September 2019
Volume 19, Issue 10
Open Access
Vision Sciences Society Annual Meeting Abstract  |   September 2019
Decision threshold in a perceptual task is influenced by information content of a pre-training stimulus
Author Affiliations
  • Tyler Barnes-Diana
    Brown University
  • Yuka Sasaki
    Brown University
  • Takeo Watanabe
    Brown University
Journal of Vision September 2019, Vol.19, 145. doi:https://doi.org/10.1167/19.10.145
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Tyler Barnes-Diana, Yuka Sasaki, Takeo Watanabe; Decision threshold in a perceptual task is influenced by information content of a pre-training stimulus. Journal of Vision 2019;19(10):145. https://doi.org/10.1167/19.10.145.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

It has been found that minor differences in pre-training procedure have lasting effects on performance on a perceptual task (Barnes-Diana, Sasaki & Watanabe, 2016). Using hierarchical Bayesian parameter estimation of the parameters of a drift diffusion model (DDM), we decompose that behavioral effect into the information threshold necessary before making a decision as well as the speed of information accumulation. In the first experiment, we used a peripheral coarse discrimination task in which subjects responded in each trial to a concurrently presented central letter (T or L) and peripheral task (tilted lines left or right from vertical). An example image, presented only once for 100ms prior to training, was varied across two groups of subjects. One group (n=11) was exposed to a peripheral noise patch and the other (n=11) a peripheral vertically oriented stimulus. DDM analysis indicated that the group that received the noise patch had a lower information threshold than the group exposed to the vertical reference. We thus found that a high information threshold was set when the first exposure to experiment-related stimulus had high information content (the vertical reference), whereas a low information threshold was set when the information content was low (the noise patch). However, the DDM analysis of the original experiment estimated non-decision times that were shorter than the interval between stimulus and prompt onset, giving rise to a question as to whether the reaction times reflected a true reaction time or a prediction error of the prompt onset. Accordingly, a second experiment was done with the same task (n=7), shortening the interval between stimulus and prompt onset. The results were consistent with a difference in threshold across groups. From these results we conclude that the decision threshold is influenced by the information content of a pre-training stimulus.

Acknowledgement: NIH R01EY019466, NIH R21EY028329, NIH R01EY027841, BSF2016058 
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×