Social Cognition
Attitude Attribution Task
Also called: attitude attribution paradigm, correspondence bias task, fundamental attribution error task, essay paradigm, Jones and Harris paradigm
Measure the correspondence bias: participants read short essays on everyday questions, written by choice or argued on an assigned side, and mark what each writer really thinks.
7-10 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
- 102
- Evidence grade
- C
- Registered clinical studies
- 0
What it measures
Measures the correspondence bias. Participants read 12 short essays on everyday questions, each by an imagined student who either chose the side it argues or was told to argue it after a coin toss, and mark what the student really thinks. The main score is how much attitude people read into the essays whose side was assigned.
Constructs
- Correspondence bias
- Attitude attribution
- Person perception
- Use of situational information
Research uses
- Social cognition research
- Person perception and attribution research
- Cross-cultural comparison of attribution
- Within-study group comparison
How the paradigm works
The attitude attribution task measures the correspondence bias: reading an attitude into what a person says even when the situation decided it. Jones and Harris (1967) had students read an essay on Castro's Cuba written either by free choice or by assignment and estimate the writer's true attitude. With choice, the pro-Castro writer was judged far more pro-Castro than the anti-Castro writer (59.62 against 17.38 on a scale from 10 to 70); with assignment the gap shrank but stayed large (44.10 against 22.87). In Many Labs 2 (Klein and colleagues, 2018), 7,197 people who read an assigned essay on capital punishment still judged the writer's attitude by the side the essay took. Here each of 12 everyday questions, such as which season is better, has one essay for each side, written for this task, of about 45 to 55 words, on harmless topics with no health, political, religious or group content. The writers are imagined students in a writing class. Before each essay, a line says that the student chose the side, or was told to argue it after a coin toss picked it. After the essay, the participant marks what the student really thinks on a bar from one side's end to the other's, with no numbers and no starting point; the middle means no preference. Each person sees each question once, 6 essays chosen and 6 assigned, with each side and each end of the bar balanced within each, drawn from the session's seed, and never more than 3 of one kind in a row. The essays are written in each of the eight languages.
More about the paradigm
A practice with the bar on the participant's own view of a day off and one check question on who picks the side, with feedback, come first. There is no time limit.
The main score is the attitude read into assigned essays: for each assigned essay, how far the mark lies toward the essay's side, from -1 (the other side's end) through 0 (the middle) to 1 (the essay's end), averaged. The same score for chosen essays and the difference between the two show how much the assignment changed the judgment. Each question's two sides are balanced across sessions, so a question's own pull cancels in a group. Each person here reads both kinds of essay; most studies compare groups, and seeing both kinds may make the assignment easier to notice. Scopelliti and colleagues (2018) measured the bias within individuals with a scale of their own. Whether the classic results are an error is debated: Walker, Smith and Vul (2022) argue that they fit unbiased inference in an uncertain world. Scores describe judgments in this task; they are research measures and do not diagnose.
Scores
- Attitude Read Into Assigned Essaysscore
- Over the essays whose side was assigned by a coin toss, how far the marks lie toward the essay's side: -1 at the other side's end, 0 in the middle, 1 at the essay's end. Needs at least 2 rated essays.
- Positive means an attitude was read into essays whose side was assigned, the correspondence bias; 0 means none was
- Attitude Read Into Chosen Essaysscore
- The same over the essays whose side the student chose. Needs at least 2 rated essays.
- Positive expected: a chosen side says something about the writer
- Chosen Minus Assignedscore
- Attitude read into chosen essays minus attitude read into assigned essays.
- Positive means the assignment lowered the attitude read into an essay
- Check Correct0/1
- 1 when the check question on who picks the side was answered correctly.
- 1 supports interpretation
- Median Reading Timems
- Median time each essay was on screen before the participant moved on to the bar.
- Informational
- Readings Under 3 Secondscount
- Essays left in under 3 seconds, too fast to read.
- 0 expected
- Median Rating Timems
- Median time from the bar's screen appearing to the mark being confirmed, over ratings at or after 100 ms.
- Informational
- Ratings Under 100 mscount
- Ratings confirmed less than 100 ms after the screen appeared.
- 0 expected
- Essayscount
- Essays rated, out of 12.
- Informational
Use and evidence
102 papers from 2021 to 2025 name this task in their title or abstract, down 32% compared with 2016 to 2020. That ranks it 210 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 18 |
| 2024 | 25 |
| 2023 | 13 |
| 2022 | 25 |
| 2021 | 21 |
| 2020 | 44 |
| 2019 | 31 |
| 2018 | 20 |
| 2017 | 28 |
| 2016 | 27 |
| 2015 | 24 |
| 2014 | 34 |
| 2013 | 27 |
| 2012 | 24 |
| 2011 | 20 |
| 2010 | 22 |
| 2009 | 19 |
| 2008 | 9 |
| 2007 | 18 |
| 2006 | 18 |
| 2005 | 16 |
| 2004 | 17 |
| 2003 | 20 |
| 2002 | 15 |
| 2001 | 15 |
| 2000 | 6 |
Use grade C
100 to 299 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
- 0
- Papers on reliability or validity
- 34
- Registered clinical studies using it as an outcome
- 0
Note on the counts: many papers discuss the bias and run no task.
