Sequence as context in category learning: An eyetracking study.

Area of Science:

• Cognitive Psychology
• Learning and Memory

Background:

• Categorization tasks are fundamental to learning.
• Understanding how learning context influences feature attention is crucial for cognitive models.

Purpose of the Study:

• To investigate if the proximity of categories in a learning sequence affects feature attention during categorization.
• To determine if the local learning environment influences how categories are represented.

Main Methods:

• Experiment 1: Participants learned 4 categories distinguished by 1 feature, with categories presented in two pairs sequentially.
• Experiment 2: Participants learned 4 categories with interleaved category presentations.
• A transfer test assessed attention to diagnostic features across all categories.

Main Results:

• Participants preferentially attended to features that distinguished categories presented adjacently in the learning sequence.
• This effect persisted even when categories were interleaved during learning.
• Equally diagnostic features were overlooked if they did not separate proximal categories.

Conclusions:

• The local learning context significantly influences attentional biases in categorization.
• Category representations are dynamically shaped by the immediate sequence of learning experiences.
• Findings suggest a context-dependent mechanism for feature selection in category learning.