Decision Making
Probability Discounting Task
Also called: probabilistic discounting, odds-against discounting, risky choice discounting, certainty equivalent task
Measure how much uncertainty lowers the value of a reward: adjusting choices between sure points and a stated chance of more, scored as the area under the discounting curve.
4-6 minUntimed
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.
- Papers, 2021 to 2025
- 109
- Evidence grade
- C
- Registered clinical studies
- 7
What it measures
Measures how much uncertainty lowers the value of a reward. Adjusting choices between some points for sure and a stated chance of 640 points, at 7 chances from 95% to 5%, give the certain amount worth as much as each chance; the area under those points is the main score.
Constructs
- Probability discounting
- Risk preference
- Decision-making under risk
- Valuation of uncertain rewards
Research uses
- Gambling and addiction research
- Behavioral economics and risk preference research
- Decision-making across the lifespan
- Within-study group comparison
How the paradigm works
The probability discounting task measures how much uncertainty lowers the value of a reward (Rachlin, Raineri and Cross, 1991). Each screen offers some points for sure or a stated chance of 640 points, otherwise nothing. At each chance the certain amount adjusts with the participant's choices, by the adjusting-amount procedure of the delay discounting task (Du, Green and Myerson, 2002): it starts at 320 points and moves down after the certain amount is chosen or up after the chance is chosen, by 160 points and then by half the step after each choice. The amount that would come after the last choice is the indifference point: the certain points worth as much as the chance.
More about the paradigm
Seven chances are asked, 95%, 90%, 75%, 50%, 25%, 10% and 5%, in an order drawn from the session's seed, with 6 choices at each (3 to 6 in the settings): 42 choices at the standard settings. The side of the certain amount is drawn on every choice. One check question with feedback comes first. At the end one choice is drawn and played: certain points are kept, and a chance is decided by a draw from the seed. The study decides what the points are worth, and the payment line says whether they are exchanged for money. There is no time limit. The points and the chances are this test's own.
The main score is the area under the curve: the indifference points as shares of 640, plotted against the odds against winning, (100 - p) / p, scaled by the largest odds against, with the point for a certain reward at 1 (Myerson, Green and Warusawitharana, 2001). It runs from 0 to 1, and a larger area means uncertain points lose less value. A hyperbola fitted to the points gives h (Rachlin, Raineri and Cross, 1991): h = 1 values each chance at its expected points, above 1 values chances below them, and below 1 above them. The indifference point at each chance, the share of certain choices and the Johnson and Bickel (2008) checks for nonsystematic data are also reported. Choices under 100 ms are counted as anticipations. Scores support research and do not diagnose.
Scores
- Area Under the Curve%
- Area under the indifference points as shares of 640 points, against the odds against winning scaled from 0 to 1, with the point (0, 1) for a certain reward.
- Higher means uncertain points lose less value
- Probability Discount Rate (ln h)score
- Natural logarithm of h in v = V / (1 + h x odds against), fitted by least squares with V held at 640 points.
- 0 matches the expected value; above 0 values chances below it
- Probability Discount Rate (h)score
- The fitted h itself, before the logarithm.
- 1 matches the expected value
- Fit (R squared)%
- Share of the variance in the indifference points, as shares of 640, that the fitted hyperbola explains.
- Higher means the hyperbola describes the points well
- Certain Choices%
- Share of all choices that took the certain points.
- Informational
- Consistency Flagflag
- 1 when the indifference points pass both Johnson and Bickel (2008) checks, from the largest chance to the smallest, with 640 points as the reference. 0 when a point rises more than 128 points above the one before it, or the last is not at least 64 points below the first.
- 1 supports interpretation
- Rise Checkflag
- 1 when an indifference point rises more than 128 points above the one for the next larger chance (Johnson and Bickel criterion 1).
- 0 supports interpretation
- Spread Checkflag
- 1 when the indifference point at 5% is not at least 64 points below the one at 95% (Johnson and Bickel criterion 2).
- 0 supports interpretation
- Certain Equivalent at 95%points
- Certain points worth as much as a 95% chance of 640 points.
- Informational
- Certain Equivalent at 90%points
- Certain points worth as much as a 90% chance of 640 points.
- Informational
- Certain Equivalent at 75%points
- Certain points worth as much as a 75% chance of 640 points.
- Informational
- Certain Equivalent at 50%points
- Certain points worth as much as a 50% chance of 640 points.
