Attention
Scene Glimpse Task
Also called: scene gist, rapid scene categorization, ultra-rapid categorization, rapid visual categorization, gist of a scene, masked scene categorization, scene categorization
Measure how short a glimpse of a photograph is enough to say what kind of scene it was, from flashes of about 17 ms up to 133 ms, each followed by a mask.
5-7 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
- 32
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
Measures how short a glimpse of a photograph is enough to say what kind of scene it was. Photographs of coasts, forests, mountains, deserts, tall buildings, bridges, roads and towns flash for 17 to 133 ms and are covered by a mask. Scores accuracy at each display time and the display time at which accuracy is halfway between chance and perfect.
Constructs
- Scene gist
- Rapid visual categorization
- Visual masking
- Perceptual processing time
Research uses
- Visual perception research
- Scene and object recognition research
- Masking and timing research
- Within-study comparison of groups
How the paradigm works
The Scene Glimpse Task asks how short a look at a photograph is enough to say what kind of scene it shows. Each trial shows a cross for 500 ms, then a photograph of a coast, forest, mountain, desert, tall buildings, bridge, road or town for about 17, 33, 50, 83 or 133 ms, timed in whole screen frames, and at once a 250 ms mask made of the same photograph cut into small tiles and shuffled. The participant then chooses one of the eight kinds by click or number key, with no time limit. Each round shows every kind at every display time once, in a new order, and no photograph is shown twice; a practice of 4 photographs, two of them for 500 ms, comes first. Two rounds (standard; 1 to 4) give 80 trials.
Scores
- Gist Timems
- The display time at which accuracy reaches halfway from chance (1 in 8) to perfect, found from the accuracy at each display time; empty when accuracy does not get there by the longest time.
- Lower means a shorter glimpse was enough
- Accuracy%
- Share of usable answers that name the photograph's kind, over all display times.
- Chance is 0.125; not above chance is flagged
- Natural Scenes%
- Accuracy on coasts, forests, mountains and deserts.
- Informational
- Built Scenes%
- Accuracy on tall buildings, bridges, roads and towns.
- Informational
- Natural or Built%
- Share of usable answers on the right side of natural against built, whichever kind was named.
- Chance is 0.5 when a person only knows the group; higher than accuracy means mistakes stayed in the group
- Median Answer Timems
- Median time from the answers appearing to the answer, on usable answers.
- Informational; the answer has no time limit
- Anticipationscount
- Answers given within 100 ms of the answers appearing, which leave the scores.
- More than max(2, 20%) is invalid
- Early Key Pressescount
- Keys pressed before the answers 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
- Displays on Time%
- Share of photographs shown within one screen frame of their planned time.
- Below 0.9 is flagged and withholds the percentile
- Completed Trialscount
- Test trials answered.
- Fewer than planned is flagged
- Planned Trialscount
- Test trials the run planned.
- Informational
Use and evidence
32 papers from 2021 to 2025 name this task in their title or abstract, down 52% compared with 2016 to 2020. That ranks it 551 of 711 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 9 |
| 2024 | 5 |
| 2023 | 7 |
| 2022 | 5 |
| 2021 | 6 |
| 2020 | 9 |
| 2019 | 14 |
| 2018 | 18 |
| 2017 | 12 |
| 2016 | 14 |
| 2015 | 27 |
| 2014 | 20 |
| 2013 | 21 |
| 2012 | 20 |
| 2011 | 10 |
| 2010 | 31 |
| 2009 | 6 |
| 2008 | 11 |
| 2007 | 3 |
| 2006 | 3 |
| 2005 | 3 |
| 2004 | 8 |
| 2003 | 3 |
| 2002 | 1 |
| 2001 | 1 |
| 2000 | 2 |
Use grade D
30 to 99 papers from 2021 to 2025.
Evidence grade D
Little published measurement evidence found.
- Papers on test-retest reliability
- 1
- Papers on reliability or validity
- 5
- Registered clinical studies using it as an outcome
- 0
Measured on ConductCognition
Pooled across the studies that feed our platform norms, by device group. A figure needs 50 or more people.
No pooled figure yet.
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 0.58.0, counted 8 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Photographs
- Scenes of eight kinds: four natural (coast, forest, mountain, desert) and four built (tall buildings, bridge, road, town). 222 CC0 photographs credited on the terms page, 21 to 29 of each kind, with no person as the subject. No photograph shows twice in a session.
