Processing Speed
Kappa Effect Task
Also called: tau effect, kappa and tau effects, three-flash task, space-time interdependence, spatiotemporal illusion, perceptual time dilation
Measure how far the spacing of three flashes pulls judged time, the kappa effect, read across a group, with a setting for the tau effect.
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
- 22
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
Measures how far spacing pulls judged time, the kappa effect. Three dots flash one after another along a row, and participants say whether the middle one came closer in time to the first or to the last, while its place is near one end. The shift is read across a group, not for one person.
Constructs
- Time perception
- Space-time interaction
- Perceptual integration
Research uses
- Time perception research, at the group level
- Research on how space and time are combined in perception
- Within-study group comparison
How the paradigm works
The kappa effect task measures how far the spacing of flashes pulls judged time, the kappa effect (Cohen, Hansel and Sylvester, 1953). Three dots flash one after another along a row. When the middle dot is near the first in space, it tends to seem closer to the first in time too. This form follows Kirsch (2023): the middle dot comes at 30%, 40%, 50%, 60% or 70% of the 1,200 ms from the first dot to the last, its place is near the first dot (20% of the distance) or near the last (80%), and the participant says whether it came closer in time to the first or to the last.
More about the paradigm
On a dark panel a cross marks the middle for 500 ms and the panel is empty for 500 ms; then three white dots 16 px across flash for 100 ms each, the first and the last 240 px either side of the middle. The dots run left to right on half the trials and right to left on the other half, in every level and spacing, so the side of the screen the middle dot falls on is not tied to an answer key. The participant presses F for the first dot or J for the last, with no time limit. The standard session has 100 trials, 10 for each of the 5 levels with each spacing, in an order drawn from the session's seed, after 8 practice trials with feedback in which the middle dot is halfway in place and clearly near one end in time. Every level is set before the session: there is no staircase. The displays switch on screen frames, and the time each one was on screen is saved, so the intervals shown are measured.
The main score is the kappa shift. In each spacing a cumulative Gaussian fitted to the share of "last" answers by the measured time gives the moment at which "first" and "last" are equally likely; the shift is that moment with the middle dot near the first minus near the last, in ms. Positive values mean the spacing pulled judged time toward it. The share of "last" answers with the middle dot near the last minus near the first gives the same pull without a fit. Only the outer levels, 30% and 70%, have a right answer; nearer the middle the effect itself makes answers go the other way, so none of them is counted as an error. A setting asks instead whether the middle dot was closer in space, with its place at the five levels and its time near the first or the last: the tau effect (Helson and King, 1931), scored as the shift in percent of the distance.
Scores
- Kappa Shiftms
- Moment judged halfway in time with the middle dot near the first dot minus near the last, from a fitted curve in each spacing.
- Positive values mean spacing pulled judged time; read across a group
- Pull of the Other Dimension%
- Share of "last" answers with the middle dot near the last minus near the first in the dimension not judged, over all five levels.
- Positive values mean the effect; read across a group
- Halfway Moment, Near the Firstms
- Time from the first dot at which "first" and "last" are equally likely, with the middle dot near the first dot in space.
- Informational; 600 ms is halfway
- Halfway Moment, Near the Lastms
- The same with the middle dot near the last dot in space.
- Informational; 600 ms is halfway
- Just Noticeable Differencems
- z(0.75) times the fitted SD, the mean over the spacings with a fit.
- Lower means finer timing
- Tau Shift%
- Place judged halfway with the middle dot early in time minus late, in percent of the distance (space setting).
- Positive values mean timing pulled judged distance; read across a group
- Halfway Place, Early%
- Place, in percent of the distance from the first dot, at which "first" and "last" are equally likely with the middle dot early in time (space setting).
- Informational; 50% is halfway
- Halfway Place, Late%
- The same with the middle dot late in time (space setting).
- Informational; 50% is halfway
- Just Noticeable Difference, Place%
- z(0.75) times the fitted SD, in percent of the distance (space setting).
