Decision Making
Columbia Card Task
Also called: CCT, hot and cold Columbia Card Task, card task of risk taking
Measure risk taking with 32 face-down cards: participants choose how many to turn over, or turn them one at a time, as the gain, the loss and the number of loss cards change.
9-12 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
- 27
- Evidence grade
- C
- Registered clinical studies
- 4
What it measures
Measures risk taking with 32 face-down cards, a few of them loss cards. In the cold part participants choose how many cards to turn over before seeing any; in the hot part they turn cards over one at a time and see each outcome at once. The main score is the mean number of cards chosen in the cold part, and the effects of the gain, the loss and the number of loss cards show how the values are used.
Constructs
- Risk taking
- Use of gain, loss and probability information
- Affective and deliberative decision making
Research uses
- Decision-making and risk research
- Adolescent and developmental research
- Impulsivity and substance use studies
- Within-study group comparison
How the paradigm works
The Columbia Card Task measures risk taking with 32 cards face down, a few of them loss cards (Figner, Mackinlay, Wilkening and Weber, 2009). Before each round the participant sees how many cards are loss cards, how many points a gain card adds and how many points a loss card takes away. Gain cards add their points; a loss card takes its points away and ends the round. In the cold part the participant chooses how many cards to turn over, from 0 to 32, before any is shown, and sees the points only after the last cold round. In the hot part the participant turns cards over one at a time and sees each outcome at once, until they stop or a loss card ends the round. In every round the loss cards are placed at random, as the instructions say.
More about the paradigm
Each part plays every combination of 1 or 3 loss cards, a gain of 20 or 40 points and a loss of 150 or 350 points three times, 24 rounds per part, after one practice round with feedback, and the order of the parts is drawn for each session. These payoffs were written for this test so that the number of cards with the highest expected points differs in each combination, from 0 to 28. A study can set 2 to 4 rounds of each combination and fix the order of the parts. The points have no set value, and money is mentioned only if the study turns on the payment line.
The main score is the mean number of cards chosen in the cold part. The effects of the gain, the loss and the number of loss cards on the cards chosen, the gap between the cards chosen and the number with the highest expected points, the mean cards turned over in the hot part and the share of hot rounds a loss card ended are also reported. In the hot part a loss card can end a round before the participant would have stopped, which lowers the count; the source placed the loss cards last on most rounds without saying so, and this test does not. Scores support research and do not diagnose.
Scores
- Cold Mean Cardscount
- Mean number of cards chosen in the cold part, where every choice is made before any card is shown.
- Higher means more risk taken
- Cold Gain Effectcount
- Mean cards chosen at a gain of 40 minus at a gain of 20, in the cold part.
- Positive values mean more cards when gain cards pay more
- Cold Loss Effectcount
- Mean cards chosen at a loss of 350 minus at a loss of 150, in the cold part.
- Negative values mean fewer cards when a loss card costs more
- Cold Loss Card Effectcount
- Mean cards chosen with 3 loss cards minus with 1, in the cold part.
- Negative values mean fewer cards when a loss card is more likely
- Cold Gap from Best Numbercount
- Mean of the cards chosen minus the number with the highest expected points, in the cold part.
- Positive values mean more cards than that number
- Hot Mean Cardscount
- Mean number of cards turned over in the hot part, counting a loss card that ended the round.
- Higher means more risk taken; a loss card found by chance lowers it
- Hot Mean Cards When Stoppedcount
- Mean cards turned over in the hot rounds the participant ended by stopping.
- Informational
- Hot Rounds Ended by a Loss Card%
- Share of hot rounds that a loss card ended.
- Informational
- Cold Decision Timems
- Median time from the start of a cold round to confirming the number of cards.
- Informational
- Cold Pointspoints
- Points over the cold rounds.
- Informational
- Hot Pointspoints
- Points over the hot rounds.
- Informational
- Total Pointspoints
- Points over all scored rounds.
- Informational
- Cold Same-Choice Share%
- Share of cold rounds with the most common number of cards.
- 90% or more is flagged for review
- Cold Rounds at 0 Cardscount
- Cold rounds in which 0 cards were chosen.
- 90% or more of the cold rounds is invalid
- Hot Rounds with No Cardcount
- Hot rounds stopped before any card was turned over.
- 90% or more of the hot rounds is invalid
- Cold Roundstrials
- Cold rounds scored.
- Informational
- Hot Roundstrials
- Hot rounds scored.
- Informational
- Planned Roundstrials
- Rounds the session planned to play across both parts.
- Under 90% played is invalid
Use and evidence
27 papers from 2021 to 2025 name this task in their title or abstract, down 23% compared with 2016 to 2020. That ranks it 310 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 7 |
| 2024 | 8 |
| 2023 | 5 |
| 2022 | 6 |
| 2021 | 1 |
| 2020 | 7 |
| 2019 | 9 |
| 2018 | 8 |
| 2017 | 6 |
| 2016 | 5 |
| 2015 | 14 |
| 2014 | 1 |
| 2013 | 5 |
| 2012 | 3 |
| 2011 | 1 |
| 2010 | 1 |
| 2009 | 3 |
| 2008 | 0 |
| 2007 | 0 |
| 2006 | 0 |
| 2005 | 0 |
| 2004 | 0 |
| 2003 | 0 |
| 2002 | 0 |
| 2001 | 0 |
| 2000 | 0 |
Use grade E
Fewer than 30 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
- 1
- Papers on reliability or validity
- 8
- Registered clinical studies using it as an outcome
- 4
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.
