Social Cognition
Others' Expectations Task
Also called: guilt aversion task, trust game with belief elicitation, second-order beliefs in the trust game, trust responsiveness
Measure guilt aversion: participants say how many points programmed partners expect back, then return points to them, scored by how their returns follow those beliefs.
5-7 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
- 59
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
Measures how much people live up to what others expect of them. In each round a programmed partner sends some of its 10 points, the participant receives 3 times the amount, says how many points the partner expects back, then chooses how many to send back. The main score is how the points sent back follow those beliefs within each person.
Constructs
- Guilt aversion
- Second-order beliefs
- Trustworthiness
- Social decision making
Research uses
- Behavioral economics research
- Trust, reciprocity and moral emotion studies
- Social decision-making research in clinical groups
- Within-study group comparison
How the paradigm works
The Others' Expectations Task measures guilt aversion: how far people live up to what they believe others expect of them (Charness and Dufwenberg, 2006; Battigalli and Dufwenberg, 2007). The participant plays the second player of the trust game against partners that are computer programs following set rules, and by default the instructions say so. A neutral wording says only "your partner", and the last screen always says the partners were computer programs.
More about the paradigm
In each round a new partner has 10 points and sends some of them, and the participant receives three times the amount, as in the Trust Game. Partners send 2, 4, 6, 8 or 10 points: each amount once in the first five rounds and once more in the next five, each half in an order drawn from the session's seed, 10 rounds as standard (5 to 10). Before each return the participant says how many points they think the partner expects back, on the Trust Game's amount bar, then chooses how many of the points received to send back. One comprehension question with feedback comes first. The study decides what the points are worth, and the payment line says whether they are exchanged for money.
The main score is the expectation slope: the least-squares slope of the points sent back on the points the participant thought the partner expected, over that person's rounds. A slope of 1 means one more point sent back for each point more expected. The expectation shortfall is the share of the expected points that were not sent back, and expectations met is the share of rounds where the return reached the expectation. Charness and Dufwenberg found that second players who cooperated had guessed higher expectations than those who did not, in all five of their conditions. When Ellingsen and colleagues (2010) told second players what their own partner had expected, returns did not follow, and they read the earlier pattern as people thinking others think like themselves. Reading beliefs and returns within one person sets aside such differences between people; the amount a partner sends moves both, so the slope describes how returns follow beliefs as trust varies. Scores describe answers in a game played for points and support research; they do not describe a person's character and do not diagnose.
Scores
- Expectation Slopescore
- Least-squares slope of the points sent back on the points the participant thought the partner expected, over the rounds; needs 3 rounds and beliefs that differ.
- Higher means returns follow beliefs more closely
- Expectation Shortfall%
- Points thought expected but not sent back, summed over rounds, as a share of the points thought expected.
- 0 means every expectation was met
- Expectations Met%
- Share of rounds in which the points sent back were at least the points thought expected.
- Higher means more expectations met
- Share Thought Expected%
- Points thought expected as a share of the points received, averaged over rounds.
- Informational
- Share Returned%
- Points sent back as a share of the points received, averaged over rounds.
- Higher means more trustworthiness
- Comprehension Checkflag
- 1 when the comprehension question was answered correctly, 0 when not.
- 1 supports interpretation
- Median Return Timems
- Median time from a return screen to Confirm, over returns confirmed at or after 100 ms.
- Informational
- Anticipationscount
- Belief and return screens confirmed under 100 ms.
- Lower is better
- Roundscount
- Rounds with both a belief and a return.
- Informational
- Planned Roundscount
- Rounds the session's settings plan.
- Informational
- Points Keptpoints
- Points received minus points sent back, over the rounds.
- Informational
Use and evidence
59 papers from 2021 to 2025 name this task in their title or abstract, up 4% compared with 2016 to 2020. That ranks it 419 of 711 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 15 |
| 2024 | 9 |
| 2023 | 13 |
| 2022 | 9 |
| 2021 | 13 |
| 2020 | 7 |
| 2019 | 7 |
| 2018 | 20 |
| 2017 | 11 |
| 2016 | 12 |
| 2015 | 8 |
| 2014 | 9 |
| 2013 | 8 |
| 2012 | 11 |
| 2011 | 15 |
| 2010 | 8 |
| 2009 | 8 |
| 2008 | 2 |
| 2007 | 3 |
| 2006 | 5 |
| 2005 | 0 |
| 2004 | 0 |
| 2003 | 0 |
| 2002 | 0 |
| 2001 | 0 |
| 2000 | 0 |
Use grade D
30 to 99 papers from 2021 to 2025.
