) offer a number of theoretical and empirical strengths with respect to understanding "configural" or "holistic" perception. One difficulty that has arisen is the question of whether a phenomenon known as mean shift integrality (MSI) might pose challenges to the identification of violations of perceptual and/or decisional separability. Unfortunately, there have been no methods available for ascertaining whether MSI might be occurring. The present project represents an initial to address this problem by inducing MSI using a visual illusion. Observers (n = 6) viewed stimuli designed to induce the Hering illusion (see Figure 1, supplmental materials) in two tasks. The first was a complete identification (CID) task, in which observers indicated perceived curvature of vertical lines presented to the left and right of center. Three levels of induced curvature for each set of lines were combined factorially to produce nine alternatives, with each assigned a unique response. The second task was a double-factorial paradigm (DFP) task in which observers gave one response if they perceived both sets of lines to have the highest possible level of curvature and gave a second response to all of the other stimuli. Sufficient trials were run to allow analysis at the individual observer level. Data from the CID task were analyzed using both frequency- and RT-based measures of marginal response invariance and report independence, and the identification/confusion matrixes were used to fit multivariate gaussian identification models. Data from the DFP task were analyzed using the distribution-based measures of processing architecture and capacity (Townsend & Wenger, 2004). Results indicated that MSI was induced for a subset of the observers, and that there were regularities relating measures of capacity to the presence or absence of MSI.
Meeting abstract presented at VSS 2018
This PDF is available to Subscribers Only