Purchase this article with an account.
David Alexander Kahn, Geoffrey Karl Aguirre; Stimulus Similarity and Dimensionality in the Processing of Non-Face Objects. Journal of Vision 2011;11(11):869. doi: https://doi.org/10.1167/11.11.869.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Absolute identification is limited to 7 ± 2 unidimensional exemplars (Miller. 1956). While neither practice nor increased gamut removes the limit, multidimensional exemplars do (Shiffrin & Nosofsky, 1994). We have recently shown P200 evoked potential amplitude indexes the visual dissimilarity of sequential stimuli drawn from a linear morphed continuum of 5 faces (Kahn et al., 2010, JOV). We hypothesize that the neural dissimilarity of stimuli of the same physical difference will be ameliorated in the presence of 9 linear exemplars, and restored with 9 exemplars drawn from two dimensions.
We designed a two dimensional stimulus space of radial frequency contours. We selected 5 and 9 exemplar linear subsets (5L and 9L) of these stimuli – with each subset having equal spacing between the exemplars – as well as a two-dimensional, 9 exemplar set (9S) in which the diagonal spacing between the stimuli was equal to that of the linear spaces. Subjects viewed the stimuli (blocked by exemplar space) in a continuous, counter-balanced stream (1s per stim with ITI of 250–350 ms) while they monitored for the infrequent appearance of a target stimulus from outside the space. 128 channel ERP was recorded and electrodes responsive to formed visual stimuli were identified in a separate localizer study.
In preliminary data from 3 subjects, we have confirmed that the P200 is modulated by the perceptual similarity of sequential stimuli in the 5L space, extending our previous result to non-face stimuli. This modulation is disrupted in the 9L space and reformed in the 9S space, despite the smaller physical difference between the stimuli. Further analyses will examine the representation of prototype within these spaces. We interpret these results within the context of a a flexible model of neural tuning, sensitive to the gamut and distribution of stimulus context.
This PDF is available to Subscribers Only