Attention
Flash-Lag Task
Also called: flash-lag effect, flash-lag illusion, flash lag, rotating bar flash-lag, motion extrapolation task
Measure how a moving object is seen ahead of a flash: a bar turns, two lines flash beyond its ends, and people judge whether the lines were ahead of the bar or behind it.
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. Try it here first, with no account.
- Papers, 2021 to 2025
- 36
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
A measure of how a moving object is seen ahead of a flash. A bar turns around a dot, and two short lines flash once beyond its ends, a little ahead of it or behind it. Participants press one key if the lines were ahead of the bar and another if they were behind it. The score is the lag: how far ahead the lines must be to look aligned with the bar, in ms of the bar's turning.
Constructs
- Perceived position of moving objects
- Visual timing
- Motion extrapolation
Research uses
- Visual perception and motion research
- Individual differences in perception
- Within-study group comparison
How the paradigm works
The Flash-Lag Task measures how a moving object is seen ahead of a brief flash at the same place (Nijhawan, 1994). A bar turns about the center of the screen, and two short lines flash once, one beyond each end of the bar. When the lines are exactly aligned with the bar, most people see them trailing behind it. Accounts differ: the brain may extrapolate the moving bar forward, process moving objects with a shorter delay, or read the bar's position a little after the flash (Whitney and Murakami, 1998; Brenner and Smeets, 2000; reviewed by Hogendoorn, 2020), and the size of the effect differs from person to person (Cottier and colleagues, 2023).
More about the paradigm
On each trial a red dot shows for 500 ms, then the bar appears and turns at 180 degrees a second, clockwise on half the trials, for 800 to 1,600 ms before the flash, which lasts about 17 ms, one frame of a 60 Hz screen. The bar keeps turning for 500 ms from the start of the flash, then disappears, and the participant presses J if the lines were ahead of the bar or F if they were behind it, with no time limit. The lines are placed at one of nine offsets from 60 ms behind the bar to 180 ms ahead of it, in steps of 30 ms of the bar's turning (5.4 degrees), 12 times each (standard; 8 to 20), with 8 catch trials 250 ms ahead or behind: 116 trials in a seeded order with a rest halfway, after 6 practice trials with feedback, about 7 to 9 minutes. Every display is set before the session: no answer changes a later trial, and there is no staircase or training mode. Desktop and laptop only.
The main score is the lag: the offset at which the lines are called ahead as often as behind, from a cumulative Gaussian fitted to the answers (Linares and colleagues, 2007), in ms of the bar's turning; positive when the lines had to lead the bar to look aligned. The same in degrees, the spread of the answers, the share of "ahead" answers at each offset and the share of "behind" answers when the lines were aligned are also reported. Scores support research and do not diagnose.
Scores
- Flash-Lagms
- The offset at which the flashed lines are called ahead of the bar as often as behind it, from a cumulative Gaussian fitted to the answers, in ms of the bar's turning; positive when the lines had to lead the bar to look aligned.
- Positive values are the usual finding: aligned lines look behind the bar
- Flash-Lag in Degreesdegrees
- The lag in degrees of the bar's turning, at 180 degrees a second.
- Informational
- Spread of Answersms
- Half the distance between the offsets at which the lines are called ahead on 25% and on 75% of trials, from the fitted curve.
- Smaller means more consistent judgments
- Lapse Rate%
- Fitted share of answers that did not depend on the offset.
- Informational
- Fit%
- Share of the variance of the answers across offsets that the fitted curve explains.
- Informational
- Ahead Answers by Offset%
- Share of "ahead" answers at each offset, keyed by its signed ms, positive ahead of the bar.
- Rises with the offset; it crosses 0.5 at the lag
- Behind Answers When Aligned%
- Share of "behind" answers when the lines were aligned with the bar.
- Above 0.5 shows the effect in its classic form
- Catch Trial Accuracy%
- Share of the catch trials, with the lines 250 ms ahead of the bar or behind it, answered right, when at least 4 were answered.
- Below 0.5 is invalid and below 0.75 is flagged for review
- Median Answer Timems
- Median time from the bar's disappearance to the answer.
- Informational; the answer has no time limit
- Anticipationscount
- Answers under 100 ms from the bar's disappearance, which are not scored.
- Lower is better
- Early Key Pressescount
- F or J pressed before the bar disappeared, 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
- Dropped Frames%
- Share of the frames the browser skipped while the bar turned.
- Above 0.05 is flagged for review
- Flash Durationms
- Median measured time the flashed lines were on screen.
- About 16.7 on a 60 Hz screen
- 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
36 papers from 2021 to 2025 name this task in their title or abstract, down 3% compared with 2016 to 2020. That ranks it 331 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 6 |
| 2024 | 9 |
| 2023 | 5 |
| 2022 | 9 |
| 2021 | 7 |
| 2020 | 4 |
| 2019 | 11 |
| 2018 | 11 |
| 2017 | 8 |
| 2016 | 3 |
| 2015 | 7 |
| 2014 | 10 |
| 2013 | 16 |
| 2012 | 11 |
| 2011 | 11 |
| 2010 | 50 |
| 2009 | 12 |
| 2008 | 25 |
| 2007 | 17 |
| 2006 | 16 |
| 2005 | 7 |
| 2004 | 19 |
| 2003 | 13 |
| 2002 | 19 |
| 2001 | 10 |
| 2000 | 9 |
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
- 9
- 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 2 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Display
- On a dark panel: a red dot for 500 ms, then a white bar of 200 by 8 px turning about the dot at 180 degrees a second, clockwise on half the trials, for 800 to 1,600 ms before the flash.
