Social Cognition
Moral Dilemma Task
Also called: trolley problem, trolley dilemma, footbridge dilemma, sacrificial dilemmas, moral judgment task
Measure judgments about causing harm for a greater good: participants judge sacrificial dilemmas in incongruent and congruent versions, scored by harm accepted and the process dissociation parameters.
8-10 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
- 508
- Evidence grade
- C
- Registered clinical studies
- 1
What it measures
Measures judgments about causing harm for a greater good. Participants read 10 sacrificial dilemmas, each in a version where the harmful action saves five people and a version where it brings only a minor benefit, and judge whether the action is appropriate. The scenarios are in English.
Constructs
- Moral judgment
- Harm aversion
- Utilitarian and deontological inclinations
- Social decision making
Research uses
- Moral psychology research
- Social cognition studies
- Personality and individual difference research
- Within-study group comparison
How the paradigm works
The Moral Dilemma Task measures judgments about causing harm for a greater good. Participants read short sacrificial dilemmas, in which an action would cost one person's life, and judge whether the action is appropriate. This version has 10 scenarios written for the task, each in two versions with the same harmful action: an incongruent version, where the action saves five people, and a congruent version, where it brings only a minor benefit, such as saving property. By the criteria of Greene and colleagues (2001), five scenarios are personal, where the participant would harm someone directly, and five are impersonal, where the participant would redirect an existing threat onto one person.
More about the paradigm
Each scenario appears on its own screen, read at the participant's own pace, and the question follows on the next screen: "Is it appropriate for you to [action] in order to [goal]?", answered Appropriate or Inappropriate, as in Greene and colleagues (2001) and Conway and Gawronski (2013). After one unscored practice item, the 20 items come in an order drawn from the session's seed, with the two versions of each scenario in different halves of the session. A setting asks "Would you [action] in order to [goal]?" instead, answered Yes or No. The scenarios are in English; the instructions are also in Italian.
Scores
- Harm Accepted on Incongruent Dilemmas%
- Share of incongruent dilemmas, where the harmful action saves five people, in which the action was accepted.
- Higher means more acceptance of harm for the greater good
- Utilitarian Parameter (U)score
- Process dissociation estimate: the rejection rate on congruent dilemmas minus the rejection rate on incongruent dilemmas (Conway and Gawronski, 2013). Ranges from -1 to 1.
- Higher means outcomes weigh more in the judgments
- Deontological Parameter (D)score
- Process dissociation estimate: the rejection rate on incongruent dilemmas divided by 1 minus U. Ranges from 0 to 1; not computed when U is 1.
- Higher means stronger rejection of causing harm whatever the outcome
- Harm Accepted on Personal Dilemmas%
- Share of incongruent personal dilemmas, where the harm is caused directly, in which the action was accepted.
- Informational
- Harm Accepted on Impersonal Dilemmas%
- Share of incongruent impersonal dilemmas, where an existing threat is redirected, in which the action was accepted.
- Informational
- Harm Accepted on Congruent Dilemmas%
- Share of congruent dilemmas, where the harmful action brings only a minor benefit, in which the action was accepted.
- Low values are expected; 0.5 or more is flagged for review
Use and evidence
508 papers from 2021 to 2025 name this task in their title or abstract, up 50% compared with 2016 to 2020. That ranks it 50 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 120 |
| 2024 | 92 |
| 2023 | 114 |
| 2022 | 89 |
| 2021 | 93 |
| 2020 | 97 |
| 2019 | 67 |
| 2018 | 65 |
| 2017 | 54 |
| 2016 | 56 |
| 2015 | 55 |
| 2014 | 39 |
| 2013 | 35 |
| 2012 | 24 |
| 2011 | 28 |
| 2010 | 15 |
| 2009 | 7 |
| 2008 | 6 |
| 2007 | 6 |
| 2006 | 1 |
| 2005 | 5 |
| 2004 | 1 |
| 2003 | 0 |
| 2002 | 1 |
| 2001 | 3 |
| 2000 | 4 |
Use grade B
300 to 999 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
- 46
- Registered clinical studies using it as an outcome
- 1
Note on the counts: includes philosophy and AI papers on the trolley problem.
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.
- Scenarios
- 10 sacrificial dilemmas written for this task, each with one person who would not survive the action, in plain and non-graphic language.
