Social Cognition
Intentional or Accidental Task
Also called: intentionality bias task, intentionality bias paradigm, intentional or accidental judgment, intention attribution task
Measure the intentionality bias. Participants judge whether actions in short sentences were done on purpose or by accident, under time pressure and with time to think.
4-6 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
- 43
- Evidence grade
- C
- Registered clinical studies
- 2
What it measures
Measures the intentionality bias, the pull to see actions as done on purpose. Participants press F or J to say whether actions in 64 short sentences, such as "She spilled the orange juice", were done on purpose or by accident, in a part with a time limit and a part without one. The main score is the share of ambiguous actions judged done on purpose under time pressure.
Constructs
- Intentionality bias
- Intention attribution
- Social cognition
- Dual-process reasoning
Research uses
- Social cognition research
- Attribution and moral judgment research
- Research on time pressure and judgment
- Within-study group comparison
How the paradigm works
The intentional or accidental task measures the intentionality bias, the pull to see what people do as done on purpose. Rosset (2008) showed 90 students sentences about actions that can be done either way, such as setting a house on fire, mixed with clearly intentional and clearly accidental ones, and asked whether each was generally done on purpose or by accident. Students who had 2,400 ms for each sentence judged 39% of the ambiguous actions done on purpose, against 33% for those who had 5,000 ms, and 22% against 15% for the actions most often done by accident; she argued that an intentional reading comes first and that judging an act accidental takes extra thought. Moore and Pope (2014) found the bias with videos of one ambiguous finger movement, judged intentional 64% of the time, and Pinkham, Harvey and Penn (2018) used 24 sentences, half with 2.4 seconds and half with 5, in a large validation study of social cognition tests. Here, 64 sentences written for this task in each language, each a person doing one everyday thing, are judged on purpose or by accident with the F and J keys: 32 ambiguous actions, such as "She spilled the orange juice", and 16 clearly intentional and 16 clearly accidental ones. One part gives 2.4 seconds per sentence by default and the other has no time limit; each part has its own 32 sentences, and which part comes first, which set each part takes and which key means on purpose are drawn from the session's seed. A cross comes before each sentence, four practice sentences start each part, and in the fast part a missed sentence brings a short message.
More about the paradigm
The main score is the share of ambiguous actions judged done on purpose in the fast part. The same share in the slow part, their difference, the share over both parts, accuracy on the clear sentences, and in the fast part the median times of on-purpose and by-accident answers to ambiguous sentences come with it. Scores describe judgments of these sentences; they are research measures for comparing groups, and they do not describe a person's character or diagnose.
Scores
- Fast Part: On Purpose%
- The share of ambiguous actions judged done on purpose in the fast part, over answers at or after 100 ms.
- Higher means more actions seen as done on purpose under time pressure, the intentionality bias
- Slow Part: On Purpose%
- The same share in the part with no time limit.
- Informational
- Speed Effectpoints
- The fast part's share minus the slow part's, in percentage points.
- Positive means more on-purpose judgments under time pressure, as Rosset (2008) found
- Both Parts: On Purpose%
- The share of ambiguous actions judged done on purpose over both parts.
- Informational
- Clear Sentences Right%
- The share of clearly intentional and clearly accidental sentences answered as their kind, over both parts.
- Near 100 expected; under 70 invalidates the run
- Fast Part: On-Purpose Timems
- Median time of on-purpose answers to ambiguous sentences in the fast part, from the sentence's onset.
- Informational
- Fast Part: By-Accident Timems
- Median time of by-accident answers to ambiguous sentences in the fast part, from the sentence's onset.
- Informational
- By-Accident Slowingms
- The by-accident time minus the on-purpose time, in the fast part.
- Positive means judging an action accidental took longer, as the account of an intentional default predicts
- Missed Sentencescount
- Sentences of the fast part left unanswered before they went.
- Informational
- Answers Under 100 mscount
- Answers made less than 100 ms after their sentence appeared.
- 0 expected
- Ambiguous Sentences Answeredcount
- Ambiguous sentences answered at or after 100 ms over both parts, out of 32.
- Informational
Use and evidence
43 papers from 2021 to 2025 name this task in their title or abstract, down 26% compared with 2016 to 2020. That ranks it 459 of 670 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 8 |
| 2024 | 7 |
| 2023 | 5 |
| 2022 | 12 |
| 2021 | 11 |
| 2020 | 11 |
| 2019 | 11 |
| 2018 | 17 |
| 2017 | 11 |
| 2016 | 8 |
| 2015 | 11 |
| 2014 | 9 |
| 2013 | 4 |
| 2012 | 7 |
| 2011 | 11 |
| 2010 | 5 |
| 2009 | 10 |
| 2008 | 4 |
| 2007 | 3 |
| 2006 | 4 |
| 2005 | 3 |
| 2004 | 1 |
| 2003 | 7 |
| 2002 | 2 |
| 2001 | 0 |
| 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
- 17
- Registered clinical studies using it as an outcome
- 2
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 0.48.0, counted 12 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Sentences
- 64 written for this task in each language, each a person doing one everyday thing: 32 ambiguous actions that could be done on purpose or by accident, such as "She spilled the orange juice", 16 clearly intentional and 16 clearly accidental. None involves injury, violence, crime or a group of people.
