Tests of an exemplar model for relating perceptual classification and recognition memory

Area of Science:

• Cognitive Psychology
• Computational Neuroscience
• Psychophysics

Background:

• Understanding how humans process visual information, particularly faces, is crucial in cognitive science.
• Previous models have attempted to explain recognition and classification based on feature analysis and exemplar retrieval.

Purpose of the Study:

• To investigate how individuals make classification, recognition, and similarity judgments for schematic faces.
• To test an exemplar-similarity model's ability to predict these judgments.
• To examine the differential allocation of attention to psychological dimensions during these tasks.

Main Methods:

• Conducted experiments involving classification, recognition, and similarity judgments on 34 schematic faces.
• Utilized multidimensional scaling (MDS) to derive a spatial solution from similarity judgments.
• Employed an exemplar-similarity model, integrating MDS solutions, to predict classification and recognition performance.

Main Results:

• The exemplar-similarity model accurately predicted participants' classification and recognition judgments.
• Evidence indicated differential attention allocation to psychological dimensions for classification versus recognition.
• Attention distribution approached ideal-observer levels for classification and showed tendencies for recognition.
• Interactive effects of exemplar frequencies and similarities were observed, aligning with the exemplar model's predictions.
• Unexpectedly, frequency effects were larger for classification than for recognition.

Conclusions:

• The exemplar-similarity model provides a robust framework for understanding face judgment tasks.
• Attention allocation strategies adapt based on the specific cognitive task (classification vs. recognition).
• Frequency effects in face perception warrant further investigation, particularly their differential impact on classification and recognition.