Executive Function
Beads Task
Also called: jumping to conclusions task, beads in a jar task, urn task, draws to decision task
Measure how much evidence people gather before deciding: beads come one at a time from a hidden jar, and the score is how many are seen before the choice.
5-8 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
- 68
- Evidence grade
- C
- Registered clinical studies
- 20
What it measures
Measures how much evidence people gather before deciding. Beads are drawn one at a time from one of two hidden jars of orange and blue beads in opposite proportions, and participants decide which jar they come from whenever they choose. The main score is the number of beads seen before deciding, with the share of decisions made after one or two beads.
Constructs
- Evidence gathering
- Probabilistic reasoning
- Jumping to conclusions
Research uses
- Psychosis and delusion research
- Reasoning and decision-making research
- Cognitive-bias intervention studies
- Within-study group comparison
How the paradigm works
The beads task measures how much evidence people gather before deciding (Huq, Garety and Hemsley, 1988). Two jars hold orange and blue beads in opposite proportions: jar A has 85 orange and 15 blue, jar B 85 blue and 15 orange. In each round the computer hides one of the jars and shows its beads one at a time, each put back before the next; the participant asks for another bead or decides which jar the beads come from, up to 20 beads. Deciding after very few beads is called jumping to conclusions, a reasoning style studied in research on delusions and psychosis.
More about the paradigm
The standard session has 6 rounds with the jars of 85 to 15, then 6 with the harder jars of 60 to 40 (Garety, Hemsley and Wessely, 1991), because several sequences in one session give steadier scores than one (McLean, Mattiske and Balzan, 2018). The bead sequences are written by our code by stated rules and are the same for every session: each starts with a bead of its jar's main color, and the beads seen never favor the other jar. The session's seed decides which jar each round comes from, half from each. A question about the jars, with feedback, and one practice round come first. Beads always come from the hidden jar; none is picked from a set on the screen.
The main score is draws to decision: the number of beads seen before deciding, averaged over the rounds, and for each pair of jars. The share of rounds decided after 1 or 2 beads, the criterion of Dudley and colleagues (2016) for jumping to conclusions, the beads seen in the first round, the probability that the chosen jar was the source given the beads seen, decisions against the evidence and the time to decide are also reported. Presses under 100 ms after a bead appears are counted as anticipations and not taken. Scores support research and do not diagnose.
Scores
- Draws to Decisioncount
- Beads seen before deciding, averaged over the rounds.
- Lower means deciding on less evidence
- Draws to Decision, 85 to 15count
- Beads seen before deciding in the rounds with the jars of 85 to 15.
- Lower means deciding on less evidence
- Draws to Decision, 60 to 40count
- Beads seen before deciding in the rounds with the jars of 60 to 40.
- Lower means deciding on less evidence
- Decisions After 1 or 2 Beads%
- Rounds decided after 1 or 2 beads.
- Higher means more jumping to conclusions
- First Round Drawscount
- Beads seen before deciding in the first round, the round a single-round study scores.
- 2 or fewer is jumping to conclusions by the usual criterion
- Probability at Decision%
- The probability, from the beads seen, that the chosen jar was the source, averaged over the rounds.
- Informational
- Decisions Against the Evidencecount
- Rounds in which the chosen jar was the one the beads seen argued against.
- Lower is better; more than chance is flagged
- Decisions After All Beadscount
- Rounds decided only after all 20 beads.
- Informational
- Decision Timems
- Median time from the last bead to the decision.
- Informational
- Anticipationscount
- Presses under 100 ms after a bead appeared, which were not taken.
- Lower is better
- Jar Question Failedflag
- 1 when the question about which jar has more blue beads was not answered right in 3 tries.
- 1 is flagged for review
- Planned Roundstrials
- Rounds the session planned to show.
- Informational
- Total Roundstrials
- Rounds shown in the scored run.
- Fewer than planned is flagged for review
Use and evidence
68 papers from 2021 to 2025 name this task in their title or abstract, down 19% compared with 2016 to 2020. That ranks it 226 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 18 |
| 2024 | 10 |
| 2023 | 9 |
| 2022 | 15 |
| 2021 | 16 |
| 2020 | 24 |
| 2019 | 13 |
| 2018 | 23 |
| 2017 | 7 |
| 2016 | 17 |
| 2015 | 18 |
| 2014 | 12 |
| 2013 | 13 |
| 2012 | 10 |
| 2011 | 9 |
| 2010 | 8 |
| 2009 | 3 |
| 2008 | 7 |
| 2007 | 0 |
| 2006 | 2 |
| 2005 | 2 |
| 2004 | 0 |
| 2003 | 2 |
| 2002 | 0 |
| 2001 | 2 |
| 2000 | 0 |
Use grade D
30 to 99 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
- 2
- Papers on reliability or validity
- 29
- Registered clinical studies using it as an outcome
- 20
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.