- Versions
- Each scenario in an incongruent version, where the action saves five people, and a congruent version, where the same action brings only a minor benefit: 20 items.
- Personal and impersonal
- 5 personal scenarios, where the participant would harm someone directly, and 5 impersonal, where an existing threat would be redirected onto one person, by the criteria of Greene and colleagues (2001).
- Screens
- The scenario on one screen, read at the participant's own pace, then the question on the next with two answer buttons. No time limit.
- Question
- "Is it appropriate for you to [action] in order to [goal]?", answered Appropriate or Inappropriate (standard), or "Would you [action] in order to [goal]?", answered Yes or No.
- Order
- Drawn from the session's seed: two halves of 10, each with one version of every scenario and 5 of each version. Which answer button is on the left is also drawn and saved.
- Practice
- One non-moral practice item, not scored.
- Scores
- Harm accepted on incongruent dilemmas, the process dissociation parameters U and D of Conway and Gawronski (2013), and harm accepted on personal, impersonal and congruent dilemmas.
The scenarios were written for this task; no published dilemma set is used. The scenarios are English; sessions run in other languages are flagged for review. Conway and Gawronski (2013) showed their 20 items in one fixed random order; here the order is drawn for each session.
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 judgments. No answer is correct, so accuracy is not judged against chance, and the same answer on every judgment is accepted. Accepting the harmful action on at least half of the congruent dilemmas is flagged for review, as is a session run in another language. Judgments under 200 ms on at least half the items invalidate.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. Rejecting every harmful action is a stable moral view, so repeated answers are not checked. In a congruent dilemma the harmful action brings only a minor benefit, so neither a deontological nor a utilitarian inclination favors it (Conway and Gawronski 2013); accepting it on at least half of these items suggests careless answering, so it is flagged for review and never invalidates. The scenarios are English, so a session run in another language is flagged for review. 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 |
|---|---|---|
| Question wording | Appropriate or inappropriate (standard) | Appropriate or inappropriate (standard), Would you: yes or no |
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
- Instructions and screens: English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese. Words and sentences: English.
- 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 Moral Dilemma Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Greene et al., 2001; Conway & Gawronski, 2013), with the standard settings (question wording: Appropriate or inappropriate). The main score was harm accepted on incongruent dilemmas (share of incongruent dilemmas, where the harmful action saves five people, in which the action was accepted).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Greene, J. D., Sommerville, R. B., Nystrom, L. E., Darley, J. M., & Cohen, J. D. (2001). An fMRI investigation of emotional engagement in moral judgment. Science, 293(5537), 2105-2108. https://doi.org/10.1126/science.1062872
- Conway, P., & Gawronski, B. (2013). Deontological and utilitarian inclinations in moral decision making: A process dissociation approach. Journal of Personality and Social Psychology, 104(2), 216-235. https://doi.org/10.1037/a0031021The incongruent and congruent versions and the U and D parameters.
- Thomson, J. J. (1985). The trolley problem. Yale Law Journal, 94(6), 1395-1415. https://doi.org/10.2307/796133The trolley and footbridge cases behind sacrificial dilemmas.
Source papers
- Greene and colleagues (2001): personal and impersonal moral dilemmas judged appropriate or inappropriate. Conway and Gawronski (2013): process dissociation of incongruent and congruent dilemmas.Paradigm source, Adults
Questions
How long does the Moral Dilemma Task take?
About 8 to 10 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?
Harm Accepted on Incongruent Dilemmas, Utilitarian Parameter (U), Deontological Parameter (D), Harm Accepted on Personal Dilemmas, Harm Accepted on Impersonal Dilemmas and Harm Accepted on Congruent Dilemmas. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Moral Dilemma Task used in research?
508 papers from 2021 to 2025 name it in their title or abstract, up 50% compared with 2016 to 2020. That ranks it 50 of 395 test families in the ConductCognition test census (use grade B). Source: ConductCognition test census 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Moral Dilemma Task have?
In the published research, 1 paper reports its test-retest reliability and 1 registered clinical study uses 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?
Question wording (Appropriate or inappropriate (standard), Would you: yes or no). Settings can be changed on yearly plans; free and pay-as-you-go studies use the standard settings.
Which languages can participants use?
The instructions and screens are in English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese. The words and sentences stay in English, so sessions in other languages are flagged for review. 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