A dimensional summation account of polymorphous category learning.

Area of Science:

• Cognitive Psychology
• Learning Sciences

Background:

• Polymorphous concepts, despite having overall similarity, are difficult to learn.
• The dimensional summation hypothesis predicts challenges in forming polymorphous concepts.

Purpose of the Study:

• To test predictions of the dimensional summation hypothesis regarding polymorphous concept formation.
• To compare the dimensional summation hypothesis with alternative accounts of concept learning.

Main Methods:

• Experiments were conducted to test specific predictions of the dimensional summation hypothesis.
• The study compared the dimensional summation hypothesis against strategic, seriality, stimulus decomposition, and error-based accounts.
• Deterministic pretraining and concurrent counting tasks were employed.

Main Results:

• Deterministic pretraining on stimulus features facilitated polymorphous concept formation compared to direct concept training.
• The dimensional summation hypothesis provided the best explanation for the experimental data.
• The single feature pretraining effect was eliminated by a concurrent counting task.

Conclusions:

• The dimensional summation hypothesis accurately predicts aspects of polymorphous concept formation.
• Natural concept acquisition may involve the serial summation of evidence, with implications for educational strategies.