Exemplar theory's predicted typicality gradient can be tested and disconfirmed

Area of Science:

• Cognitive Psychology
• Computational Neuroscience

Background:

• Categorization models explain how humans form and use categories.
• Exemplar theory posits categorization based on similarity to stored specific examples.
• Typicality gradients reflect how representative an item is within a category.

Purpose of the Study:

• To evaluate the predictive accuracy of exemplar theory regarding typicality gradients.
• To compare exemplar theory's predictions with empirical data from a dot-distortion task.
• To assess the viability of exemplar theory in light of empirical categorization findings.

Main Methods:

• Analysis of typicality gradients predicted by exemplar theory.
• Comparison of theoretical predictions with empirical data from a dot-distortion categorization task.
• Evaluation of a flexible and mathematically powerful exemplar model.

Main Results:

• Exemplar theory's predicted typicality gradients do not match empirical data.
• This discrepancy persists even with advanced versions of exemplar models.
• Prototype theories demonstrate consistency with observed typicality gradients.

Conclusions:

• Exemplar theory is disconfirmed in the domain of categorization studied.
• Empirical typicality gradients support prototype-based categorization models over exemplar-based ones.