- Display
- A cross for 500 ms, then the photograph for 1, 2, 3, 5 or 8 frames of a 60 Hz screen (about 17, 33, 50, 83 and 133 ms), timed in whole screen frames, then a 250 ms mask made of the same photograph cut into tiles and shuffled.
- Answer
- The eight kinds as answers. The participant clicks one or presses its number from 1 to 8, with no time limit.
- Rounds
- 2 rounds (standard; 1 to 4), each with every kind at every display time once, in a new order: 80 trials, with a break halfway.
- Practice
- 4 photographs, two of them for 500 ms, each followed for 1,500 ms by the right answer.
- Scores
- Accuracy at each display time and overall, the display time at which accuracy is halfway from chance to perfect, accuracy on natural and on built scenes, and how often answers fell on the right side of natural against built.
Every score is an accuracy on one screen, so no calibration is claimed; the frame interval and the share of photographs shown on time are saved. Neither a shorter nor a longer glimpse is a grade. 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: scene_glimpse_test from the run with the most trials, ties going to the latest run; the kind chosen is read from the saved key (1 to 8) and the right kind from the saved row. Every answer has a right kind (chance 1 in 8): accuracy significantly below chance invalidates and not above chance is flagged; one key on 90% of answers invalidates when accuracy is not above chance. A short run or fewer than 90% of photographs within one frame of their planned time is flagged. Responses under 100 ms follow the timed-test rule.
Why these rules
Every photograph is of one kind, so answers can be compared with chance, which is 1 in 8 across the display times pooled; the shortest glimpses are near chance for most people, so the pooled accuracy only needs to beat it. A person who keeps pressing one key is not reporting the scenes. The display time is the manipulated variable, so photographs shown off plan leave the scores.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Rounds of photographs | 2 | 1 to 4 |
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 Scene Glimpse Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Potter MC, 1976; Thorpe S, Fize D, Marlot C, 1996; Greene MR, Oliva A, 2009), with the standard settings (rounds of photographs: 2). The main score was gist time (the display time at which accuracy reaches halfway from chance (1 in 8) to perfect, found from the accuracy at each display time; empty when accuracy does not get there by the longest time).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Potter, M. C. (1976). Short-term conceptual memory for pictures. Journal of Experimental Psychology: Human Learning and Memory, 2(5), 509-522. https://doi.org/10.1037/0278-7393.2.5.509Pictures shown in rapid succession and picked out by name: the display times of a brief-picture task.
- Thorpe, S., Fize, D., & Marlot, C. (1996). Speed of processing in the human visual system. Nature, 381(6582), 520-522. https://doi.org/10.1038/381520a0Natural photographs flashed for 20 ms and categorized, with the visual processing needed done in under 150 ms.
- Greene, M. R., & Oliva, A. (2009). The briefest of glances: The time course of natural scene understanding. Psychological Science, 20(4), 464-472. https://doi.org/10.1111/j.1467-9280.2009.02316.xExposure thresholds for scene categories and for global properties of scenes, shorter for the global properties.
- Bacon-Mace, N., Mace, M. J. M., Fabre-Thorpe, M., & Thorpe, S. J. (2005). The time course of visual processing: Backward masking and natural scene categorisation. Vision Research, 45(11), 1459-1469. https://doi.org/10.1016/j.visres.2005.01.004A mask after each photograph: accuracy rose with the delay between photograph and mask and levelled off within tens of milliseconds.
- Fei-Fei, L., Iyer, A., Koch, C., & Perona, P. (2007). What do we perceive in a glance of a real-world scene? Journal of Vision, 7(1), 10. https://doi.org/10.1167/7.1.10Photographs shown for 27 to 500 ms and masked at once, with free reports of what was seen.
Source papers
- Greene MR, Oliva A (2009). The briefest of glances: the time course of natural scene understanding. Psychological Science, 20(4):464-472.Paradigm source, Adults
Questions
How long does the Scene Glimpse Task take?
About 5 to 7 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?
Gist Time, Accuracy, Natural Scenes, Built Scenes, Natural or Built, Median Answer Time, Anticipations, Early Key Presses, Frame Interval, Refresh Rate, Displays on Time, Completed Trials and Planned Trials. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Scene Glimpse Task used in research?
32 papers from 2021 to 2025 name it in their title or abstract, down 52% compared with 2016 to 2020. That ranks it 551 of 711 test families in the ConductCognition test census (use grade D). Source: ConductCognition test census 0.58.0, counted 8 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Scene Glimpse Task have?
In the published research, 1 paper reports 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?
Rounds of photographs (1 to 4). 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.