Social Cognition
Hidden Ball Beliefs Task
Also called: belief reasoning task, self-other conflict task, cups false belief task
Measure how another person's belief and your own knowledge interfere: participants say where a player thinks a hidden ball is, or where they think it is, after a clue only they saw.
12-14 minTimed: desktop or laptop with a mouse or trackpad
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
- 5
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
Measures how another person's belief and one's own knowledge interfere. A ball is hidden under one of three cups; a clue shows where it is or that one cup is empty, and two players point to where they think it is. Participants press the cup a player thinks the ball is under, or the cup they think it is under themselves. The main score is the self-other conflict: how much slower answers are when the player's belief conflicts with what the participant knows.
Constructs
- Theory of mind
- Belief reasoning
- Self-perspective inhibition
- Executive control
Research uses
- Theory of mind research in adults
- Aging and social cognition research
- Perspective taking studies
- Within-study group comparison
How the paradigm works
The Hidden Ball Beliefs Task measures how another person's belief and one's own knowledge interfere. Rahman, Kessler, Apperly, Hansen, Javed, Holland and Hartwright (2021) built it on the belief reasoning tasks of Apperly and colleagues (2011) and Hartwright, Apperly and Hansen (2012) to separate three sources of conflict that classic false belief tasks mix: knowing where something really is, a belief that conflicts with one's own, and two cued places to keep apart. As in Apperly and colleagues (2008), the beliefs are shown rather than inferred. In their 100 younger and older adults, answers were about 97 ms slower when the player's belief conflicted with what the participant knew, about 48 ms slower when the two players disagreed, and about 48 ms faster when the ball was shown.
More about the paradigm
In each round a ball is hidden under one of three cups and the cups are shuffled for 1,000 ms. Three pictures of 1,000 ms follow in a seeded order, each with a 500 ms mark after it: a clue, in which one cup is lifted over the ball or over a green X that shows the cup empty, and the cup that each of two players, blue and red, points to. The players do not see the clue. A question picture then asks where the blue player thinks the ball is, where the red player thinks it is, or where the participant thinks it is, answered with the 1, 2 and 3 keys within 3,000 ms. The rounds about a player cross three factors: the ball shown or a cup shown empty, the players at the same cup or different cups, and the asked player's cup fitting what the participant knows or ruled out by it. The rounds that ask the participant come from the same conditions, so the question cannot be told from the pictures, and they show whether the clue was followed. The standard session has 2 blocks of 36 rounds, 6 of each condition about a player and 24 that ask the participant; 6 blocks give the 216 rounds of Rahman and colleagues. Six practice rounds have feedback. The pictures are drawn by ConductScience's code; no picture or script of the original task is used. About 12 to 14 minutes.
The main score is the self-other conflict: the mean time of the right answers in the four conditions where the asked player's belief conflicts with what the participant knows, minus the four where it fits. The test also reports the same in errors, the conflict when the two players disagree, the difference when the ball is shown, accuracy on the rounds about a player and on the rounds that ask the participant, and the mean answer time. A condition's time needs 3 right answers, and times beyond 2 SD of the condition's mean are left out, as in the source. A session needs two thirds of the rounds that ask the participant right. Scores support research and do not diagnose.
Scores
- Self-Other Conflictms
- Mean right-answer time when the asked player's belief conflicts with what the participant knows, minus when it fits.
- Above zero means the conflict slowed the answers
- Self-Other Conflict, Errors%
- The same difference in percent errors, a missed answer counted as an error.
- Above zero means more errors under the conflict
- Other-Other Conflictms
- Mean right-answer time when the two players point to different cups, minus the same cup.
- Above zero means two places to keep apart slowed the answers
- Other-Other Conflict, Errors%
- The same difference in percent errors.
- Above zero means more errors when the players disagree
- Ball Shown Minus Cup Shown Emptyms
- Mean right-answer time when the clue showed the ball, minus when it showed an empty cup.
- Informational; below zero means faster when the ball was shown
- Ball Shown Minus Cup Shown Empty, Errors%
- The same difference in percent errors.