- Informational
- Certain Equivalent at 25%points
- Certain points worth as much as a 25% chance of 640 points.
- Informational
- Certain Equivalent at 10%points
- Certain points worth as much as a 10% chance of 640 points.
- Informational
- Certain Equivalent at 5%points
- Certain points worth as much as a 5% chance of 640 points.
- Informational
- Check Correctflag
- 1 when the check question about what a chance gives was answered correctly, 0 when not.
- 1 supports interpretation
- Choice Timems
- Median time to make a choice, over choices of at least 100 ms.
- Informational
- Anticipationscount
- Choices under 100 ms. They stay in the curve, because each choice sets the next offer.
- Lower is better
- Chances Completedcount
- Chances with an indifference point, out of 7.
- 7 supports interpretation
- Choices Madetrials
- Adjusting choices made: 42 at the standard settings.
- Informational
- Points Wonpoints
- Points from the drawn choice: its certain points, or 640 or 0 from its chance.
- Informational
Use and evidence
109 papers from 2021 to 2025 name this task in their title or abstract, down 26% compared with 2016 to 2020. That ranks it 179 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 19 |
| 2024 | 15 |
| 2023 | 22 |
| 2022 | 22 |
| 2021 | 31 |
| 2020 | 26 |
| 2019 | 31 |
| 2018 | 38 |
| 2017 | 25 |
| 2016 | 27 |
| 2015 | 26 |
| 2014 | 33 |
| 2013 | 28 |
| 2012 | 21 |
| 2011 | 9 |
| 2010 | 12 |
| 2009 | 20 |
| 2008 | 14 |
| 2007 | 12 |
| 2006 | 10 |
| 2005 | 4 |
| 2004 | 1 |
| 2003 | 5 |
| 2002 | 2 |
| 2001 | 1 |
| 2000 | 1 |
Use grade C
100 to 299 papers from 2021 to 2025.
Evidence grade C
At least 2 test-retest papers, 30 reliability or validity papers or 3 registered clinical studies.
- Papers on test-retest reliability
- 3
- Papers on reliability or validity
- 24
- Registered clinical studies using it as an outcome
- 7
How the grades work
Use: A is 1,000 or more papers from 2021 to 2025, B 300 to 999, C 100 to 299, D 30 to 99 and E fewer than 30. Evidence: A means at least 25 papers report its test-retest reliability and at least 25 registered clinical studies use it as an outcome; B, at least 8 test-retest papers, and either 5 registered clinical studies or 150 reliability or validity papers; C, at least 2 test-retest papers, 30 reliability or validity papers or 3 registered clinical studies; D, little published measurement evidence found. The counts measure use. They include every paper that names the task in its title or abstract. Grades describe the published research on this task, not ConductCognition's version of it.
Source: ConductCognition test census 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Choices
- Some points for sure or a stated chance of 640 points, otherwise nothing, as two buttons whose sides are drawn from the session's seed. No time limit.
- Chances
- 95%, 90%, 75%, 50%, 25%, 10% and 5%, in an order drawn from the session's seed, with the choices at one chance together.
- Adjusting
- At each chance the certain amount starts at 320 points and moves down after it is chosen or up after the chance is chosen, by 160, then 80, 40, 20, 10 and 5 points. 6 choices at each chance (3 to 6 in the settings), 42 in all.
- Practice
- One check question with feedback: a chance gives its points or nothing.
- Draw
- One choice drawn from the session's seed is played: certain points are kept, and a chance is decided by a draw.
- Points
- The study decides what points are worth, and the payment line says whether they are exchanged for money.
- Scores
- The area under the indifference points as shares of 640, against the odds against winning scaled from 0 to 1 (Myerson and colleagues, 2001). Also h of a fitted hyperbola, the indifference point at each chance and the share of certain choices. Choices under 100 ms are counted as anticipations.
- Checks
- Choosing one screen side on 90% of choices invalidates. The Johnson and Bickel (2008) checks for nonsystematic data, a chance without an indifference point and a wrong check answer are flagged and never exclude.
