Executive Function
Contingency Judgment Task
Also called: illusion of causality task, causal illusion task, contingency learning task, outcome density task, covariation judgment task, causal judgment task
Measure the illusion of causality: participants see plants given a fertilizer or not, predict which will flower, then rate each fertilizer, scored as the rating of one that made no difference.
6-9 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
- 73
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
Measures the illusion of causality: judging that a cause works when the outcome is just as likely without it. Participants see plants one at a time, each given a fertilizer or not, predict whether each will flower, then rate how effective each garden's fertilizer is. The main score is the rating of a fertilizer that made no difference while plants usually flowered.
Constructs
- Causal learning
- Contingency judgment
- Illusion of causality
- Judgment under uncertainty
Research uses
- Research on causal learning and contingency judgment
- Research on beliefs in pseudoscience, the paranormal and superstition
- Within-study comparison of judgments under uncertainty
How the paradigm works
This test tells a garden story, with no medicine or illness. In each of three gardens, three plants in four get the garden's fertilizer. For each plant the participant sees whether it got the fertilizer, predicts whether it will flower (Yes or No, by click, tap, or F and J), with no time limit, and then sees it in flower or not. After the garden's last plant, the participant rates how effective its fertilizer is on a bar from not effective at all to perfectly effective, scored 0 to 100. In garden 1 the fertilizer makes no difference and plants usually flower (75% with it and 75% without it); in garden 2 it makes no difference and plants seldom flower (25% either way); in garden 3 it works (75% with it, 25% without it). The gardens always come in this order, so garden 1's rating is the single-problem measure of the literature, made before any other garden. The plants come in blocks of 16 that hold each garden's exact proportions, in an order drawn from the session's seed; 48 plants per garden by default, or 32 as a setting. About 6 to 9 minutes.
More about the paradigm
The main score is the illusion of causality: the rating of garden 1's fertilizer, where the contingency is zero. The outcome density effect follows (garden 1's rating minus garden 2's, two fertilizers that made no difference), then the ratings of gardens 2 and 3, contingency sensitivity (garden 3's rating minus garden 1's; the two gardens differ only in the plants that got no fertilizer) and, in garden 1, the share of fertilized plants predicted to flower minus the share of unfertilized ones. A session that rates the working fertilizer no higher than garden 2's is flagged for review. A high rating in garden 1 is a common judgment under these conditions; the scores support research and do not diagnose.
Scores
- Illusion of Causalityscore
- Rating of garden 1's fertilizer, from 0 (not effective at all) to 100 (perfectly effective). It made no difference: plants flowered on 75% of trials with it and without it.
- 0 matches the contingency; higher values mean a fertilizer that made no difference was judged effective
- Outcome Density Effectscore
- Garden 1's rating minus garden 2's. Both fertilizers made no difference; plants flowered on 75% of trials in garden 1 and 25% in garden 2.
- Positive: the ineffective fertilizer was rated higher when plants flowered often
- Rare Outcome Ratingscore
- Rating of garden 2's fertilizer, which made no difference while plants seldom flowered (25% with it and 25% without it), from 0 to 100.
- 0 matches the contingency
- Working Fertilizer Ratingscore
- Rating of garden 3's fertilizer, which worked: plants flowered on 75% of trials with it and 25% without it, from 0 to 100.
- Informational; expected to be high
- Contingency Sensitivityscore
- Garden 3's rating minus garden 1's. Fertilized plants flowered on 75% of trials in both gardens; only the plants without fertilizer differ.
- Higher means the ratings used the plants that got no fertilizer
- Prediction Difference%
- In garden 1, the share of fertilized plants predicted to flower minus the share of unfertilized plants.
- 0 matches the contingency; positive values show the illusion in the predictions
- Median Prediction Timems
- Median time to answer whether a plant will flower.
- Informational
- Anticipationscount
- Presses within 100 ms of a plant appearing. They are not taken as predictions.
- Quality check
- Ratings Madecount
- Gardens rated in the scored run.
- Fewer than planned is flagged for review
- Planned Ratingscount
- Gardens to rate: 3.
- Informational
Use and evidence
73 papers from 2021 to 2025 name this task in their title or abstract, up 78% compared with 2016 to 2020. That ranks it 252 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 24 |
| 2024 | 19 |
| 2023 | 13 |
| 2022 | 10 |
| 2021 | 7 |
| 2020 | 8 |
| 2019 | 6 |
| 2018 | 14 |
| 2017 | 6 |
| 2016 | 7 |
| 2015 | 5 |
| 2014 | 5 |
| 2013 | 5 |
| 2012 | 6 |
| 2011 | 5 |
| 2010 | 5 |
| 2009 | 4 |
| 2008 | 4 |
| 2007 | 10 |
| 2006 | 8 |
| 2005 | 10 |
| 2004 | 12 |
| 2003 | 5 |
| 2002 | 6 |
| 2001 | 4 |
| 2000 | 4 |
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
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 3 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Story and pictures
- A gardener tries three fertilizers, each in its own garden. Potted plants and fertilizer bottles are drawn by code; each garden has its own flower color and fertilizer color, and a line under each plant names its fertilizer or says it got none. No medicine or illness appears.
- Gardens
- Three gardens in a fixed order, with the fertilizer on 75% of plants in each. Garden 1: plants flower on 75% of trials with the fertilizer and 75% without it, no contingency with a frequent outcome. Garden 2: 25% and 25%, no contingency with a rare outcome. Garden 3: 75% and 25%, a delta P of 0.5.
