Social Cognition
Who Is Responsible Task
Also called: responsibility judgments, responsibility attribution in groups, blame in teams, sharing out responsibility, pivotality and criticality, structural model of responsibility, diffusion of responsibility
Measure how people share out responsibility: participants read short stories in which three people each had a part in a plan that went wrong, with one or several failing, and rate how responsible each person is.
6-8 minUntimed
Free to start: 10 free participants per account, then $0.99 per participant for academic research, or any yearly plan. Try it here first, with no account.
- Papers, 2021 to 2025
- 48
- Evidence grade
- D
- Registered clinical studies
- 0
Ben brought the music.
Cleo brought the decorations.
What it measures
Measures how people share out responsibility among several people. Participants read 12 short stories in which 3 people each had a part in a plan that went wrong, such as a surprise party with no cake, where every part was needed or any one part was enough, and rate how responsible each person is. The main score is how much more responsible a lone failure is judged than one of three failures.
Constructs
- Responsibility attribution
- Pivotality
- Criticality
- Counterfactual reasoning about groups
Research uses
- Moral psychology research
- Social cognition and group processes research
- Causal reasoning research
- Within-study group comparison
How the paradigm works
The who is responsible task measures how people share out responsibility when several people contribute to one outcome. Gerstenberg and Lagnado (2010) had 69 people play a team game with three computer players and, after each round, rate each player's responsibility for the team's result from 0 (not responsible at all) to 10 (very responsible). Ratings followed a structural model of responsibility: a player who made the difference is fully responsible, and one who did not gets less the more of the other players' results would have had to change for them to make the difference; this model fit 52 of the 69 people best. Zultan, Gerstenberg and Lagnado (2012) showed that the way the parts combine matters, not only how many people failed: when a team of four lost, a member's blame on a 7-point scale fell from 5.55 to 4.46 when a teammate who could stand in for them had succeeded, but not when a teammate whose part was also needed had succeeded (5.43), and a member who became the one who made the difference drew the most blame. Lagnado, Gerstenberg and Zultan (2013) found that responsibility rises both with pivotality, how close a person was to making the difference, and with criticality, how much the outcome depended on them beforehand. Here each story has three people whose parts were all needed, with one, two or all three failing, or any one of whose parts would have been enough, with all three failing; each person is rated for how responsible they are for what happened. The main score is the mean rating of a lone failure minus that of each of three failures when every part was needed.
Scores
- Pivotality Effectscore
- Mean responsibility rating of a lone failure minus that of each of three failures, both when every part was needed, from -6 to 6.
- Positive means being the one who made the difference drew more responsibility
- Any One Enough Minus All Neededscore
- Mean rating of each of three failures when any one part was enough minus that when every part was needed, from -6 to 6.
- Positive means each failure counted more when any one person could have prevented the outcome; negative means it counted more when every part was needed
- Lone Failure Ratingscore
- Mean rating, from 1 to 7, of the one person who failed when every part was needed.
- Informational
- One of Two Failures Ratingscore
- Mean rating, from 1 to 7, of each of two people who failed when every part was needed.
- Informational
- One of Three Failures Ratingscore
- Mean rating, from 1 to 7, of each of three people who failed when every part was needed.
- Informational
- Any One Enough Ratingscore
- Mean rating, from 1 to 7, of each of three people who failed when any one part would have been enough.
- Informational
- Did Their Part Ratingscore
- Mean rating, from 1 to 7, of people who did their part.
- Low expected
- Responsibility Added Up, One Failedscore
- With one failure, the three people's ratings read from 0 (rated 1) to 1 (rated 7) and added up, from 0 to 3.
- Informational
- Responsibility Added Up, Two Failedscore
- The same sum in stories where two of the three failed.
- Informational
- Responsibility Added Up, All Failedscore
- The same sum in stories where all three failed and every part was needed.
- About 1 means a fixed amount shared out; higher means each failure kept much of a full share
- Responsibility Added Up, Any One Enoughscore
- The same sum in stories where all three failed and any one part would have been enough.
- Informational
- Median Reading Timems
- Median time spent on a story screen before pressing Next.
- Informational
- Median Rating Timems
- Median time from a rating screen appearing to a rating, over ratings at or after 100 ms.
- Informational
- Ratings Under 100 mscount
- Ratings made less than 100 ms after their screen appeared.
- 0 expected
- Stories Ratedcount
- Stories with all three people rated.
- Informational
Use and evidence
48 papers from 2021 to 2025 name this task in their title or abstract, down 9% compared with 2016 to 2020. That ranks it 473 of 711 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 14 |
| 2024 | 10 |
| 2023 | 11 |
| 2022 | 5 |
| 2021 | 8 |
| 2020 | 17 |
| 2019 | 9 |
| 2018 | 7 |
| 2017 | 12 |
| 2016 | 8 |
| 2015 | 13 |
| 2014 | 14 |
| 2013 | 5 |
| 2012 | 6 |
| 2011 | 8 |
| 2010 | 9 |
| 2009 | 12 |
| 2008 | 7 |
| 2007 | 1 |
| 2006 | 8 |
| 2005 | 3 |
| 2004 | 2 |
| 2003 | 9 |
| 2002 | 0 |
| 2001 | 2 |
| 2000 | 3 |
Use grade D
30 to 99 papers from 2021 to 2025.