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.
- Essays
- 24 written for this task: 12 everyday questions, such as which season is better, each with one essay for each side, of about 45 to 55 words. No health, political, religious or group topics. They are shown in the session's language.
- Writers
- Imagined students in a writing class. A line above each essay says the student chose the side, or was told to argue it after a coin toss picked it.
- Rating
- After each essay, a bar from one side's end to the other's, with no numbers and no starting point; the middle means no preference. Which end names the essay's side changes from screen to screen.
- Order
- Each question once: 6 essays chosen and 6 assigned, with 3 on each side and 3 at each end of the bar in each, drawn from the session's seed, and never more than 3 of one kind in a row.
- Practice
- One practice with the bar on the participant's own view of a sunny or a rainy day off, then one question, with feedback, on who picks the side when a coin toss does.
- Scores
- The attitude read into assigned essays: how far toward the essay's side the marks lie, from -1 to 1, over the assigned essays. Also the same for chosen essays and the difference between the two.
Jones and Harris (1967) gave each condition to a different group, and Many Labs 2 replicated the effect with 7,197 people reading assigned essays (Klein and colleagues, 2018). Here each person reads both kinds of essay, on everyday questions, so the comparison is within the person, as Scopelliti and colleagues (2018) measured the bias in individuals; seeing both kinds may make the assignment easier to notice. Whether reading attitudes into assigned essays is an error is debated (Walker, Smith and Vul, 2022). 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: the scored run's marks on the bar, one per essay. No mark is correct, so accuracy is not judged against chance. A run with fewer than half its planned essays is invalid and any other shortfall is flagged; essays left in under 3 seconds on at least half the screens, the same mark on every essay and a wrong check answer are flagged. Ratings under 200 ms on at least half the essays invalidate.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. The scores describe how far the marks follow the essays' sides when the side was chosen and when it was assigned, so any mark is a possible answer, including the middle throughout. An essay of about 50 words cannot be read in under 3 seconds, so many such readings suggest the essays were skipped, and one mark for every essay suggests the bar was not used; both are flagged for review because they can also be real answers. 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
This test runs with fixed settings.
Need a different version of this test?
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 Attitude Attribution Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Jones EE, Harris VA, 1967). The main score was attitude read into assigned essays (over the essays whose side was assigned by a coin toss, how far the marks lie toward the essay's side: -1 at the other side's end, 0 in the middle, 1 at the essay's end).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Jones, E. E., & Harris, V. A. (1967). The attribution of attitudes. Journal of Experimental Social Psychology, 3(1), 1-24. https://doi.org/10.1016/0022-1031(67)90034-0
- Miyamoto, Y., & Kitayama, S. (2002). Cultural variation in correspondence bias: The critical role of attitude diagnosticity of socially constrained behavior. Journal of Personality and Social Psychology, 83(5), 1239-1248. https://doi.org/10.1037/0022-3514.83.5.1239Assigned essays read in Japan and the United States.
- Klein, R. A., Vianello, M., Hasselman, F., et al. (2018). Many Labs 2: Investigating variation in replicability across samples and settings. Advances in Methods and Practices in Psychological Science, 1(4), 443-490. https://doi.org/10.1177/2515245918810225The correspondence bias replicated across samples and settings.
- Scopelliti, I., Min, H. L., McCormick, E., Kassam, K. S., & Morewedge, C. K. (2018). Individual differences in correspondence bias: Measurement, consequences, and correction of biased interpersonal attributions. Management Science, 64(4), 1879-1910. https://doi.org/10.1287/mnsc.2016.2668Measuring the bias within each person.
- Walker, D., Smith, K. A., & Vul, E. (2022). Reconsidering the "bias" in "the correspondence bias". Decision, 9(3), 263-284. https://doi.org/10.1037/dec0000180Whether the classic results are an error.
Source papers
- Many Labs 2 (Klein and colleagues, 2018): across 7,197 participants, writers assigned to argue for capital punishment were judged more in favor of it than writers assigned to argue against it (10.98 against 4.45 on a scale from 1 to 15; d = 1.82).Paradigm source, Adults
Questions
How long does the Attitude Attribution Task take?
About 7 to 10 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?
Attitude Read Into Assigned Essays, Attitude Read Into Chosen Essays, Chosen Minus Assigned, Check Correct, Median Reading Time, Readings Under 3 Seconds, Median Rating Time, Ratings Under 100 ms and Essays. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Attitude Attribution Task used in research?
102 papers from 2021 to 2025 name it in their title or abstract, down 32% compared with 2016 to 2020. That ranks it 210 of 495 test families in the ConductCognition test census (use grade C). Source: ConductCognition test census 0.22.0, counted 3 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Attitude Attribution 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 C in the ConductCognition test census).
Does it report a percentile?
No. Scores and quality flags only.
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.