Social Cognition
Indirect Hints Task
Also called: indirect request task, indirect speech comprehension task, intention inference task
Measure how people understand hints: participants read short scenes that end with a hint and choose what the speaker means, with a plainer hint after a miss.
8 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
- 78
- Evidence grade
- B
- Registered clinical studies
- 33
Daniel leaves the front door wide open. Mila says, “My cat is very quick when she sees an open door.”
What it measures
Measures how well people understand a hint: what someone means when they ask for something without saying it. Participants read short scenes that end with a hint and choose what the speaker means from four choices. After a miss, the scene goes on with a plainer hint and the question is asked again. The main score is hint points as a percent of the most possible. The scenes are in English.
Constructs
- Theory of mind
- Understanding intentions
- Indirect speech
- Social cognition
Research uses
- Theory of mind research in adults
- Research on how people understand indirect requests
- Social cognition batteries
- Within-study group comparison
How the paradigm works
The Indirect Hints Task measures how well people understand a hint: what someone means when they ask for something without saying it. It follows the design of Corcoran, Mercer and Frith (1995), with our own name, scenes and choices. Each short scene has two people and ends with one of them dropping a hint, and the participant chooses what the speaker means from four choices: the meaning, the hint taken at face value, and two other wishes that the scene rules out. If the first choice is not the meaning, the scene goes on with a plainer hint and the same question, the original's two steps as Klein and colleagues (2020) describe them, and the original's scoring follows: 2 points for the meaning at the first hint and 1 at the second.
More about the paradigm
A session has 10 scenes by default, the number of items in the original, or 16 as a setting. The scenes are about everyday needs, such as a borrowed ladder, a shared printer or a seat on a train, and were written for this test; no item of the published task is used. The session's seed orders the scenes and their choices, with the meaning equally often in each place, and all of it is saved. The scene stays on screen and there is no time limit. Two practice scenes have feedback.
The main score is hint points as a percent of the most possible. The test also reports the points, the share of scenes understood at the first hint and at the second, and how often a first answer took the hint at face value. The published task was among the tasks with the strongest measurement properties in the Social Cognition Psychometric Evaluation study (Pinkham and colleagues, 2016, 2018), and many adults score at or near its top (Klein and colleagues, 2020); those findings are for its items, not these. A choice among four may bring more scores to the top, so the test is best used to compare groups. The scenes are English; sessions run in other languages are flagged for review. Scores support research and do not diagnose.
Scores
- Hint Score%
- Points as a percent of the most possible: 2 for the meaning at the first hint, 1 at the second, 0 otherwise.
- Higher is better
- Hint Pointspoints
- The points themselves. The 10 scenes of the standard session allow 20.
- Higher is better
- Understood at First Hint%
- Percent of scenes where the meaning was chosen at the first hint.
- Higher is better
- Understood at Second Hint%
- Of the scenes that went on to the plainer hint, the percent where the meaning was chosen there.
- Higher is better
- Face-Value Answers%
- Percent of first answers that took the hint at face value.
- Informational
- Median Time to a Right First Answerms
- Median time from the scene's appearance to a right first answer, reading included.
- Informational
- Quick First Answerscount
- First answers within 3 seconds of the scene appearing, too soon to have read it.
- At least half invalidates
- Scenes Shownitems
- Scenes the session reached.
- Informational
- Scenes Setitems
- Scenes the session set out to show.
- Informational
- Anticipationscount
- Answers under 100 ms.
- Lower supports interpretation
Use and evidence
78 papers from 2021 to 2025 name this task in their title or abstract, down 24% compared with 2016 to 2020. That ranks it 243 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 23 |
| 2024 | 14 |
| 2023 | 11 |
| 2022 | 13 |
| 2021 | 17 |
| 2020 | 20 |
| 2019 | 34 |
| 2018 | 19 |
| 2017 | 16 |
| 2016 | 14 |
| 2015 | 9 |
| 2014 | 13 |
| 2013 | 9 |
| 2012 | 11 |
| 2011 | 2 |
| 2010 | 7 |
| 2009 | 3 |
| 2008 | 7 |
| 2007 | 3 |
| 2006 | 4 |
| 2005 | 3 |
| 2004 | 0 |
| 2003 | 3 |
| 2002 | 1 |
| 2001 | 0 |
| 2000 | 0 |
Use grade D
30 to 99 papers from 2021 to 2025.
Evidence grade B
At least 8 test-retest papers, and either 5 registered clinical studies or 150 reliability or validity papers.
- Papers on test-retest reliability
- 17
- Papers on reliability or validity
- 43
- Registered clinical studies using it as an outcome
- 33
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 31 October 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Scenes
- 10 by default, the number of items in the original task, or 16 as a setting. Each has two people and ends with one of them dropping a hint.
