Attention
Event Segmentation Task
Also called: event segmentation paradigm, unitization task, Newtson unitization, event boundary task, segmentation of everyday activity
Record where people see one action end and the next begin while they watch films of everyday activity, at the smallest and the largest units.
8-10 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
- 262
- Evidence grade
- C
- Registered clinical studies
- 0
What it measures
Participants watch short films of hands doing everyday tasks and press a key each time one meaningful unit of activity ends and another begins, once marking the smallest units and once the largest. Records the number and film time of every boundary.
Constructs
- Event perception
- Event segmentation
- Hierarchical structure of activity
- Attention to ongoing activity
Research uses
- Research on memory for everyday activity
- Research on aging and event perception
- Within-study comparison of segmentation agreement
How the paradigm works
The Event Segmentation Task records how people divide ongoing activity into meaningful units. Newtson (1973) had observers press a button whenever, in their judgment, one meaningful action ended and another began, and Zacks, Tversky and Iyer (2001) asked for the smallest units that seem natural and meaningful on one viewing and the largest on another. In their films, single takes of everyday activities such as making a bed, coarse units averaged 34.3 s and fine units 12.8 s, and coarse boundaries fell closer to fine ones than chance would place them: people see activity as nested parts within larger parts.
More about the paradigm
This version shows three short films of hands doing everyday tasks, each in one continuous shot with no sound: making a spring roll (60 s), folding a paper boat (30 s) and finishing a stew (64 s), all CC0 films from Pixabay. In the standard design the person watches all three twice, once marking the smallest units and once the largest, in an order drawn from the session's seed. Each pass opens with a 30 s practice film of a paper plane being folded; a practice with too few boundaries, or with no more fine than coarse ones, is followed by the note that people usually identify more (or fewer) units and one more viewing, as in Zacks and colleagues (2006). The person presses the space bar at each boundary, or taps one large button on a tablet. Each press is saved in the film's own time, and the films cannot be paused or skipped.
The scores of one session are the boundaries at each grain, the mean unit length (a film's length over its boundaries, as in Zacks and colleagues 2006), and two measures of how coarse boundaries sit against fine ones: alignment, the mean distance from each coarse boundary to the nearest fine one against what chance would give, and enclosure, the share of coarse boundaries that come at or after their nearest fine one. The measure most linked to memory, segmentation agreement, compares each person's boundaries with the rest of the group: films are cut into 1 s bins, the group's share of people with a boundary in each bin is correlated with the person's own bins, and the correlation is scaled by the lowest and highest values possible for that person's number of boundaries (Kurby and Zacks 2011; Sasmita and Swallow 2023). In Sargent and colleagues (2013), agreement correlated .48 with recall of the films. Agreement needs a group, so it is computed across a study's own participants from the saved press times, never against a published norm; Sasmita and Swallow found that it separates real segmentation from chance with 6 to 8 people and steadies by about 18. The films here are short, so one person's counts are small, and the scores support research and do not diagnose.
Scores
- Fine Boundariescount
- Boundaries pressed over the three test films while marking the smallest units.
- Informational: more boundaries means smaller units
- Coarse Boundariescount
- Boundaries pressed over the three test films while marking the largest units.
- Informational; fewer than the fine boundaries when the instructions were followed
- Fine Unit Lengthseconds
- Each film's length over its fine boundaries, averaged over the films with at least one.
- Informational
- Coarse Unit Lengthseconds
- Each film's length over its coarse boundaries, averaged over the films with at least one.
- Informational; longer than the fine units when the instructions were followed
- Hierarchical Alignmentscore
- How much closer coarse boundaries fall to the nearest fine boundary than chance placement would put them, as chance distance minus observed distance over chance distance, over the films watched at both grains.
- 0 is chance and 1 is every coarse boundary on a fine one; higher means nested units
- Enclosure%
- Share of coarse boundaries that come at or after their nearest fine boundary in the same film.
