Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Overestimation of base-rate differences in complex perceptual categories

W T Maddox1, C J Bohil

  • 1Department of Psychology, University of Texas, Austin 78712, USA. maddox@psy.utexas.edu

Perception & Psychophysics
|June 18, 1998
PubMed
Summary

This study found that people often overestimate base-rate differences in perceptual categorization tasks, regardless of stimulus complexity. This suggests a consistent learning bias in how individuals perceive category probabilities.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Sex differences in cognitive performance and alcohol consumption in High Alcohol-Drinking (HAD-1) rats.

Behavioural brain research·2020
Same author

Sex-specific ultrasonic vocalization patterns and alcohol consumption in high alcohol-drinking (HAD-1) rats.

Physiology & behavior·2017
Same author

Toward a dual-learning systems model of speech category learning.

Frontiers in psychology·2014
Same author

Neural correlates of rule-based and information-integration visual category learning.

Cerebral cortex (New York, N.Y. : 1991)·2006
Same author

Separating perceptual processes from decisional processes in identification and categorization.

Perception & psychophysics·2002
Same author

On the relation between base-rate and cost-benefit learning in simulated medical diagnosis.

Journal of experimental psychology. Learning, memory, and cognition·2001

Area of Science:

  • Cognitive Psychology
  • Perception
  • Machine Learning

Background:

  • Perceptual categorization involves assigning stimuli to categories based on features.
  • Decision-bound theory explains categorization by defining boundaries in a stimulus space.
  • Base-rate information (prior probabilities of categories) influences categorization decisions.

Purpose of the Study:

  • To examine the optimality of multidimensional perceptual categorization performance.
  • To investigate how base-rate ratios affect categorization with integral and separable dimensions.
  • To analyze complex category structures and decision bounds.

Main Methods:

  • Experimental sessions with human observers.
  • Analysis at the single-observer level using nested models from decision-bound theory.

Related Experiment Videos

  • Testing continuous-valued analogues of base-rate learning theories.

    • Main Results:

    • Optimal decision bounds were highly nonlinear across all conditions.
    • Observers were sensitive to base-rate manipulations but generally overestimated base-rate differences.
    • Decision-bound theory with accurate category knowledge and overestimated base-rates best explained the data.

    Conclusions:

    • Base-rates are learned similarly regardless of decision bound complexity.
    • A consistent overestimate of base-rate differences was observed in observers.
    • Decision-bound theory provides a robust framework for understanding perceptual categorization with base-rate information.