- Jars
- Jar A holds 85 orange and 15 blue beads, jar B 85 blue and 15 orange; the harder jars hold 60 and 40. Each jar is shown as a drawing with its counts, and the round's jar is hidden.
- Sequences
- 20 beads each, the same for every session: 3 of the jar's minority color with the jars of 85 to 15, 8 with the jars of 60 to 40. The first bead is the jar's main color and the beads seen never favor the other jar.
- Rounds
- 6 rounds with the jars of 85 to 15, then 6 with the jars of 60 to 40 (standard; 3 to 8 each, or one pair of jars only). The session's seed decides which jar each round comes from, half from each.
- Answer
- Draw another bead or choose jar A or jar B, up to 20 beads, with no time limit. A press under 100 ms after a bead appears is counted as an anticipation and not taken.
- Practice
- A question about which jar has more blue beads, with feedback and up to 3 tries, then one practice round that ends by saying which jar it was. There is no feedback in the test.
- Data
- The beads shown in order with the time of each, the decision and its time, and the sequence and jar of each round, from which the scorer counts the beads again.
Huq, Garety and Hemsley used one sequence; several rounds in one session follow McLean and colleagues, who found one sequence unsteady from one session to the next. Draws to decision has no better direction, so no percentile is given.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: beads_task_round from the scored run, one per round. A decision counts as correct when it names the jar the beads seen favor (chance 50%); a round decided when the beads seen are even has no correct answer. The number of beads is the measure and is not judged. Below chance invalidates and not above chance is flagged; one jar on at least 90% of decisions invalidates when accuracy is not above chance; decisions under 200 ms after the last bead on at least half the rounds invalidate. A jar question never answered right, or a short run, is flagged for review.
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. 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. Every sequence favors the jar it comes from or is even, so naming the other jar means the jars were confused, and the question before the practice checks that they were understood. Those two limits are ConductScience choices: the sources set none.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Rounds for each pair of jars | 6 | 3 to 8 |
| Jars | 85 to 15, then 60 to 40 (standard) | 85 to 15, then 60 to 40 (standard), 85 to 15 only, 60 to 40 only |
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 Beads Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Huq, Garety and Hemsley, 1988), with the standard settings (rounds for each pair of jars: 6; jars: 85 to 15, then 60 to 40). The main score was draws to decision (beads seen before deciding, averaged over the rounds).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Huq, S. F., Garety, P. A., & Hemsley, D. R. (1988). Probabilistic judgements in deluded and non-deluded subjects. The Quarterly Journal of Experimental Psychology Section A, 40(4), 801-812. https://doi.org/10.1080/14640748808402300The method: two jars in opposite proportions, beads drawn one at a time until the participant decides which jar they come from.
- Garety, P. A., Hemsley, D. R., & Wessely, S. (1991). Reasoning in deluded schizophrenic and paranoid patients: Biases in performance on a probabilistic inference task. The Journal of Nervous and Mental Disease, 179(4), 194-201. https://doi.org/10.1097/00005053-199104000-00003The harder jars of 60 to 40.
- McLean, B. F., Mattiske, J. K., & Balzan, R. P. (2018). Towards a reliable repeated-measures beads task for assessing the jumping to conclusions bias. Psychiatry Research, 265, 200-207. https://doi.org/10.1016/j.psychres.2018.04.043Several sequences in one session for a steadier score.
- Dudley, R., Taylor, P., Wickham, S., & Hutton, P. (2016). Psychosis, delusions and the "jumping to conclusions" reasoning bias: A systematic review and meta-analysis. Schizophrenia Bulletin, 42(3), 652-665. https://doi.org/10.1093/schbul/sbv150Draws to decision and the criterion of a decision after 1 or 2 beads.
Source papers
- Huq SF, Garety PA, Hemsley DR (1988). Probabilistic judgements in deluded and non-deluded subjects. The Quarterly Journal of Experimental Psychology Section A, 40(4), 801-812.Paradigm source, Adults in the source
Questions
How long does the Beads Task take?
About 5 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?
Draws to Decision, Draws to Decision, 85 to 15, Draws to Decision, 60 to 40, Decisions After 1 or 2 Beads, First Round Draws, Probability at Decision, Decisions Against the Evidence, Decisions After All Beads, Decision Time, Anticipations, Jar Question Failed, Planned Rounds and Total Rounds. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Beads Task used in research?
68 papers from 2021 to 2025 name it in their title or abstract, down 19% compared with 2016 to 2020. That ranks it 226 of 395 test families in the ConductCognition test census (use grade D). Source: ConductCognition test census 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Beads Task have?
In the published research, 2 papers report its test-retest reliability and 20 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 for each pair of jars (3 to 8) and jars (85 to 15, then 60 to 40 (standard), 85 to 15 only, 60 to 40 only). 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.