Recognition-based inference: When is less more in the real world?

Area of Science:

• Cognitive Psychology
• Decision Making
• Heuristics and Biases

Background:

• The recognition heuristic suggests that recognizing more objects generally leads to better inferences.
• A counterintuitive
• less-is-more
• effect exists where recognizing fewer objects can yield more accurate inferences.
• This effect is often explained by recognition validity exceeding knowledge validity.

Purpose of the Study:

• To investigate how dependencies between validity and the number of recognized objects (n) affect the
• less-is-more
• effect.
• To identify cognitive and ecological factors contributing to these validity dependencies.

Main Methods:

• Analysis of empirical datasets to identify validity dependencies.
• Computer simulations to model the impact of dependencies on the
• less-is-more
• effect.

Main Results:

• Validity dependencies (correlations of validities with n) are common in real-world data.
• These dependencies significantly limit the occurrence and strength of the
• less-is-more
• effect.
• Cognitive and ecological factors can influence the relationship between recognition validity and n.

Conclusions:

• The assumption of independence between validity and n is often violated, challenging the universality of the
• less-is-more
• effect.
• Understanding validity dependencies is crucial for accurately predicting inferential accuracy in real-world scenarios.