- Informational
- Accuracy About the Players%
- Percent of rounds about a player answered with that player's cup.
- Higher is better
- Accuracy, Your Own View%
- Percent of rounds that ask the participant answered with a cup that can hold the ball.
- Below two thirds invalidates
- Mean Answer Timems
- Mean time from the question to a right answer about a player, at or after 100 ms.
- Informational
- Missed Answerscount
- Rounds with no answer within 3,000 ms.
- Lower supports interpretation
- Anticipationscount
- Answers under 100 ms, left out of the times.
- Lower supports interpretation
- Rounds Reachedtrials
- Rounds in the scored run.
- Fewer than planned is flagged for review
- Rounds Plannedtrials
- Rounds the session set out to show.
- Informational
Use and evidence
5 papers from 2021 to 2025 name this task in their title or abstract. That ranks it 472 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 0 |
| 2024 | 0 |
| 2023 | 4 |
| 2022 | 0 |
| 2021 | 1 |
| 2020 | 0 |
| 2019 | 0 |
| 2018 | 0 |
| 2017 | 0 |
| 2016 | 0 |
| 2015 | 0 |
| 2014 | 0 |
| 2013 | 0 |
| 2012 | 0 |
| 2011 | 0 |
| 2010 | 0 |
| 2009 | 0 |
| 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 D
Little published measurement evidence found.
- Papers on test-retest reliability
- 0
- Papers on reliability or validity
- 0
- Registered clinical studies using it as an outcome
- 0
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 0.22.0, counted 2 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Round
- The cups are shuffled for 1,000 ms; then the clue and each player's guess, 1,000 ms each in a seeded order, with a 500 ms mark after each; then the question until a key or 3,000 ms.
- Clue
- One cup lifted over the ball, so its place is known, or over a green X, so that cup is empty. The players do not see it.
- Question
- Where the blue player or the red player thinks the ball is, or where the participant thinks it is, answered with 1, 2 or 3 for the left, middle or right cup.
- Conditions
- 8 conditions about a player: the ball shown or a cup shown empty, the players at the same cup or different cups, and the asked player's cup fitting what the participant knows or ruled out by it.
- Session
- 2 blocks of 36 rounds by default, or 4 or 6 as a setting. Each pair of blocks holds 6 rounds of each condition about a player and 24 rounds that ask the participant, drawn from the same conditions. The session's seed draws every round, with no more than 3 rounds in a row with the same right cup.
- Practice
- 6 rounds with feedback. The test has no feedback.
The pictures were drawn for this test. Rahman and colleagues (2021) ran 216 rounds; the standard session here has 72, and 6 blocks give 216. The question stays on the screen until an answer, where theirs showed for 1,000 ms.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: hidden_ball_test from the scored run, every round: 1, 2 or 3 for the left, middle or right cup within 3,000 ms of the question. Chance is one in three for a round about a player and for a "you" round with the ball shown, and two in three for a "you" round with a cup shown empty. Below chance invalidates and not above chance is flagged. One key on 90% of answers invalidates when accuracy is not above chance. Missed responses and responses under 100 ms follow the timed-test rule. Fewer than two thirds of the "you" rounds right invalidates. A cost that cannot be computed, with fewer than 3 right answers in a condition, and a run short of its rounds are flagged for review. A hidden tab withholds the percentile.
