Visual prototype formation with discontinuous representation of dimensions of variability

Area of Science:

• Cognitive Psychology
• Perception Science

Background:

• Prototype abstraction is a key concept in understanding how humans generalize from experience.
• Two prominent models, prototype-distance and attribute-frequency, offer different predictions for prototype formation.

Purpose of the Study:

• To investigate which model better predicts prototype formation under varying conditions.
• To determine if prototypes are always centered on experienced stimuli or can include unexperienced values.

Main Methods:

• Three experiments were conducted to test predictions of the prototype-distance and attribute-frequency models.
• Stimuli varied across multiple dimensions, and recognition accuracy was measured.

Main Results:

• Prototype formation varied: sometimes unexperienced values were included, other times most frequent values were best recognized.
• The prototype-distance model failed to explain recognition of stimuli deviating from the central tendency.
• The attribute-frequency model, with added assumptions, could account for both observed outcomes.

Conclusions:

• Prototype formation is more complex than initially proposed by the prototype-distance model.
• The attribute-frequency model provides a more flexible framework for understanding prototype abstraction, accommodating both central tendency and frequency-based effects.