The Dilution Effect and Information Integration in Perceptual Decision Making.

Area of Science:

• Cognitive Science
• Decision Making
• Information Integration

Background:

• Human judgment often deviates from optimal behavior when integrating multiple information sources.
• Optimal models, particularly Bayesian ones, successfully explain information integration in perception and memory.
• Processing differences between quantitative/linguistic and perceptual stimuli may explain this paradox.

Purpose of the Study:

• To test if controlled information integration (vs. automatic) leads to suboptimal performance.
• To investigate the dilution effect in a controlled face categorization task.
• To evaluate the Multi-component Information Accumulation model's ability to explain observed effects.

Main Methods:

• Participants categorized faces based on resemblance to a patriarch.
• Evidence from the top and bottom halves of test faces was independently manipulated.
• The study examined accuracy, response time, and dilution effects.

Main Results:

• Splitting faces to encourage controlled integration led to behavior farther from optimal.
• This manipulation also resulted in larger dilution effects.
• The Multi-component Information Accumulation model accurately predicted accuracy, response time, and dilution effects.

Conclusions:

• Controlled information integration, as induced by splitting faces, promotes suboptimal decision-making and amplifies the dilution effect.
• The Multi-component Information Accumulation model provides a unified account of optimal and suboptimal information integration.
• Findings highlight the role of processing type in judgment and decision-making performance.