Learning Categories from Individual Examples
Mentor 1
Caitlin Bowman
Start Date
10-5-2022 10:00 AM
Description
The ability to learn categories and classify new items or experiences is an essential function for everyday life as it allows us to process information and organize our thoughts. However, there has been considerable debate in cognitive psychology about how people represent categories. The prototype model posits that individuals create an abstract image (prototype) of an item that contains all the typical features of individual category members. The exemplar model suggests categories are represented by each individual category member. These two models have each been fit to categorization data for decades, often showing impressive fit. A seminal study by Mack and colleagues (2013) found support for the exemplar model by showing evidence of only exemplar correlates in the brain even when there were no differences between the fits of the models to behavior. The current study adapted the work of Mack et al. (2013) in an attempt to replicate the behavioral findings with analytical technics that might be more sensitive to differences in the relative fits of the prototype vs. exemplar model.
Learning Categories from Individual Examples
The ability to learn categories and classify new items or experiences is an essential function for everyday life as it allows us to process information and organize our thoughts. However, there has been considerable debate in cognitive psychology about how people represent categories. The prototype model posits that individuals create an abstract image (prototype) of an item that contains all the typical features of individual category members. The exemplar model suggests categories are represented by each individual category member. These two models have each been fit to categorization data for decades, often showing impressive fit. A seminal study by Mack and colleagues (2013) found support for the exemplar model by showing evidence of only exemplar correlates in the brain even when there were no differences between the fits of the models to behavior. The current study adapted the work of Mack et al. (2013) in an attempt to replicate the behavioral findings with analytical technics that might be more sensitive to differences in the relative fits of the prototype vs. exemplar model.