- Plants
- 48 per garden by default (32 as a setting), in blocks of 16 (12 with the fertilizer, 4 without) that each hold the garden's exact proportions, in a seeded order.
- Each plant
- The participant predicts whether it will flower, Yes or No, by click, tap, or the F and J keys, with no time limit. The plant then shows in flower or not, with a line saying which, for 1,000 ms, then a 500 ms blank. Presses under 100 ms are not taken.
- Ratings
- After each garden's last plant, how effective its fertilizer is at making the plants flower, on a bar from not effective at all to perfectly effective, scored 0 to 100, with no numbers shown and no starting point.
- Scores
- The illusion of causality (garden 1's rating), the outcome density effect (garden 1's rating minus garden 2's), the ratings of gardens 2 and 3, contingency sensitivity (garden 3's rating minus garden 1's) and the prediction difference in garden 1.
The sources used a medicine and a patient or a button and a light; this test uses a garden, drawn plants and a 0 to 100 bar, so its ratings are not directly comparable with theirs. The three gardens share one session in a fixed order, and what people learn in one garden can carry into the next (Blanco and Matute, 2019), so garden 1's rating, made first, is the main score. The scores support research 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: contingency_judgment_trial from the scored run, a prediction of whether each plant will flower, and contingency_judgment_rating, one rating from 0 to 100 for each of the 3 gardens, with no time limit. A prediction is a guess and a rating is a judgment, so accuracy is not judged against chance. The working fertilizer not rated above the one whose plants seldom flowered, a run short of its planned ratings and one prediction on at least 95% of plants are flagged for review. Predictions under 200 ms on at least half the plants invalidate.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. Garden 3's fertilizer worked and garden 2's plants seldom flowered with or without it, so any reading of what was shown rates garden 3 higher, including one that ignores the plants without fertilizer (the illusion this test measures); the same rating for every garden fails it too. It is flagged, not excluded. Blanco, Matute and Vadillo (2013) excluded a participant who answered No to every prediction; one prediction for nearly every plant is flagged here. 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 |
|---|---|---|
| Plants per garden | 48 plants (standard) | 32 plants (shorter), 48 plants (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 Contingency Judgment Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Alloy and Abramson, 1979; Blanco, Matute and Vadillo, 2013; Matute et al., 2015), with the standard settings (plants per garden: 48 plants). The main score was illusion of causality (rating of garden 1's fertilizer, from 0 (not effective at all) to 100 (perfectly effective)).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Alloy, L. B., & Abramson, L. Y. (1979). Judgment of contingency in depressed and nondepressed students: Sadder but wiser? Journal of Experimental Psychology: General, 108(4), 441-485. https://doi.org/10.1037/0096-3445.108.4.441Nondepressed students overestimated their control over a light when uncontrolled outcomes were frequent.
- Allan, L. G. (1980). A note on measurement of contingency between two binary variables in judgment tasks. Bulletin of the Psychonomic Society, 15(3), 147-149. https://doi.org/10.3758/BF03334492Delta P: the chance of the outcome with the cause minus its chance without it.
- Blanco, F., Matute, H., & Vadillo, M. A. (2013). Interactive effects of the probability of the cue and the probability of the outcome on the overestimation of null contingency. Learning & Behavior, 41(4), 333-340. https://doi.org/10.3758/s13420-013-0108-8Null contingencies overestimated most when both the cause and the outcome are frequent.
- Matute, H., Blanco, F., Yarritu, I., Díaz-Lago, M., Vadillo, M. A., & Barberia, I. (2015). Illusions of causality: How they bias our everyday thinking and how they could be reduced. Frontiers in Psychology, 6, 888. https://doi.org/10.3389/fpsyg.2015.00888The illusion of causality and the outcome and cue density effects.
- Torres, M. N., Barberia, I., & Rodríguez-Ferreiro, J. (2020). Causal illusion as a cognitive basis of pseudoscientific beliefs. British Journal of Psychology, 111(4), 840-852. https://doi.org/10.1111/bjop.12441Garden 1's design: 48 trials with the cause and the outcome each on 75%; the rating went with pseudoscientific beliefs.
- Barreiro, A., Del Carpio, A., Rodríguez-Ferreiro, J., & Barberia, I. (2025). Presentation format influences the strength of causal illusions. Memory & Cognition, 53(8), 2347-2363. https://doi.org/10.3758/s13421-025-01714-zA 32-trial version shown trial by trial, which gave a stronger illusion than a frequency tree.
- Blanco, F., & Matute, H. (2019). Base-rate expectations modulate the causal illusion. PLoS ONE, 14(3), e0212615. https://doi.org/10.1371/journal.pone.0212615Learning an outcome's rate beforehand changed a later null-contingency rating.
Source papers
- Torres MN, Barberia I, Rodríguez-Ferreiro J (2020). Causal illusion as a cognitive basis of pseudoscientific beliefs. British Journal of Psychology, 111(4):840-852.N 225, University students
- No normative data for this task.Paradigm source, N/A
Questions
How long does the Contingency Judgment Task take?
About 6 to 9 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?
Illusion of Causality, Outcome Density Effect, Rare Outcome Rating, Working Fertilizer Rating, Contingency Sensitivity, Prediction Difference, Median Prediction Time, Anticipations, Ratings Made and Planned Ratings. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Contingency Judgment Task used in research?
73 papers from 2021 to 2025 name it in their title or abstract, up 78% compared with 2016 to 2020. That ranks it 252 of 495 test families in the ConductCognition test census (use grade D). Source: ConductCognition test census 0.22.0, counted 3 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Contingency Judgment 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?
Plants per garden (32 plants (shorter), 48 plants (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. 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.