- Answer
- F or J for on purpose or by accident; which key means on purpose is drawn from the session's seed.
- Parts
- A fast part, in which each sentence goes after 2.4 seconds by default (1.5 to 5 as a setting), and a slow part with no time limit. Each part has its own 32 sentences, 16 ambiguous, 8 intentional and 8 accidental, in 8 groups of 4. Which part comes first and which set each part takes are drawn from the session's seed, or the study fixes the order.
- Trial
- A cross for 500 ms, then the sentence on a dark panel with the keys named under it. In the fast part, a missed sentence brings a short message to answer sooner.
- Practice
- 4 clear sentences before each part, with the answer shown back.
- Scores
- The share of ambiguous actions judged done on purpose in the fast part, the main score, and in the slow part; their difference; the share over both parts; accuracy on the clear sentences; and in the fast part the median times of the two answers to ambiguous sentences and their difference.
- Languages
- Sentences and instructions in English, Italian, German, Spanish, French, Japanese, Portuguese and Chinese.
Rosset (2008) compared two groups, one shown each sentence for 2,400 ms and the other for 5,000 ms, who marked their answers on paper; here each person takes both parts, with keys and response times, as the 24-sentence version of Pinkham, Harvey and Penn (2018) also gave each person fast and slow sentences. The sentences are this task's own, so levels are best compared between groups within a study; they describe judgments of these sentences, not 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: intentional_or_accidental_test from the scored run, F or J for on purpose or by accident, within the fast part's time limit (standard 2.4 seconds) or with no limit in the slow part. The 32 clear sentences have a right answer (chance 50%, half of each kind); the 32 ambiguous ones have none and are never counted as errors. Below chance on the clear sentences invalidates and not above chance is flagged; one key on 90% of answers invalidates when the clear sentences are not above chance. Missed responses and responses under 100 ms follow the timed-test rule. Under 70% right on the clear sentences invalidates.
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. An ambiguous sentence can be read either way, so its answer describes a judgment, not accuracy. Rosset (2008) included clearly intentional and clearly accidental sentences to track reading and flag indiscriminate answering; her participants answered 97% of them as their kind. The 70% limit is a ConductScience choice.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Time to answer in the fast part (ms) | 2,400 | 1,500 to 5,000 |
| Part that comes first | Drawn for each session (standard) | Drawn for each session (standard), The fast part, The slow part |
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 Intentional or Accidental Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Rosset E, 2008; Moore JW, Pope A, 2014), with the standard settings (time to answer in the fast part (ms): 2,400; part that comes first: Drawn for each session). The main score was fast Part: on purpose (the share of ambiguous actions judged done on purpose in the fast part, over answers at or after 100 ms).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Rosset, E. (2008). It's no accident: Our bias for intentional explanations. Cognition, 108(3), 771-780. https://doi.org/10.1016/j.cognition.2008.07.001
- Moore, J. W., & Pope, A. (2014). The intentionality bias and schizotypy. Quarterly Journal of Experimental Psychology, 67(11), 2218-2224. https://doi.org/10.1080/17470218.2014.911332The bias with videos of one ambiguous movement instead of sentences.
- Pinkham, A. E., Harvey, P. D., & Penn, D. L. (2018). Social Cognition Psychometric Evaluation: Results of the final validation study. Schizophrenia Bulletin, 44(4), 737-748. https://doi.org/10.1093/schbul/sbx117A 24-sentence version with fast and slow sentences for each person, and its retest reliability.
Source papers
- Pinkham, Harvey and Penn (2018): 154 healthy adults judged 40% of 24 sentences intentional (SD 15); retest r = .51.N 154, Adults
Questions
How long does the Intentional or Accidental Task take?
About 4 to 6 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?
Fast Part: On Purpose, Slow Part: On Purpose, Speed Effect, Both Parts: On Purpose, Clear Sentences Right, Fast Part: On-Purpose Time, Fast Part: By-Accident Time, By-Accident Slowing, Missed Sentences, Answers Under 100 ms and Ambiguous Sentences Answered. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Intentional or Accidental Task used in research?
43 papers from 2021 to 2025 name it in their title or abstract, down 26% compared with 2016 to 2020. That ranks it 459 of 670 test families in the ConductCognition test census (use grade D). Source: ConductCognition test census 0.48.0, counted 12 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Intentional or Accidental Task have?
In the published research, 2 papers report its test-retest reliability and 2 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?
Time to answer in the fast part (ms) (1,500 to 5,000) and part that comes first (Drawn for each session (standard), The fast part, The slow part). 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.