- Flash
- Two lines of 60 by 8 px, one beyond each end of the bar with a 10 px gap, for 17 ms (one frame of a 60 Hz screen), placed by the bar's position on the frame that shows them. The bar keeps turning for 500 ms from the start of the flash, then disappears.
- Offsets
- The lines 60 or 30 ms behind the bar, aligned with it, or 30, 60, 90, 120, 150 or 180 ms ahead of it, in ms of the bar's turning (30 ms is 5.4 degrees); 8 catch trials 250 ms ahead or behind.
- Answer
- Press J if the lines were ahead of the bar or F if they were behind it, with no time limit. Keys pressed before the bar disappears are counted and not taken as answers.
- Trials
- 12 at each of the 9 offsets (standard; 8 to 20) and 8 catch trials: 116 trials in a seeded order, with a rest halfway.
- Practice
- 6 trials with the lines 250 ms ahead or behind, with feedback for 1,000 ms.
- Fixed displays
- Every offset, speed and display time is set before the session, and no answer changes a later trial: there is no staircase and no training mode.
- Score
- The offset at which the lines are called ahead as often as behind, from a cumulative Gaussian fitted to the answers, in ms; positive when the lines had to lead the bar to look aligned. Answers under 100 ms leave the score.
The display follows Nijhawan's rotating rod, drawn in screen pixels with no calibration; the offsets are times and angles of the bar's turning, so they do not depend on the screen. The flash lasts one frame of a 60 Hz screen and two of a 120 Hz screen.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: flash_lag_test from the scored run, J when the flashed lines were ahead of the turning bar or F when they were behind it, with no time limit. Most offsets are near the point where the lines look aligned, so no answer there is right or wrong for chance and accuracy is not judged. Answer times run from the bar's disappearance; responses under 100 ms follow the timed-test rule. Under half the catch trials (the lines 250 ms ahead or behind, at least 4 answered) answered right invalidates, and under three quarters is flagged. No fitted lag, a run short of its planned trials, or more than 5% of motion frames dropped is flagged for review.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. An answer under 100 ms from the bar's disappearance was pressed before the display could be judged. On the catch trials the lines are 45 degrees along the turn, far beyond any flash-lag, so wrong answers there mean the keys did not follow the display. The motion and the flash are timed in screen frames, so a display that drops frames changed what was shown, and a hidden tab withholds the percentile.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Trials at each offset | 12 | 8 to 20 |
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 Flash-Lag Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Nijhawan, 1994; Hubbard, 2014), with the standard settings (trials at each offset: 12). The main score was flash-lag (the offset at which the flashed lines are called ahead of the bar as often as behind it, from a cumulative Gaussian fitted to the answers, in ms of the bar's turning; positive when the lines had to lead the bar to look aligned).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Nijhawan, R. (1994). Motion extrapolation in catching. Nature, 370(6487), 256-257. https://doi.org/10.1038/370256b0The paradigm: a rotating rod whose outer segments are flashed and look as if they trail the moving middle.
- Whitney, D., & Murakami, I. (1998). Latency difference, not spatial extrapolation. Nature Neuroscience, 1(8), 656-657. https://doi.org/10.1038/3659A flash had to lead a moving bar by about 45 ms to look aligned with it, read as a shorter delay for moving objects.
- Brenner, E., & Smeets, J. B. J. (2000). Motion extrapolation is not responsible for the flash-lag effect. Vision Research, 40(13), 1645-1648. https://doi.org/10.1016/S0042-6989(00)00067-5Dots turning about the fixation point and a bar flashed for one frame; the errors matched reading the dots' position about 60 ms after the flash.
- Linares, D., López-Moliner, J., & Johnston, A. (2007). Motion signal and the perceived positions of moving objects. Journal of Vision, 7(7), 1. https://doi.org/10.1167/7.7.1Flashed dots set ahead of or behind turning dots by the method of constant stimuli, and a cumulative Gaussian whose center measures the effect.
- Cottier, T. V., Turner, W., Holcombe, A. O., & Hogendoorn, H. (2023). Exploring the extent to which shared mechanisms contribute to motion-position illusions. Journal of Vision, 23(10), 8. https://doi.org/10.1167/jov.23.10.8A rod turning about the fixation point at 180 degrees a second, a rod flashed beyond its end, and the judgment of which was further along the turn.
- Hogendoorn, H. (2020). Motion extrapolation in visual processing: Lessons from 25 years of flash-lag debate. Journal of Neuroscience, 40(30), 5698-5705. https://doi.org/10.1523/JNEUROSCI.0275-20.2020
Source papers
- Nijhawan R (1994). Motion extrapolation in catching. Nature, 370(6487):256-257.Paradigm source, Adults
- Hubbard TL (2014). The flash-lag effect and related mislocalizations: Findings, properties, and theories. Psychological Bulletin, 140(1):308-338.Paradigm source, Adults
Questions
How long does the Flash-Lag 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?
Flash-Lag, Flash-Lag in Degrees, Spread of Answers, Lapse Rate, Fit, Ahead Answers by Offset, Behind Answers When Aligned, Catch Trial Accuracy, Median Answer Time, Anticipations, Early Key Presses, Frame Interval, Dropped Frames, Flash Duration, Completed Trials and Planned Trials. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Flash-Lag Task used in research?
36 papers from 2021 to 2025 name it in their title or abstract, down 3% compared with 2016 to 2020. That ranks it 331 of 495 test families in the ConductCognition test census (use grade D). Source: ConductCognition test census 0.22.0, counted 2 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Flash-Lag 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?
Trials at each offset (8 to 20). 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.