Generalized Identity Matching to Sample after Multiple-Exemplar Training in Capuchin Monkeys.

Area of Science:

• Behavioral Psychology
• Animal Cognition
• Primate Learning

Background:

• Generalized identity matching-to-sample (IDMTS) is crucial for complex cognitive abilities.
• Understanding how training procedures influence generalized IDMTS in non-human primates is essential for comparative cognition research.

Purpose of the Study:

• To investigate the efficacy of two training procedures, Mask and Shuffled S-, in developing generalized IDMTS in capuchin monkeys.
• To examine the role of "select" relations (sample-S+ control) in achieving generalized IDMTS.

Main Methods:

• Established a multiple-exemplar IDMTS baseline with three adult male capuchins.
• Employed Mask (blank comparison) and Shuffled S- procedures during training and testing.
• Utilized probe trials with novel stimulus sets to assess generalization.
• Analyzed stimulus control relations and the development of "select" relations.

Main Results:

• All capuchins achieved high accuracy on generalized IDMTS tests.
• The Mask procedure resulted in predominant "select" relations.
• Monkeys trained with the Shuffled S- procedure showed higher accuracy on Shuffled S- IDMTS trials compared to non-shuffled trials.
• "Select" relations were identified as a positive predictor for generalized IDMTS.

Conclusions:

• Both Mask and Shuffled S- procedures can facilitate generalized IDMTS in capuchin monkeys.
• The development of "select" relations is a key factor in achieving generalized IDMTS.
• Excluding familiar non-matching stimuli may enhance performance in Shuffled S- IDMTS tests.