September 2015
Volume 15, Issue 12
Free
Vision Sciences Society Annual Meeting Abstract  |   September 2015
Sensory measurement and motor planning are not separable in interval timing
Author Affiliations
  • Evan Remington
    McGovern Institute for Brain Research, Massachusetts Institute of Technology Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology
  • Mehrdad Jazayeri
    McGovern Institute for Brain Research, Massachusetts Institute of Technology Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology
Journal of Vision September 2015, Vol.15, 977. doi:10.1167/15.12.977
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Evan Remington, Mehrdad Jazayeri; Sensory measurement and motor planning are not separable in interval timing. Journal of Vision 2015;15(12):977. doi: 10.1167/15.12.977.

      Download citation file:


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

      ×
  • Supplements
Abstract

The brain must integrate various sources of information and use that information to guide actions. Bayesian models of sensorimotor function describe how an ideal observer integrates information to generate behavioral responses that maximize the probability of a desired outcome. In most Bayesian models, it is assumed that the computations associated with estimating the state of the environment are distinct from those that generate the desired motor plans. Implicit in this formulation is the prediction that sensory estimates can be flexibly applied to different sensorimotor transformations. For example, a typical Bayesian observer would bias its sensory estimates based on prior knowledge of stimulus statistics independent of how the estimate will guide behavioral responses. We tested this prediction by asking human subjects to perform a time measurement and production task similar to a previous study (Jazayeri and Shadlen 2010). Subjects had to measure a sample interval that was drawn from a prior distribution, then immediately produce an interval that was equal to the sample interval multiplied by a gain factor. To test whether sensory estimation and motor planning computations are independent, each subject participated in multiple experimental sessions in which the sensory estimation parameters were kept fixed while the corresponding motor planning parameters were changed. In particular, across sessions, the prior distribution of the sample intervals was held constant while the gain factor was varied. We found that the effect of sensory prior on behavior depended on the gain factor suggesting that the computations underlying sensory estimation are not independent of the motor plan. Results were not explained by variants of the Bayesian model that consider additional uncertainty about the gain factor and/or additional noise in the production epoch. These findings suggest that, at least in this task, sensory estimation is not computationally separable from motor planning.

Meeting abstract presented at VSS 2015

×
×

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.

×