Attention
Contour Integration Task
Also called: contour detection task, path detection, Gabor path paradigm, association field, snake contours, perceptual organization task
Measure contour integration: how far a path of aligned patches can bend before people stop seeing it among patches at random angles.
7-9 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. Preview any test free after you sign up.
- Papers, 2021 to 2025
- 80
- Evidence grade
- C
- Registered clinical studies
- 6
What it measures
Measures how well vision links separate edges into a contour. Each trial shows two fields of small striped patches, one after the other; in one, ten patches line up along a hidden path. Participants press 1 or 2 for the field that held the path. The path bends by one of six fixed angles, from straight to zigzag, and the score is the bend at which 75 percent of answers are right.
Constructs
- Contour integration
- Perceptual grouping
- Visual integration
Research uses
- Schizophrenia research on visual integration
- Amblyopia and strabismus research
- Visual perception research
How the paradigm works
The Contour Integration Task measures how well vision links separate edges into one contour. Each trial shows two fields of small striped patches (Gabor patches), one after the other, each for 1,000 ms with a cross for 200 ms between them. In one field, ten patches line up along a hidden path, each turned along it; every other patch has a random place and orientation. The participant then presses 1 or 2 for the field that held the path, with no time limit.
More about the paradigm
The path bends at each patch by the trial's path angle, 0, 18, 36, 54, 72 or 90 degrees, to the left or right at random, with up to 10 degrees of jitter. The fields follow the construction of Field, Hayes and Hess (1993) as Hansen, May and Hess (2014) describe it: an invisible grid of 15 by 15 cells with one patch in each, the path placed first. The path's patches are as far apart as the field's patches are from their nearest neighbours, so the path is no denser than the field around it and only the alignment of its patches shows it. Each path angle appears 16 times (standard; 10 to 24) with the path in each field on half the trials, for 96 trials in a seeded order with a break halfway, after 6 practice trials whose feedback shows the path circled. Nothing a participant answers changes a path angle: there is no staircase. The fields switch on screen frames, and their measured durations are saved on every trial. Desktop and laptop only; about 7 to 9 minutes.
The main score is the path angle threshold: the bend at which 75 percent of answers are right, from a two-choice Weibull falling with the path angle, fitted by maximum likelihood to the share right at each angle. Higher means paths that bent more were still seen. Accuracy at each path angle, any lean toward the first or the second field and the screen's timing are also reported. Scores support research and do not diagnose.
Scores
- Path Angle Thresholddegrees
- The bend between neighbouring patches on the path at which 75 percent of answers are right, from a two-choice Weibull falling with the angle, fitted by maximum likelihood to the share right at each of the six path angles.
- Higher is better; paths that bent more were still seen
- Psychometric Slopescore
- Slope of the fitted Weibull function.
- Higher means accuracy fell more steeply as the path bent
- Lapse Rate%
- Fitted share of errors on paths straight enough to see.
- Lower means fewer lapses
- Accuracy%
- Share of usable answers that named the field with the path.
- Higher is better; guessing gives 0.5
- Accuracy by Path Angle%
- Share right at each path angle, keyed by the angle in degrees.
- Falls as the path bends more
- First Field Answers%
- Share of usable answers that named the first field.
- 0.5 means no lean toward either field
- Accuracy, Path in the First Field%
- Share right when the path was in the first field.
- Informational
- Accuracy, Path in the Second Field%
- Share right when the path was in the second field.
- Informational
- Median Answer Timems
- Median time from the question to the answer.
- Informational; the answer has no time limit
- Anticipationscount
- Answers given within 100 ms of the question, which leave the score.
- More than max(2, 20%) is invalid
- Early Key Pressescount
- 1 or 2 keys pressed before the question appeared, which are not taken as answers.
- Informational
- Frame Intervalms
- Median time between screen frames during the trials.
- About 16.7 on a 60 Hz screen
- Refresh RateHz
- Screen refresh rate from the frame interval.
- Informational
- Fields on Time%
- Share of answered trials whose two fields each ran within one frame of plan; only these are scored.
- Under 0.9 is flagged for review
- Completed Trialstrials
- Trials answered in the scored run.
- Fewer than planned is flagged for review
- Planned Trialstrials
- Trials the run set out to show.
- Informational
Use and evidence
80 papers from 2021 to 2025 name this task in their title or abstract, about the same as from 2016 to 2020. That ranks it 204 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 25 |
| 2024 | 20 |
| 2023 | 12 |
| 2022 | 14 |
| 2021 | 9 |
| 2020 | 13 |
| 2019 | 19 |
| 2018 | 9 |
| 2017 | 24 |
| 2016 | 15 |
| 2015 | 31 |
| 2014 | 41 |
| 2013 | 43 |
| 2012 | 29 |
| 2011 | 37 |
| 2010 | 63 |
| 2009 | 18 |
| 2008 | 22 |
| 2007 | 11 |
| 2006 | 14 |
| 2005 | 14 |
| 2004 | 19 |
| 2003 | 13 |
| 2002 | 14 |
| 2001 | 10 |
| 2000 | 15 |
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
- 19
- Registered clinical studies using it as an outcome
- 6
Note on the counts: a few counted papers use the words in computational vision.
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 2 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Display
- On a gray panel 480 px square: a cross for 500 ms, a field of about 215 striped patches for 1,000 ms, the cross for 200 ms, then a second field for 1,000 ms. Each patch is a grating of 8 px per cycle in a round window, at a random place, orientation and phase, one in each cell of an invisible 15 by 15 grid.
