Gradual growth versus shape invariance in perceptual decision making.

Area of Science:

• Cognitive psychology
• Computational neuroscience
• Decision making

Background:

• Models of human perceptual judgment often assume sensory neural activity integrates until a boundary is reached.
• These models predict changes in response time distribution shapes across conditions, though these changes can be subtle.
• An alternative perspective posits shape invariance in response time distributions, as seen in some race models.

Purpose of the Study:

• To competitively assess a one-bound diffusion model (gradual growth) against a shape-invariant inverse Gaussian distribution model.
• To evaluate subtle shape changes versus shape invariance in response time distributions.
• To determine which model better explains human perceptual judgment data.

Main Methods:

• Employed a Bayesian approach to pool information across multiple participants for robust assessment.
• Competitively evaluated the one-bound diffusion model against a shape-invariant inverse Gaussian distribution model.
• Analyzed response time distributions to detect subtle shape changes.

Main Results:

• Found subtle shape changes in response time distributions.
• These observed shape changes were highly concordant with the predictions of a simple diffusion gradual growth model.
• The results were discordant with the predictions of a shape-invariant model.

Conclusions:

• Subtle shape changes in response time distributions provide evidence supporting gradual growth models.
• The findings challenge the notion of strict shape invariance in response time distributions for perceptual judgments.
• The one-bound diffusion model offers a better account of human perceptual judgment data compared to shape-invariant models.