- Above 0.5 means coarse units tend to close after a fine unit ends
- Films Watchedcount
- Test films that played to their end, over both passes.
- Fewer than planned invalidates
- Films With No Boundarycount
- Test films watched with no boundary pressed.
- Any is flagged for review
- Double Pressescount
- Presses within 100 ms of the last press counted, left out of every count.
- Lower is better
- Time With the Page Hiddenseconds
- Seconds the page was out of view during the films; each film paused meanwhile.
- Lower is better
Use and evidence
262 papers from 2021 to 2025 name this task in their title or abstract, up 96% compared with 2016 to 2020. That ranks it 105 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 84 |
| 2024 | 60 |
| 2023 | 46 |
| 2022 | 36 |
| 2021 | 36 |
| 2020 | 34 |
| 2019 | 28 |
| 2018 | 25 |
| 2017 | 23 |
| 2016 | 24 |
| 2015 | 9 |
| 2014 | 16 |
| 2013 | 19 |
| 2012 | 22 |
| 2011 | 19 |
| 2010 | 20 |
| 2009 | 8 |
| 2008 | 10 |
| 2007 | 7 |
| 2006 | 3 |
| 2005 | 0 |
| 2004 | 3 |
| 2003 | 2 |
| 2002 | 0 |
| 2001 | 1 |
| 2000 | 0 |
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
- 30
- 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.
- Films
- Three films of hands at everyday tasks, each one continuous shot with no sound: making a spring roll (60 s), folding a paper boat (30 s) and finishing a stew (64 s). They are CC0 films from Pixabay, shown at up to 640 by 360, in the same order every session.
- Passes
- Each film twice: once marking the smallest units of activity that seem natural and meaningful, once the largest. The order of the two passes comes from the session's seed, fine first in half the sessions; a setting fixes the order or keeps one grain.
- Practice
- Before each pass, a 30 s practice film of hands folding a paper plane. Fewer than 4 boundaries at the fine grain or 1 at the coarse grain, or no more fine than coarse boundaries, brings the note that people usually identify more (or fewer) units, and one more viewing.
- Press
- The space bar, or one large button on a tablet, at each boundary. Each press is saved in the film's own time. A film cannot be paused or skipped, and it pauses while the page is hidden.
- Scores
- Boundaries and mean unit length at each grain, and how coarse boundaries sit against fine ones: hierarchical alignment and enclosure. Segmentation agreement is computed across a study's own participants from the saved press times, in 1 s bins.
The films are short, 154 s per pass where published studies used films of several minutes, so one person's counts are small. Agreement compares a person with the rest of a study's own participants; it needs at least 6 to 8 people per film and grain and steadies by about 18 (Sasmita and Swallow 2023). The spring roll film ends with a dissolve to the finished rolls, and an edit in a film can draw boundaries of its own at the fine grain (Magliano and Zacks 2011). 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: event_segmentation_film from the scored run, one row per film and grain with the film time of each press. No boundary is right or wrong, accuracy is not judged against chance, and any number of presses is accepted. A film not watched to its end, or a pass with no boundary in any film, invalidates. A film with no boundary, no more boundaries at the smallest units than at the largest, more than one press a second in a film and the page hidden for at least a tenth of the films are each flagged for review.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. People differ in where and how often they mark boundaries, and that is what the test records, so the rule judges only whether the films ran and the presses were given, never what they say about the person. Participants who pressed no boundary in a film failed to follow the instructions in Zacks and colleagues (2006), and those who marked no more fine than coarse units were left out by Kurby and Zacks (2011). A film pauses while the page is hidden, so its press times keep their place in the film, but a long hidden spell breaks the viewing, so a percentile is withheld after one.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Grains and their order | Fine and coarse, order from the session's seed (standard) | Fine and coarse, order from the session's seed (standard), Fine, then coarse, Coarse, then fine, Fine only, Coarse only |
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 and tablets. 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 Event Segmentation Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Newtson, 1973; Zacks, Tversky & Iyer, 2001), with the standard settings (grains and their order: Fine and coarse, order from the session's seed). The main score was fine boundaries (boundaries pressed over the three test films while marking the smallest units).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Newtson, D. (1973). Attribution and the unit of perception of ongoing behavior. Journal of Personality and Social Psychology, 28(1), 28-38. https://doi.org/10.1037/h0035584The original unitization procedure, as Zacks, Tversky and Iyer (2001) describe it.