- Path
- In one of the two fields, 10 patches line up along a path, each turned along it. At each patch the path bends by the trial's path angle, 0, 18, 36, 54, 72 or 90 degrees, to the left or right at random, plus up to 10 degrees of jitter. The path's patches are as far apart as the field's patches are from their nearest neighbours, so only their alignment shows the path.
- Answer
- Press 1 or 2 for the field that held the path, with no time limit. Keys pressed before the question are counted and not taken as answers.
- Trials
- 16 at each path angle (standard; 10 to 24): 96 trials in a seeded order, with the path in each field on half the trials at every angle, and a break after every 48 trials.
- Practice
- 6 trials at the three straightest path angles, each followed for 2,500 ms by the field that held the path, with its patches circled.
- Fixed angles
- Every path angle and display time is set before the session, and no answer changes it: there is no staircase. The fields are shown in whole frames of the screen's own rate.
- Score
- The path angle at which 75 percent of answers are right, from a two-choice Weibull falling with the angle, fitted by maximum likelihood to the share right at each angle. Answers under 100 ms and trials whose fields ran more than one frame off plan leave the score.
Studies build contours in several ways: open paths or closed loops, Gabor patches or lines, and a difficulty set by path angle, spacing or orientation jitter, often with a staircase. This version uses open paths of Gabor patches, six fixed path angles (the method of constant stimuli) and two fields one after the other, so its threshold is compared within this version.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: contour_integration_test from the scored run, 1 or 2 for the field that held the path, after both fields, with no time limit (chance 50%; the path is in each field on half the trials at every angle), graded from the saved key and path field. Below chance invalidates and not above chance is flagged; one key on 90% of answers invalidates when accuracy is not above chance. Answer times run from the question; responses under 100 ms follow the timed-test rule. No fitted path angle, under 75 percent right with straight paths, a run short of its planned trials, or fewer than 90% of trials with both fields within one frame of plan 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. 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 100 ms from the question was pressed before the question could be read. Straight paths are the easiest the test shows, so many errors on them mean the paths could not be seen or the task was not followed. The display time is part of the design, so trials whose fields ran off plan leave the score.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Trials per path angle | 16 | 10 to 24 |
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.
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Contour Integration Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Field, Hayes and Hess, 1993; Kovacs and Julesz, 1993), with the standard settings (trials per path angle: 16). The main score was path angle threshold (the bend between neighbouring patches on the path at which 75 percent of answers are right, from a two-choice Weibull falling with the angle, fitted by maximum likelihood to the share right at each of the six path angles).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Field, D. J., Hayes, A., & Hess, R. F. (1993). Contour integration by the human visual system: Evidence for a local "association field". Vision Research, 33(2), 173-193. https://doi.org/10.1016/0042-6989(93)90156-QThe path paradigm: a jagged path of aligned Gabor patches among patches at random orientations, seen reliably up to path angles of about 60 degrees.
- Kovacs, I., & Julesz, B. (1993). A closed curve is much more than an incomplete one: Effect of closure in figure-ground segmentation. Proceedings of the National Academy of Sciences, 90(16), 7495-7497. https://doi.org/10.1073/pnas.90.16.7495Contours of separate patches found in clutter, with closed contours seen at wider spacings than open ones.
- Hansen, B. C., May, K. A., & Hess, R. F. (2014). One "shape" fits all: The orientation bandwidth of contour integration. Journal of Vision, 14(13), 17. https://doi.org/10.1167/14.13.17The grid construction, the path angle jitter and the two fields of 1,000 ms this version follows.
- Kovacs, I., Polat, U., Pennefather, P. M., Chandna, A., & Norcia, A. M. (2000). A new test of contour integration deficits in patients with a history of disrupted binocular experience during visual development. Vision Research, 40(13), 1775-1783. https://doi.org/10.1016/S0042-6989(00)00008-0Contour integration losses in strabismic amblyopia.
- Silverstein, S. M., Keane, B. P., Barch, D. M., Carter, C. S., Gold, J. M., Kovacs, I., MacDonald, A., Ragland, J. D., & Strauss, M. E. (2012). Optimization and validation of a visual integration test for schizophrenia research. Schizophrenia Bulletin, 38(1), 125-134. https://doi.org/10.1093/schbul/sbr141Contour integration as a measure of visual integration in schizophrenia.
Source papers
- Field DJ, Hayes A, Hess RF (1993). Contour integration by the human visual system: evidence for a local "association field". Vision Research, 33(2):173-193.Paradigm source, Adults
Questions
How long does the Contour Integration Task take?
About 7 to 9 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?
Path Angle Threshold, Psychometric Slope, Lapse Rate, Accuracy, Accuracy by Path Angle, First Field Answers, Accuracy, Path in the First Field, Accuracy, Path in the Second Field, Median Answer Time, Anticipations, Early Key Presses, Frame Interval, Refresh Rate, Fields on Time, Completed Trials and Planned Trials. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Contour Integration Task used in research?
80 papers from 2021 to 2025 name it in their title or abstract, about the same as from 2016 to 2020. That ranks it 204 of 395 test families in the ConductCognition test census (use grade D). Source: ConductCognition test census 0.14.0, counted 2 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Contour Integration Task have?
In the published research, 2 papers report its test-retest reliability and 6 registered clinical studies use it as an outcome (evidence grade C in the ConductCognition test census).
Which settings can a researcher change?
Trials per path angle (10 to 24). 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.