Transitive inference after minimal training in rhesus macaques (Macaca mulatta).

Area of Science:

• Cognitive Science
• Animal Behavior
• Neuroscience

Background:

• Rhesus macaques demonstrate transitive inference (TI) abilities on ordered lists.
• The learning process underlying TI acquisition in primates remains unclear.
• Understanding how training amount influences TI performance is crucial.

Purpose of the Study:

• To investigate the relationship between the quantity of transitive inference training and response accuracy in rhesus macaques.
• To model subtle learning effects in the presence of significant response variability and bias.

Main Methods:

• Four rhesus macaques were trained on seven-item ordered lists.
• Training conditions varied from minimal (one presentation per adjacent pair) to extensive (24 and 114 trials).
• A descriptive statistical model was employed to analyze learning progression.

Main Results:

• Subjects demonstrated incremental learning of serial order.
• Performance accuracy increased with the amount of training.
• Learning appeared statistical, refining uncertainty with accumulated evidence.

Conclusions:

• Serial order learning in rhesus macaques is an incremental, statistical inference process.
• This contrasts with a purely logical or deductive approach to transitive inference.
• Learning progresses from initial uncertainty to increased accuracy as training data accumulates.