Scale (In)Variance in a Unified Diffusion Model of Decision Making and Timing

Abstract

Weber’s law is the canonical scale-invariance law in psychology: when the intensities of 2 stimuli are scaled by any value k, the just-noticeable-difference between them also scales by k. A diffusion model that approximates a spike-counting process accounts for Weber’s law (Link, 1992), but there exist surprising corollaries of this account that have not yet been described or tested. We show that (a) this spike-counting diffusion model predicts time-scale invariant decision time distributions in perceptual decision making, and time-scale invariant response time (RT) distributions in interval timing; (b) for 2-choice perceptual decisions, the model predicts equal accuracy but faster responding for stimulus pairs with equally scaled-up intensities; (c) the coefficient of variation (CV) of decision times should remain constant across average intensity scales, but should otherwise decrease as a specific function of stimulus discriminability and speed–accuracy trade-off; and (d) for timing tasks, RT CVs should be constant for all durations, and RT skewness should always equal 3 times the CV. We tested these predictions using visual, auditory and vibrotactile decision tasks and visual interval timing tasks in humans. The data conformed closely to the predictions in all modalities. These results support a unified theory of decision making and timing in terms of a common, underlying spike-counting process, compactly represented as a diffusion process. (PsycINFO Database Record (c) 2015 APA, all rights reserved)

Publisher

US: American Psychological Association

Publication Date

3-1-2016

Publication Title

Psychological Review

Department

Neuroscience

Document Type

Article

DOI

https://dx.doi.org/10.1037/rev0000014

Keywords

Diffusion, Decision making, Interval timing, Weber, Scalar invariance

Language

English

Format

text

Link to Full Text

Share

COinS