Why these rules
Accuracy significantly below chance (one-sided exact test, p < .05) cannot come from guessing; it means the response mapping was reversed or the task was not followed (Slick, Sherman and Iverson 1999 treat below-chance forced-choice performance as definite response bias). Chance-level accuracy can be a real result in impaired participants, so it is only flagged. One option on at least 90% of answered trials, when the correct options are balanced and accuracy is not above chance, means the answers did not depend on the items (the task counterpart of long-string responding in surveys; Curran 2016). Missed responses and responses under 100 ms keep the platform's timed-test rule: more than max(3, 25%) missed responses or more than max(2, 20%) responses under 100 ms invalidate, and any is flagged. A response under 100 ms cannot be a reaction to the stimulus. The "you" rounds ask where the ball is, which needs only the clue: Rahman and colleagues (2021) used them to find participants who were not following the clue, without which no belief conflicts with what the participant knows. A participant who guesses on them gets about half right, so below two thirds the answers do not show the clue was followed. A mean time from fewer than 3 right answers would be too unstable to compare. A hidden tab lets the study, delay or response timers run without the participant, so a percentile is withheld after one.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Blocks | 2 blocks, 72 rounds (standard) | 2 blocks, 72 rounds (standard), 4 blocks, 144 rounds, 6 blocks, 216 rounds |
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
- Timed test: participants need a desktop or laptop with a mouse or trackpad. 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 Hidden Ball Beliefs Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Rahman F, Kessler K, Apperly IA, Hansen PC, Javed S, Holland CA, Hartwright CE, 2021; Apperly IA, Warren F, Andrews BJ, Grant J, Todd S, 2011; Hartwright CE, Apperly IA, Hansen PC, 2012), with the standard settings (blocks: 2 blocks, 72 rounds). The main score was self-other conflict (mean right-answer time when the asked player's belief conflicts with what the participant knows, minus when it fits).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Rahman, F., Kessler, K., Apperly, I. A., Hansen, P. C., Javed, S., Holland, C. A., & Hartwright, C. E. (2021). Sources of cognitive conflict and their relevance to theory-of-mind proficiency in healthy aging: A preregistered study. Psychological Science, 32(12), 1918-1936. https://doi.org/10.1177/09567976211017870The task: three sources of conflict in belief reasoning, in younger and older adults.
- Apperly, I. A., Warren, F., Andrews, B. J., Grant, J., & Todd, S. (2011). Developmental continuity in theory of mind: Speed and accuracy of belief-desire reasoning in children and adults. Child Development, 82(5), 1691-1703. https://doi.org/10.1111/j.1467-8624.2011.01635.xSpeed and accuracy of belief reasoning, on which the task builds.
- Hartwright, C. E., Apperly, I. A., & Hansen, P. C. (2012). Multiple roles for executive control in belief-desire reasoning: Distinct neural networks are recruited for self perspective inhibition and complexity of reasoning. NeuroImage, 61(4), 921-930. https://doi.org/10.1016/j.neuroimage.2012.03.012Inhibiting one's own perspective, on which the task builds.
- Apperly, I. A., Back, E., Samson, D., & France, L. (2008). The cost of thinking about false beliefs: Evidence from adults' performance on a non-inferential theory of mind task. Cognition, 106(3), 1093-1108. https://doi.org/10.1016/j.cognition.2007.05.005A task in which people are told the belief rather than inferring it.
Source papers
- Rahman F, Kessler K, Apperly IA, Hansen PC, Javed S, Holland CA, Hartwright CE (2021). Sources of cognitive conflict and their relevance to theory-of-mind proficiency in healthy aging: A preregistered study. Psychological Science, 32(12):1918-1936.Paradigm source, Adults 18-29 and 60-79
Questions
How long does the Hidden Ball Beliefs Task take?
About 12 to 14 minutes, as listed in the test library.
What device does a participant need?
Timed test: participants need a desktop or laptop with a mouse or trackpad. The browser must be Chrome, Edge, Firefox or Safari.
What scores does it produce?
Self-Other Conflict, Self-Other Conflict, Errors, Other-Other Conflict, Other-Other Conflict, Errors, Ball Shown Minus Cup Shown Empty, Ball Shown Minus Cup Shown Empty, Errors, Accuracy About the Players, Accuracy, Your Own View, Mean Answer Time, Missed Answers, Anticipations, Rounds Reached and Rounds Planned. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Hidden Ball Beliefs Task used in research?
5 papers from 2021 to 2025 name it in their title or abstract. That ranks it 472 of 495 test families in the ConductCognition test census (use grade E). Source: ConductCognition test census 0.22.0, counted 2 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Hidden Ball Beliefs 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?
Blocks (2 blocks, 72 rounds (standard), 4 blocks, 144 rounds, 6 blocks, 216 rounds). 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.