Wan and colleagues (2026) used hypothetical dollar amounts and five probabilities from 5% to 95%. This test uses one amount of its own points and seven chances, and plays one drawn choice. The area weighs the small chances most, because the odds against grow fastest there, and larger uncertain amounts are discounted more steeply, so compare results within a study. The scores support research and do not diagnose.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: the scored run's choices. Choices have no correct answer. Choosing one screen side on 90% of choices invalidates, because the side of the certain amount is drawn at random each time. Answers under 200 ms on at least half the choices invalidate. A run without an indifference point at every chance, failing the Johnson and Bickel (2008) checks and a wrong check answer are flagged.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. A preference for certain points or for a chance moves between screen sides, so choosing by side alone means the options were not read. Johnson and Bickel (2008) present their two checks for nonsystematic data as a tool for identification, not automatic exclusion, and someone who values every chance at its full amount fails the second one genuinely. A participant who misreads what a chance gives may choose the opposite of what they intend, so a wrong check answer is flagged for review and never invalidates. An answer under 200 ms is faster than a simple visual reaction, so it cannot reflect the item. Following response-time effort screening for untimed tests (Wise and Kong 2005), the session is invalid only when at least half the answers are that fast.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Choices per chance | 6 | 3 to 6 |
| Payment line | Hypothetical points (standard) | Hypothetical points (standard), The study pays for points |
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
- Languages: English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese.
- Available on: Every plan, including the free plan.
- No percentile. Scores and quality flags only. How norms work
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Probability Discounting Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Rachlin H, Raineri A, Cross D, 1991), with the standard settings (choices per chance: 6; payment line: Hypothetical points). The main score was area under the curve (area under the indifference points as shares of 640 points, against the odds against winning scaled from 0 to 1, with the point (0, 1) for a certain reward).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Rachlin, H., Raineri, A., & Cross, D. (1991). Subjective probability and delay. Journal of the Experimental Analysis of Behavior, 55(2), 233-244. https://doi.org/10.1901/jeab.1991.55-233
- Du, W., Green, L., & Myerson, J. (2002). Cross-cultural comparisons of discounting delayed and probabilistic rewards. The Psychological Record, 52(4), 479-492. https://doi.org/10.1007/BF03395199The adjusting-amount procedure.
- Myerson, J., Green, L., & Warusawitharana, M. (2001). Area under the curve as a measure of discounting. Journal of the Experimental Analysis of Behavior, 76(2), 235-243. https://doi.org/10.1901/jeab.2001.76-235The main score.
- Wan, H., Myerson, J., Green, L., Strube, M. J., & Hale, S. (2026). Age, income, and the discounting of delayed and probabilistic rewards. Frontiers in Psychology, 17, 1765142. https://doi.org/10.3389/fpsyg.2026.1765142The procedure as run with 592 adults aged 20 to 80.
- Johnson, M. W., & Bickel, W. K. (2008). An algorithm for identifying nonsystematic delay-discounting data. Experimental and Clinical Psychopharmacology, 16(3), 264-274. https://doi.org/10.1037/1064-1297.16.3.264The checks for nonsystematic data.
Source papers
- Wan, Myerson, Green, Strube and Hale (2026): 592 adults aged 20 to 80 made 6 adjusting choices at each of five probabilities from 5% to 95%, for $150, $2,500 and $30,000; among those aged 35 to 80 the mean area under the curve was 0.15 (SD 0.16), and larger amounts were discounted more steeply.Paradigm source, Adults
Questions
How long does the Probability Discounting Task take?
About 4 to 6 minutes, as listed in the test library.
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?
Area Under the Curve, Probability Discount Rate (ln h), Probability Discount Rate (h), Fit (R squared), Certain Choices, Consistency Flag, Rise Check, Spread Check, Certain Equivalent at 95%, Certain Equivalent at 90%, Certain Equivalent at 75%, Certain Equivalent at 50%, Certain Equivalent at 25%, Certain Equivalent at 10%, Certain Equivalent at 5%, Check Correct, Choice Time, Anticipations, Chances Completed, Choices Made and Points Won. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Probability Discounting Task used in research?
109 papers from 2021 to 2025 name it in their title or abstract, down 26% compared with 2016 to 2020. That ranks it 179 of 395 test families in the ConductCognition test census (use grade C). Source: ConductCognition test census 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Probability Discounting Task have?
In the published research, 3 papers report its test-retest reliability and 7 registered clinical studies use it as an outcome (evidence grade C in the ConductCognition test census).
Does it report a percentile?
No. Scores and quality flags only.
Which settings can a researcher change?
Choices per chance (3 to 6) and payment line (Hypothetical points (standard), The study pays for points). Settings can be changed on yearly plans; free and pay-as-you-go studies use the standard settings.
Which languages can participants use?
Participants can take this test in English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese. English is on every plan, and the other languages come with yearly plans. Have a native speaker read the machine-translated text before launch.
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.