- Cards
- 32 cards face down in 4 rows of 8, drawn by our code: a patterned back, a gain face with an upward mark and a loss face with a downward mark, each with its points written on it.
- Values
- Before each round: 1 or 3 loss cards, a gain card worth 20 or 40 points and a loss card costing 150 or 350 points, in all 8 combinations. The payoffs are our own.
- Cold part
- Choose how many cards to turn over, from 0 to 32; the computer turns them over in a random order, out of sight. Points show only after the last cold round.
- Hot part
- Turn cards over one at a time, each outcome shown at once, until stopping or until a loss card ends the round. The result shows for 1,500 ms before the next round.
- Rounds
- Every combination 3 times in each part (standard, 2 to 4), 24 rounds per part, in cycles that never repeat a combination twice in a row. The order of the parts is drawn for each session (standard) or fixed.
- Loss cards
- Placed at random in every round, as the instructions say.
- Practice
- One round before each part, with 2 loss cards, a gain of 30 and a loss of 200, and its result shown.
- Points
- Each round starts at 0. Money is mentioned only if the study turns on the payment line.
Figner and colleagues placed the loss cards last on most hot rounds without telling participants, so every stop was a choice. This test places them at random, as it says, so a hot round can end on a loss card before the participant would have stopped, and the cold part's count is the main score. The payoffs differ from the source's so that the number of cards with the highest expected points differs in every combination. 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: columbia_card_round from the scored run, one per round of either part. Taking risk has no correct amount, so accuracy is not judged against chance. Fewer than 90% of the planned rounds, 0 cards in at least 90% of the cold rounds or no card turned in at least 90% of the hot rounds invalidates; one number of cards in at least 90% of the cold rounds is flagged for review.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. The measures average the cards chosen and compare them across the round values, so a part played at 0 cards throughout was not played and gives no measure, as in the balloon task's rule, and one number chosen whatever the values means the values were not used. The limits are ConductScience choices: the source sets none.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Rounds of each combination | 3 | 2 to 4 |
| Order of the two parts | Drawn for each session (standard) | Drawn for each session (standard), Cold part first, Hot part first |
| Payment line | No mention of money (standard) | No mention of money (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 and tablets. 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 Columbia Card Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Figner, Mackinlay, Wilkening and Weber, 2009), with the standard settings (rounds of each combination: 3; order of the two parts: Drawn for each session; payment line: No mention of money). The main score was cold mean cards (mean number of cards chosen in the cold part, where every choice is made before any card is shown).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Figner, B., Mackinlay, R. J., Wilkening, F., & Weber, E. U. (2009). Affective and deliberative processes in risky choice: Age differences in risk taking in the Columbia Card Task. Journal of Experimental Psychology: Learning, Memory, and Cognition, 35(3), 709-730. https://doi.org/10.1037/a0014983
- Kapadia, K., Tang, Y., & John, R. (2026). Jokers in the deck: A new temperature setting for the Columbia Card Task. Behavior Research Methods, 58(2), 45. https://doi.org/10.3758/s13428-025-02932-8Loss cards placed at random, and the gap from the number with the highest expected points.
- Markiewicz, Ł., & Kubińska, E. (2015). Information use differences in hot and cold risk processing: When does information about probability count in the Columbia Card Task? Frontiers in Psychology, 6, 1727. https://doi.org/10.3389/fpsyg.2015.01727How the gain, the loss and the number of loss cards are used in each part.
- de Groot, K., & van Strien, J. W. (2019). Event-related potentials in response to feedback following risk-taking in the hot version of the Columbia Card Task. Psychophysiology, 56(9), e13390. https://doi.org/10.1111/psyp.13390A hot part with the cards shuffled at random, and the rounds a loss card ends.
Source papers
- Figner B, Mackinlay RJ, Wilkening F, Weber EU (2009). Affective and deliberative processes in risky choice: age differences in risk taking in the Columbia Card Task. Journal of Experimental Psychology: Learning, Memory, and Cognition, 35(3):709-730.Paradigm source, Adolescents and adults
Questions
How long does the Columbia Card Task take?
About 9 to 12 minutes, as listed in the test library.
What device does a participant need?
Untimed test: runs on desktops, laptops and tablets. The browser must be Chrome, Edge, Firefox or Safari.
What scores does it produce?
Cold Mean Cards, Cold Gain Effect, Cold Loss Effect, Cold Loss Card Effect, Cold Gap from Best Number, Hot Mean Cards, Hot Mean Cards When Stopped, Hot Rounds Ended by a Loss Card, Cold Decision Time, Cold Points, Hot Points, Total Points, Cold Same-Choice Share, Cold Rounds at 0 Cards, Hot Rounds with No Card, Cold Rounds, Hot Rounds and Planned Rounds. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Columbia Card Task used in research?
27 papers from 2021 to 2025 name it in their title or abstract, down 23% compared with 2016 to 2020. That ranks it 310 of 395 test families in the ConductCognition test census (use grade E). Source: ConductCognition test census 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Columbia Card Task have?
In the published research, 1 paper reports its test-retest reliability and 4 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?
Rounds of each combination (2 to 4), order of the two parts (Drawn for each session (standard), Cold part first, Hot part first) and payment line (No mention of money (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.
Related tests