Related Experiment Video
Updated: Jul 9, 2025

06:11
A Conflict Model of Reward-seeking Behavior in Male Rats
Published on: February 20, 2019
7.5K
Predictions about reward outcomes in rhesus monkeys
Yiyun Huang1, Hayoung Chang2, Laurie R Santos3
1Department of Psychology, University of Michigan.
Behavioral Neuroscience
|December 7, 2023
Summary
Monkeys exhibit reward prediction errors by looking longer at unexpected rewards. Adolescent monkeys show heightened sensitivity to unexpected rewards, mirroring human developmental patterns.
Area of Science:
- Cognitive Science
- Comparative Psychology
- Neuroscience
Background:
- Infants and animals look longer at surprising events, indicating prediction and detection of prediction failures.
- Reward prediction error (RPE) quantifies the difference between expected and actual reward outcomes.
- Existing RPE research often relies on neurobiological methods not feasible in all populations.
Purpose of the Study:
- To develop a novel behavioral task assessing monkeys' reward predictions using looking time.
- To investigate how monkeys respond to positive and negative reward prediction errors.
- To examine lifespan changes in reward prediction sensitivity in monkeys.
Main Methods:
- A novel behavioral task using looking time responses was developed for monkeys.
- Study 1: 210 semi-free-ranging monkeys were tested on positive error, negative error, and number control conditions.
- Study 2: 363 monkeys across the lifespan were assessed for responses to positive reward prediction errors.
Main Results:
- Monkeys looked longer at unexpectedly large or small rewards compared to expected rewards.
- Adolescent monkeys exhibited heightened responses to unexpected rewards.
- No significant changes in reward prediction sensitivity were observed during aging in monkeys.
Conclusions:
- Monkeys' looking responses effectively track their reward predictions.
- Monkeys share developmental patterns of reward sensitivity with humans.
- This novel approach provides insights into cognitive processes underlying reward-based decision-making.