Evidence grade D
Little published measurement evidence found.
- Papers on test-retest reliability
- 1
- Papers on reliability or validity
- 14
- Registered clinical studies using it as an outcome
- 0
Measured on ConductCognition
Pooled across the studies that feed our platform norms, by device group. A figure needs 50 or more people.
No pooled figure yet.
Note on the counts: also counts responsibility judgments outside moral psychology.
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 0.58.0, counted 14 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Stories
- 12 written for this task, each three people's plan that went wrong, such as a surprise party where one forgot the cake, with one sentence that says how their parts combine. In 9 every part was needed and one, two or all three failed; in 3 any one part would have been enough and all three failed. None has political, religious, group, health or injury content, and none is a published story. They are shown in the session's language.
- Ratings
- After each story, one screen for each person in turn: how responsible that person is for what happened, from 1 (not at all) to 7 (very much), on seven buttons, with the person's part and the story's rule on screen. A check with feedback first makes sure people know they rate each of the three.
- Order
- 12 stories as standard, or the first 8. The four kinds of story come in blocks of four, one of each, in an order drawn from the session's seed, so every part of the session holds every kind and no kind comes twice in a row.
- Scores
- The pivotality effect: the mean rating of a lone failure minus that of each of three failures when every part was needed, from -6 to 6. Also the contrast between failures when any one part was enough and when every part was needed, the five means behind them, and for each kind of story the three ratings added up.
Gerstenberg and Lagnado (2010) found people give less responsibility to a member who would have needed more of the others' results to change to make the difference, and Lagnado, Gerstenberg and Zultan (2013) showed responsibility also rises with how much the outcome depended on that member. This task uses their method with everyday stories of its own, none from their games. Scores describe how a person shares out responsibility and are best compared between groups; they 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 ratings of each person in each story. No rating is correct, so accuracy is not judged against chance. A run with no lone failure or no failure shared by all three rated, or with fewer than half the stories rated for all three people, is invalid, and any other shortfall is flagged; one number chosen for at least 90% of the ratings is flagged. Ratings under 200 ms on at least half the screens invalidate.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. The scores describe how a person's part in a group outcome moves ratings of their responsibility, so any rating is a possible answer. The main score needs a lone failure and failures shared by all three rated. One number for nearly every rating suggests the scale was not used; it is flagged for review because it can also be a real answer. 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 |
|---|---|---|
| Stories | 12 stories (standard) | 8 stories, 12 stories (standard) |
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 Who Is Responsible Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Gerstenberg T, Lagnado DA, 2010; Zultan R, Gerstenberg T, Lagnado DA, 2012; Lagnado DA, Gerstenberg T, Zultan R, 2013), with the standard settings (stories: 12 stories). The main score was pivotality effect (mean responsibility rating of a lone failure minus that of each of three failures, both when every part was needed, from -6 to 6).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Gerstenberg, T., & Lagnado, D. A. (2010). Spreading the blame: The allocation of responsibility amongst multiple agents. Cognition, 115(1), 166-171. https://doi.org/10.1016/j.cognition.2009.12.011
- Zultan, R., Gerstenberg, T., & Lagnado, D. A. (2012). Finding fault: Causality and counterfactuals in group attributions. Cognition, 125(3), 429-440. https://doi.org/10.1016/j.cognition.2012.07.014A member's blame for a team loss fell when a teammate who could stand in for them succeeded, but not when a teammate whose part was also needed did.
- Lagnado, D. A., Gerstenberg, T., & Zultan, R. (2013). Causal responsibility and counterfactuals. Cognitive Science, 37(6), 1036-1073. https://doi.org/10.1111/cogs.12054Responsibility rose with pivotality when criticality was held constant, and with criticality when pivotality was held constant.
Source papers
- Zultan and colleagues (2012), first experiment: 83 students rated each of four chefs' blame for a team loss from 1 to 7; a member whose stand-in had succeeded 4.46, a member when all four failed 5.55.Paradigm source, Adults
Questions
How long does the Who Is Responsible Task take?
About 6 to 8 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?
Pivotality Effect, Any One Enough Minus All Needed, Lone Failure Rating, One of Two Failures Rating, One of Three Failures Rating, Any One Enough Rating, Did Their Part Rating, Responsibility Added Up, One Failed, Responsibility Added Up, Two Failed, Responsibility Added Up, All Failed, Responsibility Added Up, Any One Enough, Median Reading Time, Median Rating Time, Ratings Under 100 ms and Stories Rated. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Who Is Responsible Task used in research?
48 papers from 2021 to 2025 name it in their title or abstract, down 9% compared with 2016 to 2020. That ranks it 473 of 711 test families in the ConductCognition test census (use grade D). Source: ConductCognition test census 0.58.0, counted 14 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Who Is Responsible Task have?
In the published research, 1 paper reports its test-retest reliability and 0 registered clinical studies use it as an outcome (evidence grade D in the ConductCognition test census).
Does it report a percentile?
No. Scores and quality flags only.
Which settings can a researcher change?
Stories (8 stories, 12 stories (standard)). 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, with its words in the same language. 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.