Evidence grade D
Little published measurement evidence found.
- Papers on test-retest reliability
- 0
- Papers on reliability or validity
- 5
- 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.
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.
- Partners
- Computer programs that follow set rules, named by letter, a new one each round. The standard instructions say so; a neutral wording says only "your partner". The last screen always says the partners were computer programs.
- Rounds
- 10 (standard; 5 to 10). Each partner has 10 points and sends 2, 4, 6, 8 or 10 of them: each amount once in the first 5 rounds and once more in the next 5, in orders drawn from the session's seed. The participant receives three times the amount.
- Each round
- First, how many points the participant thinks the partner expects back; then how many of the points received to send back. Both on the Trust Game's amount bar, from 0 to the points received.
- Check
- One comprehension question with feedback before the first round.
- Points
- The study decides what points are worth, and the payment line says whether they are exchanged for money.
- Scores
- The expectation slope: the least-squares slope of the points sent back on the points thought expected, over each person's rounds. Also the share of expected points not sent back, the share of rounds where the return met the expectation, and the mean shares expected and returned.
Charness and Dufwenberg (2006) asked second players after the game what their partners expected and compared those guesses across people. Here each person states a belief before every return, against partners who send different amounts, so returns are read against beliefs within each person. When Ellingsen and colleagues (2010) told second players their partner's actual expectation, returns did not follow it, so read the slope as how returns follow stated beliefs and compare groups. Scores do not describe a person's character 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 belief and return screens. No amount is correct, so accuracy is not judged against chance, and the same amount on every screen is accepted. Screens confirmed under 200 ms on at least half the screens invalidate, and so does a run with fewer than half its planned rounds. A shorter run, beliefs that never change and a wrong comprehension answer are flagged for review.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. Returning or expecting the same share every time is a stable answer, so repeated amounts are not checked. With fewer than half the rounds, the returns are compared with the beliefs over too few partners. Beliefs that never change leave the slope empty without making the answers wrong. A participant who misreads how points are tripled or kept may choose the opposite of what they intend, so a wrong comprehension 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 |
|---|---|---|
| Rounds | 10 | 5 to 10 |
| Partner wording | A computer program (standard) | A computer program (standard), Neutral: your partner |
| 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 Others' Expectations Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Charness G, Dufwenberg M, 2006; Battigalli P, Dufwenberg M, 2007), with the standard settings (rounds: 10; partner wording: A computer program; payment line: Hypothetical points). The main score was expectation slope (least-squares slope of the points sent back on the points the participant thought the partner expected, over the rounds; needs 3 rounds and beliefs that differ).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Charness, G., & Dufwenberg, M. (2006). Promises and partnership. Econometrica, 74(6), 1579-1601. https://doi.org/10.1111/j.1468-0262.2006.00719.xSecond players who cooperated had guessed higher expectations than those who did not, in all five conditions.
- Battigalli, P., & Dufwenberg, M. (2007). Guilt in games. American Economic Review, 97(2), 170-176. https://doi.org/10.1257/aer.97.2.170Guilt as a dislike of letting others down relative to what they expect.
- Vanberg, C. (2008). Why do people keep their promises? An experimental test of two explanations. Econometrica, 76(6), 1467-1480. https://doi.org/10.3982/ECTA7673Promises kept beyond what changed expectations explain, a reason for returning trust beside guilt aversion.
- Ellingsen, T., Johannesson, M., Tjøtta, S., & Torsvik, G. (2010). Testing guilt aversion. Games and Economic Behavior, 68(1), 95-107. https://doi.org/10.1016/j.geb.2009.04.021Returns did not follow partners' expectations when second players were told them.
Source papers
- Charness and Dufwenberg (2006): second-order beliefs and cooperation in a trust game. Ellingsen, Johannesson, Tjøtta and Torsvik (2010): returns when partners' expectations are told.Paradigm source, Adults
Questions
How long does the Others' Expectations Task take?
About 5 to 7 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?
Expectation Slope, Expectation Shortfall, Expectations Met, Share Thought Expected, Share Returned, Comprehension Check, Median Return Time, Anticipations, Rounds, Planned Rounds and Points Kept. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Others' Expectations Task used in research?
59 papers from 2021 to 2025 name it in their title or abstract, up 4% compared with 2016 to 2020. That ranks it 419 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 Others' Expectations Task have?
In the published research, 0 papers report 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?
Rounds (5 to 10), partner wording (A computer program (standard), Neutral: your partner) 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.
Related tests