Attention
Two-Tone Pictures Task
Also called: Mooney images, Mooney faces task, two-tone images task, perceptual closure task, face detection in two-tone images
Measure perceptual closure: participants see pictures made only of black and white patches, upright or upside down, and say whether each shows a face or which food it shows, scored by how much more often upright pictures are seen than scrambled ones.
6-8 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
- 54
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
Measures perceptual closure: seeing a whole face or object in a picture made only of black and white patches. Participants see each two-tone picture for a moment, upright or upside down, and say whether it shows a face, or which food it shows; half the pictures are scrambled patches that show neither. The main score is how much more often upright pictures are seen than the scrambled ones.
Constructs
- Perceptual closure
- Face detection from fragments
- Face inversion effects
- Object recognition from incomplete pictures
Research uses
- Perceptual organization research, for example in psychosis, compared at the group level
- Development and aging of visual closure
- Batteries of face perception tasks with the part-whole and Thatcher illusion tasks
How the paradigm works
The two-tone pictures task measures perceptual closure: seeing a whole face or object in a picture made only of black and white patches. Mooney (1957) made a test of closure from two-tone pictures of faces, and pictures of this kind, often called Mooney images, are used to study how vision joins fragments into a whole. Each trial shows one picture briefly, upright or upside down, a face for 500 ms and a food for 1,500 ms, and then the participant answers with no time limit. In the face block the question is whether the picture shows a face; in the food block, which of four foods it shows, or no food. Half the pictures of each block are scrambled controls, made from the same pictures with the same share of black and white, that show neither. With their own two-tone faces shown for 100 ms, Schwiedrzik, Melloni and Schurger (2018) found a d prime of 1.78 for upright faces against scrambled ones and of 0.83 for upside-down faces.
More about the paradigm
The pictures are our own, made by code from photographs the platform already carries: the 3/4 views of 22 people of the Face Research Lab London Set (DeBruine and Jones, 2017), 20 of them in the test, and plated pictures of 13 foods from the CROCUFID database (Toet and colleagues, 2019), 12 in the test. Each photograph is blurred and cut into two tones; each control is the same picture warped by a smooth random distortion until its patches form nothing, after the diffeomorphic scrambling of Stojanoski and Cusack (2014). The published Mooney and Gollin picture sets are not used. Each picture appears once as itself and once as a control, and which ones are upside down is drawn for each session. Practice with feedback comes before each block: 4 face pictures and 2 food pictures.
The main score is perceptual closure: d prime for upright pictures against the controls, faces and foods together, a face counting as seen when the answer is Yes and a food when any food is named, with the log-linear correction (Hautus, 1995). The task also reports d prime for upside-down pictures against the same controls, the inversion cost, d prime for upright faces and for upright foods, the percent of faces seen upright and upside down, the percent of foods named right, the controls answered as a face or a food, accuracy and the median answer time. Answers within 100 ms of the buttons are counted as anticipations and left out. Scores support research and do not diagnose.
Scores
- Perceptual Closured'
- z(share of upright pictures seen) minus z(share of controls seen), faces and foods together: a face is seen when the answer is Yes, a food when any food is named. Both shares log-linear corrected (Hautus, 1995).
- Higher means upright pictures were told from scrambled ones better
- Closure Upside Downd'
- The same for upside-down pictures, against the same controls.
- Higher means upside-down pictures were told from scrambled ones better
- Inversion Costd'
- Closure upright minus closure upside down. Positive means turning a picture upside down made it harder to see.
- Informational; an effect of orientation
- Face Closured'
- d prime for upright faces against the face controls, from the face block's Yes and No answers.
- Higher means faces were told from scrambled pictures better
- Food Closured'
- d prime for upright foods against the food controls, counting any food named as seen.
- Higher means foods were told from scrambled pictures better
- Upright Faces Seen%
- Percent of upright face pictures answered Yes.
- Informational
- Upside-Down Faces Seen%
- Percent of upside-down face pictures answered Yes.
- Informational
- Face Controls Answered Yes%
- Percent of scrambled face pictures answered Yes.
- Quality check
- Upright Foods Named%
- Percent of upright food pictures given the right food among four.
- Informational
- Upside-Down Foods Named%
- Percent of upside-down food pictures given the right food among four.
- Informational
- Food Controls Named%
- Percent of scrambled food pictures given a food name rather than No food.
- Quality check
- Accuracy%
- Percent of answered trials answered right: Yes for a face, No for a face control, the right food, No food for a food control.
- Quality check
- Median Answer Timems
- Median time from the buttons appearing to the answer, over answers at or after 100 ms.