- Lower means finer judgments of place
- Last Answers, Near the First%
- Share of "last" answers at each level with the middle dot near the first in the other dimension, keyed by the level in percent.
- The psychometric function
- Last Answers, Near the Last%
- The same with the middle dot near the last.
- The psychometric function
- Accuracy at the Outer Levels%
- Share right at 30% and 70% of the judged span, both spacings together.
- Below chance invalidates the session
- Median Answer Timems
- Median time from the third dot to the answer.
- Informational; the answer has no time limit
- Anticipationscount
- Answers under 100 ms from the third dot, which leave the score.
- Lower is better
- Keys During the Dotscount
- F or J pressed before the third dot, which do not count as answers.
- Informational
- Frame Intervalms
- Median time between screen frames, about 16.7 ms on a 60 Hz screen.
- Informational
- Intervals on Time%
- Share of answered trials whose two intervals ran within one frame of plan.
- Under 90% is flagged
- 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
22 papers from 2021 to 2025 name this task in their title or abstract, up 47% compared with 2016 to 2020. That ranks it 394 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 6 |
| 2024 | 3 |
| 2023 | 5 |
| 2022 | 3 |
| 2021 | 5 |
| 2020 | 2 |
| 2019 | 4 |
| 2018 | 0 |
| 2017 | 4 |
| 2016 | 5 |
| 2015 | 2 |
| 2014 | 5 |
| 2013 | 1 |
| 2012 | 2 |
| 2011 | 3 |
| 2010 | 1 |
| 2009 | 3 |
| 2008 | 3 |
| 2007 | 5 |
| 2006 | 3 |
| 2005 | 3 |
| 2004 | 1 |
| 2003 | 1 |
| 2002 | 1 |
| 2001 | 0 |
| 2000 | 0 |
Use grade E
Fewer than 30 papers from 2021 to 2025.
Evidence grade D
Little published measurement evidence found.
- Papers on test-retest reliability
- 0
- Papers on reliability or validity
- 6
- 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 3 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Display
- On a dark panel, a cross at the middle for 500 ms, the empty panel for 500 ms, then three white dots 16 px across, each for 100 ms, on the middle row: the first and the last 240 px either side of the middle, 480 px apart.
- Timing
- 1,200 ms from the first dot's onset to the last dot's onset. The middle dot comes at 30%, 40%, 50%, 60% or 70% of that time.
- Spacing
- The middle dot sits near the first dot (20% of the distance) or near the last (80%). The dots run left to right on half the trials and right to left on the other half.
- Answer
- After the third dot, F if the middle dot came closer in time to the first dot, or J if it came closer to the last, with no time limit.
- Trials
- 10 for each level with each spacing (standard; 8, 12 or 16), half in each direction, in an order drawn from the session's seed: 100 trials, with no feedback.
- Practice
- 8 trials with the middle dot halfway in place and at 20% or 80% of the time, with feedback for 1,000 ms.
- Fixed levels
- Every level is set before the session, and no answer changes the next one: there is no staircase. The displays switch on screen frames, so a 100 ms dot is six frames of a 60 Hz screen, and each trial saves how long each display was on the screen.
- Score
- In each spacing, a cumulative Gaussian fitted by maximum likelihood to the share of "last" answers by the measured time gives the moment judged halfway. The kappa shift is that moment with the middle dot near the first minus near the last, in ms. Answers under 100 ms and trials whose intervals ran more than a frame off plan leave the score.
- Space setting
- The middle dot's place at 30% to 70% of the distance and its time at 20% or 80%; participants judge place, and the tau shift is given in percent of the distance.
