Purchase this article with an account.
Feitong Yang, Sheng Li; Learning of uncertain stimuli transfers from criterion-based to noise-based perceptual decision, but not vice versa. Journal of Vision 2011;11(11):806. doi: 10.1167/11.11.806.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Making meaningful judgment under various uncertain situations plays an important role in our daily life. It remains unclear whether there is a general neural mechanism dealing with different types of uncertainty. Investigating the transfer of learning effects between different uncertainties paves the way to the fully understanding of this issue. Previous behavioral literatures have shown that uncertainty in decision making can originate from different sources. Converging evidences from fMRI studies have also indicated the neural correlates associated with particular types of uncertainty. In the present study, we generated a two dimensional stimulus space of Glass patterns incorporating uncertainties of criterion-based or noise-based origins and tested the transfer of the learning effect between them. For the criterion-based dimension, intermediate spiral patterns were generated by morphing between radial (0 deg of spiral angle) and concentric (90 deg of spiral angle) patterns (Li, Mayhew, & Kourtzi, 2009). For the noise-based dimension, different proportions of noise were embedded in radial and concentric patterns. We trained two groups of participants in four successive days on categorizing the Glass patterns along the criterion-based and the noise-based dimensions respectively, and tested on the uncertain stimuli of both dimensions before and after the training. Our psychophysics experiments showed that the learning effect can transfer from criterion-based dimension to noise-based dimension, but not vice versa. Especially, the transfer happened only when participants were aware of the two stimulus dimensions. Further fMRI experiment in which participants encountered both uncertainty dimensions showed that the two uncertainties activated different brain areas. That is, the criterion-based uncertainty induced more activation in ventral prefrontal cortex whereas noise-based uncertainty led to more activation in cingulated cortex. Altogether, the current results suggested that two separate but not independent neural networks are engaged in processing the different sources of uncertainty.
This PDF is available to Subscribers Only