- Choices
- 4 after each scene: the meaning, the hint taken at face value, and two other wishes the scene rules out. The meaning is in each place equally often.
- Second hint
- After a first choice that misses, the scene goes on with a plainer hint and the same question, with the same choices in the same order.
- Scoring
- 2 points for the meaning at the first hint, 1 at the second and 0 otherwise, the original's scoring.
- Time
- No time limit. The scene stays on screen while the participant chooses.
- Order
- The session's seed orders the scenes and their choices, with the meaning in one place no more than 2 times in a row, and all of it is saved.
- Practice
- 2 scenes with feedback. The test has no feedback.
The scenes were written for this test; no published item is used, and the published task's norms and measurement figures do not apply to them. The scenes are English; sessions run in other languages are flagged for review. 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: the first answer to each scene in indirect_hints_answer, four choices with the meaning equally often in each place, so chance is 25%. Accuracy below or not above chance is flagged for review and never invalidates. The same place on at least 90% of first answers invalidates unless accuracy is above chance. Answers under 200 ms on at least half of 4 or more answers invalidate, and so do first answers within 3 seconds of the scene on at least half of 4 or more scenes. Reaching fewer than 75% of the scenes set invalidates. A session run in another language is flagged for review.
Why these rules
Choosing the hint taken at face value, or another wish, on most scenes is a pattern this test can show, so accuracy at or below chance is flagged for review and never invalidates. 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). 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. A test scene and its four choices run to 75 to 96 words, which take most adults well over 3 seconds to read, so a first answer within 3 seconds cannot rest on the scene; when at least half of them come that soon, the answers do not show that the scenes were read. With fewer than three quarters of the scenes reached, the score rests on too few scenes to compare with a full session. The scenes are English; a session in another language shows them to a participant who may not read English fluently, so it is flagged for review.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Scenes | 10 scenes (standard) | 10 scenes (standard), 16 scenes |
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
- Instructions and screens: English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese. Words and sentences: English.
- Available on: Every plan, including the free plan.
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Indirect Hints Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Corcoran R, Mercer G, Frith CD, 1995; Pinkham AE, Penn DL, Green MF, Harvey PD, 2016; Pinkham AE, Harvey PD, Penn DL, 2018; Klein HS and colleagues, 2020), with the standard settings (scenes: 10 scenes). The main score was hint score (points as a percent of the most possible: 2 for the meaning at the first hint, 1 at the second, 0 otherwise).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Corcoran, R., Mercer, G., & Frith, C. D. (1995). Schizophrenia, symptomatology and social inference: Investigating "theory of mind" in people with schizophrenia. Schizophrenia Research, 17(1), 5-13. https://doi.org/10.1016/0920-9964(95)00024-G
- Klein, H. S., Springfield, C. R., Bass, E., Ludwig, K., Penn, D. L., Harvey, P. D., & Pinkham, A. E. (2020). Measuring mentalizing: A comparison of scoring methods for the hinting task. International Journal of Methods in Psychiatric Research, 29(2), e1827. https://doi.org/10.1002/mpr.1827The two steps, the 2, 1 and 0 scoring, and how many adults reach the top score.
- Pinkham, A. E., Penn, D. L., Green, M. F., & Harvey, P. D. (2016). Social cognition psychometric evaluation: Results of the initial psychometric study. Schizophrenia Bulletin, 42(2), 494-504. https://doi.org/10.1093/schbul/sbv056Measurement properties of the published task.
- 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/sbx117Measurement properties of the published task.
Source papers
- Klein HS, Springfield CR, Bass E, Ludwig K, Penn DL, Harvey PD, Pinkham AE (2020). Measuring mentalizing: A comparison of scoring methods for the hinting task. International Journal of Methods in Psychiatric Research, 29(2):e1827.Paradigm source, Adults
Questions
How long does the Indirect Hints Task take?
About 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?
Hint Score, Hint Points, Understood at First Hint, Understood at Second Hint, Face-Value Answers, Median Time to a Right First Answer, Quick First Answers, Scenes Shown, Scenes Set and Anticipations. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Indirect Hints Task used in research?
78 papers from 2021 to 2025 name it in their title or abstract, down 24% compared with 2016 to 2020. That ranks it 243 of 495 test families in the ConductCognition test census (use grade D). Source: ConductCognition test census 0.22.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Indirect Hints Task have?
In the published research, 17 papers report its test-retest reliability and 33 registered clinical studies use it as an outcome (evidence grade B in the ConductCognition test census).
Which settings can a researcher change?
Scenes (10 scenes (standard), 16 scenes). Settings can be changed on yearly plans; free and pay-as-you-go studies use the standard settings.
Which languages can participants use?
The instructions and screens are in English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese. The words and sentences stay in English, so sessions in other languages are flagged for review. 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.