The kappa effect is read across a group: one person's shift is noisy, and early three-flash studies found it in some observers but not others (Collyer, 1977; Huang and Jones, 1982). The dots appear on whatever screen the participant has, so their distance in degrees varies between people, and the effect grows with the speed the display implies (Kirsch, 2023). Kirsch found no tau effect in this display, so the space setting is for studies of that question. 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: spaced_flash_timing_test from the scored run, F (closer to the first dot) or J (closer to the last) after the third dot, with no time limit, graded from the saved key and plan. The levels nearer the middle are the measure: the spacing makes some of them look the other way, so no answer there is an error. Only the outer levels, 30% and 70% of the judged span, have a right answer (chance 50%; each comes equally often): below chance invalidates and not above chance is flagged. One key on 90% of answers invalidates when the outer levels are not above chance. Answer times run from the third dot; responses under 100 ms follow the timed-test rule. No shift from the fits, a run short of its planned trials, or fewer than 90% of trials with both intervals 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). A middle dot at 30% or 70% of the time lies beyond the shifts the sources report (Kirsch, 2023), so answers there should follow the plan; nearer the middle a "wrong" answer is the effect itself. An answer under 100 ms from the third dot was pressed before it could be judged. The intervals are the design, so trials whose dots ran late leave the score.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Judgment | Time, place varies (standard) | Time, place varies (standard), Place, time varies |
| Trials per level and spacing | 10 (standard) | 8, 10 (standard), 12, 16 |
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 Kappa Effect Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Cohen, Hansel and Sylvester, 1953; Kirsch, 2023), with the standard settings (judgment: Time, place varies; trials per level and spacing: 10). The main score was kappa shift (moment judged halfway in time with the middle dot near the first dot minus near the last, from a fitted curve in each spacing).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Cohen, J., Hansel, C. E. M., & Sylvester, J. D. (1953). A new phenomenon in time judgment. Nature, 172(4385), 901. https://doi.org/10.1038/172901a0The first report of the kappa effect.
- Helson, H., & King, S. M. (1931). The tau effect: An example of psychological relativity. Journal of Experimental Psychology, 14(3), 202-217. https://doi.org/10.1037/h0071164The tau effect, the reverse pull, first shown with touch.
- Collyer, C. E. (1977). Discrimination of spatial and temporal intervals defined by three light flashes: Effects of spacing on temporal judgments and of timing on spatial judgments. Perception & Psychophysics, 21(4), 357-364. https://doi.org/10.3758/BF03199487Three light flashes judged in time and in space, with effects that differed between observers.
- Huang, Y. L., & Jones, B. (1982). On the interdependence of temporal and spatial judgments. Perception & Psychophysics, 32(1), 7-14. https://doi.org/10.3758/BF03204862With longer sequences the kappa effect appeared for the harder judgments near the middle.
- Kirsch, W. (2023). On the interplay between time and space perception in discontinuous stimulus displays. Attention, Perception, & Psychophysics, 85(7), 2406-2421. https://doi.org/10.3758/s13414-023-02678-5The three-dot design online, the shift of the point of subjective equality across spacings, and no tau effect.
- Chen, Y., Zhang, B., & Kording, K. P. (2016). Speed constancy or only slowness: What drives the kappa effect. PLoS ONE, 11(4), e0154013. https://doi.org/10.1371/journal.pone.0154013Judged time between two flashes grew with their distance, a speed-based account of the effect.
Source papers
- Kirsch (2023), Attention, Perception, & Psychophysics, 85(7), 2406-2421: three dots online, the shift of the point of subjective equality across spacings.Paradigm source, Adults
Questions
How long does the Kappa Effect 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?
Kappa Shift, Pull of the Other Dimension, Halfway Moment, Near the First, Halfway Moment, Near the Last, Just Noticeable Difference, Tau Shift, Halfway Place, Early, Halfway Place, Late, Just Noticeable Difference, Place, Last Answers, Near the First, Last Answers, Near the Last, Accuracy at the Outer Levels, Median Answer Time, Anticipations, Keys During the Dots, Frame Interval, Intervals on Time, Completed Trials and Planned Trials. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Kappa Effect Task used in research?
22 papers from 2021 to 2025 name it in their title or abstract, up 47% compared with 2016 to 2020. That ranks it 394 of 495 test families in the ConductCognition test census (use grade E). Source: ConductCognition test census 0.22.0, counted 3 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Kappa Effect 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?
Judgment (Time, place varies (standard), Place, time varies) and trials per level and spacing (8, 10 (standard), 12, 16). 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.