- Zacks, J. M., Tversky, B., & Iyer, G. (2001). Perceiving, remembering, and communicating structure in events. Journal of Experimental Psychology: General, 130(1), 29-58. https://doi.org/10.1037/0096-3445.130.1.29The fine and coarse instructions, single-take films of everyday activity, and coarse boundaries aligned with fine ones.
- Zacks, J. M., Speer, N. K., Vettel, J. M., & Jacoby, L. L. (2006). Event understanding and memory in healthy aging and dementia of the Alzheimer type. Psychology and Aging, 21(3), 466-482. https://doi.org/10.1037/0882-7974.21.3.466Practice with a target number of units, unit length as film length over boundaries, and agreement in 1 s bins.
- Kurby, C. A., & Zacks, J. M. (2011). Age differences in the perception of hierarchical structure in events. Memory & Cognition, 39(1), 75-91. https://doi.org/10.3758/s13421-010-0027-2Agreement scaled by the lowest and highest values possible for a person's number of boundaries, and hierarchical alignment in younger and older adults.
- Sargent, J. Q., Zacks, J. M., Hambrick, D. Z., Zacks, R. T., Kurby, C. A., Bailey, H. R., Eisenberg, M. L., & Beck, T. M. (2013). Event segmentation ability uniquely predicts event memory. Cognition, 129(2), 241-255. https://doi.org/10.1016/j.cognition.2013.07.002Agreement correlated .48 with recall of the films; enclosure.
- Sasmita, K., & Swallow, K. M. (2023). Measuring event segmentation: An investigation into the stability of event boundary agreement across groups. Behavior Research Methods, 55(1), 428-447. https://doi.org/10.3758/s13428-022-01832-5Agreement against the rest of the group, and how many participants it needs.
- Hard, B. M., Recchia, G., & Tversky, B. (2011). The shape of action. Journal of Experimental Psychology: General, 140(4), 586-604. https://doi.org/10.1037/a0024310Alignment as the chance distance minus the observed distance, over the chance distance.
- Magliano, J. P., & Zacks, J. M. (2011). The impact of continuity editing in narrative film on event segmentation. Cognitive Science, 35(8), 1489-1517. https://doi.org/10.1111/j.1551-6709.2011.01202.xFilm edits and fine boundaries.
Source papers
- Zacks JM, Tversky B, Iyer G (2001). Perceiving, remembering, and communicating structure in events. Journal of Experimental Psychology: General, 130(1):29-58.Paradigm source
- Sasmita K, Swallow KM (2023). Measuring event segmentation: An investigation into the stability of event boundary agreement across groups. Behavior Research Methods, 55(1):428-447.Paradigm source
Questions
How long does the Event Segmentation Task take?
About 8 to 10 minutes, as listed in the test library.
What device does a participant need?
Untimed test: runs on desktops, laptops and tablets. The browser must be Chrome, Edge, Firefox or Safari.
What scores does it produce?
Fine Boundaries, Coarse Boundaries, Fine Unit Length, Coarse Unit Length, Hierarchical Alignment, Enclosure, Films Watched, Films With No Boundary, Double Presses and Time With the Page Hidden. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Event Segmentation Task used in research?
262 papers from 2021 to 2025 name it in their title or abstract, up 96% compared with 2016 to 2020. That ranks it 105 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 Event Segmentation 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 settings can a researcher change?
Grains and their order (Fine and coarse, order from the session's seed (standard), Fine, then coarse, Coarse, then fine, Fine only, Coarse only). 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.