How long does the Probabilistic Reversal Learning Task take?
About 4 to 6 minutes, as listed in the test library.
Executive Function
Probabilistic Reversal Learning Task
Measure reward learning and flexibility as the better of two symbols switches, with points shown after each choice.
Useful for reinforcement learning, set shifting, cognitive flexibility, and intervention studies.
Free to start: 10 free participants per account, then $0.99 per participant for academic research, or any yearly plan. Preview any test free after you sign up.
Executive function4-6 minUntimed
A
WinB
LoseContingency reversed
A now worseB now better
Flexible updating
Learn reward patterns as they reverse
What this task measures
Measures reward learning and flexibility when the better of two symbols switches without warning.
Core constructs
- Reward learning
- Cognitive flexibility
- Perseveration
- Feedback sensitivity
Research fit
- Reinforcement-learning research
- Cognitive flexibility studies
- Psychiatric and neurological research batteries
- Within-study group comparison
Paradigm overview
Probabilistic Reversal Learning measures how people learn from rewards and adjust when the rewarded choice changes. On each trial, participants choose one of two abstract symbols and then see whether they won 10 points.
The better symbol pays on 8 of every 10 trials and the other symbol on 2 of every 10, so some feedback is misleading. When 8 of the last 10 choices pick the better symbol, the two symbols switch roles without warning. If that does not happen within 40 trials, the roles switch anyway. Symbol sides change at random from trial to trial. The reward schedule comes from a seeded generator and is saved with the trial data.
Scores include trials to the first learning criterion, reversals completed, reversal errors, perseverative errors, and win-stay and lose-shift rates.
Key scoring outputs
Perseverative Errors
countAfter each reversal that followed a reached criterion, repeated choices of the previously better symbol before the first switch. The first trial after a reversal is not counted.
Lower is better
Trials to Criterion
trialsTrials until 8 of the last 10 choices first picked the better symbol.
Lower means faster learning
Reversals Completed
countTimes the participant reached the learning criterion again after a reversal.
Higher is better
Reversal Errors
countChoices of the worse symbol after the first reversal.
Lower is better
Win-Stay Rate
proportionShare of trials after a win on which the participant chose the same symbol again.
Informational
Lose-Shift Rate
proportionShare of trials after no win on which the participant switched symbols.
Informational
Task design as built
The trial sequence participants see, as the test runs in ConductCognition. Settings a researcher can change are listed below.
- Symbols
- Two abstract symbols whose sides change at random on every trial.
- Rewards
- The better symbol pays 10 points on 8 of every 10 trials, the other on 2 of every 10.
- Feedback
- +10 points or No points for 1 second, with the running total.
- Reversal
- When 8 of the last 10 choices pick the better symbol, or after 40 trials in a phase.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Number of trials | 90 | 60 to 150 |
Yearly plans can change settings within these limits. Free and pay-as-you-go studies use the standard settings.
Need a setting this test does not have?
Devices and languages
- Untimed test: runs on desktops, laptops, tablets and phones. Participant devices
- Task text: English and Italian.
- Available on: Every plan, including the free plan.
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite this test
ConductCognition versions: task version 2.0.0, scoring version 2.0.0. Every export records the versions each session used, for a methods section.
- Cools, R., Clark, L., Owen, A. M., & Robbins, T. W. (2002). Defining the neural mechanisms of probabilistic reversal learning using event-related functional magnetic resonance imaging. Journal of Neuroscience, 22(11), 4563-4567. https://doi.org/10.1523/JNEUROSCI.22-11-04563.2002
Source papers
Cools, Clark, Owen and Robbins (2002): probabilistic reversal learning with abstract patterns, randomized sides and reversal errors.
Paradigm sourceAdults
Research FAQ
Common questions about this online cognitive test
What device does a participant need?
Untimed test: runs on desktops, laptops, tablets and phones. The browser must be Chrome, Edge, Firefox or Safari.
What scores does it produce?
Perseverative Errors, Trials to Criterion, Reversals Completed, Reversal Errors, Win-Stay Rate and Lose-Shift Rate. Each session is also marked valid, needs review or invalid, with the reason.
Which settings can a researcher change?
Number of trials (60 to 150). Settings can be changed on yearly plans; free and pay-as-you-go studies use the standard settings.
Which languages can participants use?
The test's own text is in English and Italian. Italian participant pages come with yearly plans.
Is this a clinical test?
No. ConductCognition is for research use only. It runs, scores and exports the test; clinical interpretation is outside the platform.
Every test: the full test list.
Run this test in your next study
Start free with 10 free participants on any test, then use participant credits or a yearly plan.