- Informational; the answer has no time limit
- Trials Answeredtrials
- Trials answered in the scored run, leaving out anticipations.
- Fewer than planned is flagged for review
- Planned Trialstrials
- Trials the session planned: 64.
- Informational
- Anticipationscount
- Answers within 100 ms of the buttons appearing, left out of the other scores.
- Quality check
Use and evidence
54 papers from 2021 to 2025 name this task in their title or abstract, about the same as from 2016 to 2020. That ranks it 296 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 17 |
| 2024 | 11 |
| 2023 | 8 |
| 2022 | 8 |
| 2021 | 10 |
| 2020 | 9 |
| 2019 | 12 |
| 2018 | 13 |
| 2017 | 10 |
| 2016 | 10 |
| 2015 | 7 |
| 2014 | 13 |
| 2013 | 12 |
| 2012 | 5 |
| 2011 | 5 |
| 2010 | 19 |
| 2009 | 12 |
| 2008 | 12 |
| 2007 | 1 |
| 2006 | 5 |
| 2005 | 5 |
| 2004 | 7 |
| 2003 | 7 |
| 2002 | 2 |
| 2001 | 2 |
| 2000 | 0 |
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
- 7
- 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.
- Pictures
- Two-tone pictures made by our code from photographs: 20 faces from the 3/4 views of Face Research Lab London Set photographs, 10 women and 10 men, white patches on black up to 280 by 320 px, and 12 foods from CROCUFID plated pictures, black patches on white up to 480 by 370 px. Each photograph is blurred and cut into two tones, 40% of the face white and 45% of the food black.
- Controls
- Each picture also appears as a scrambled control: the same picture warped by a smooth random distortion until its patches form nothing, with the same share of black and white.
- Each trial
- A 500 ms cross, then the picture, a face for 500 ms or a food for 1,500 ms, upright or upside down. Then the question with no time limit: "Do you see a face?" with Yes and No, or "Which food is it?" with four foods and No food. Answers within 100 ms of the buttons are counted as anticipations and left out. A 500 ms blank follows each answer.
- Trials
- 64: 40 faces, then 24 foods. Each picture and each control appears once, half upright and half upside down, in an order drawn for each session with no more than 4 pictures or 4 controls in a row.
- Practice
- 4 face pictures of 2 people the test does not use before the faces, and 2 pictures of a food it does not use before the foods, with feedback for 1,500 ms.
The face block follows the face detection task with two-tone faces of Schwiedrzik, Melloni and Schurger (2018): face or no face, upright and upside down, against scrambled controls, scored in d prime with the log-linear correction. The food block follows the four-choice naming of two-tone objects of Reining and Wallis (2024). The pictures are our own, made by code; the published Mooney and Gollin sets are not used. The controls are warped by code where the published ones were moved patch by patch by hand and checked in a pilot, which these were not, so report the false-alarm rates with the closure score. Faces show for 500 ms where Schwiedrzik and colleagues used 100 ms, so the timing holds on phones and tablets. Seeing a photograph first makes its two-tone picture easier to see (Teufel and colleagues, 2015). Its faces are made from the 3/4 photographs of 22 of the Face Memory Test's 24 distractor faces, and its foods from 10 of the Food-Picture Dot-Probe's pictures, 13 of the Food Choice Task's, 12 of the Food Cue Reactivity Task's and 1 of the Evaluative Conditioning Task's, so in a study with any of them, run this task first or in a separate session. Its 22 people are also among the Verbal Overshadowing Task's 36 and include 19 of the Face Matching Task's 40, 8 of the Chimeric Faces Task's 18, 13 of the Part-Whole Face Task's 30 and 4 of the Thatcher Illusion Task's 14, tests that show only their front photographs. No screen names or labels a face. 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: two_tone_pictures_test from the scored run, one answer per picture with no time limit: Yes or No for a face (chance 1 in 2), one of 4 foods or No food for a food (chance 1 in 5). Accuracy significantly below chance invalidates and not above chance is flagged. One kind of answer, seen or not seen, on 90% of trials invalidates when accuracy is not above chance; half the trials are pictures and half controls. Answers under 200 ms on at least half the trials invalidate. A run short of its planned trials is flagged for review, and so is a closure score at or below 0.
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 200 ms is faster than a simple visual reaction, so it cannot reflect the item. Following response-time effort screening for untimed tests (Wise and Kong 2005), the session is invalid only when at least half the answers are that fast. Upright two-tone faces are seen more often than scrambled patches (Mooney, 1957), so no difference means the pictures were not looked at, and each picture shows for a set time, so a hidden tab withholds the percentile.
Running this test
Settings you can change
This test runs with fixed settings.
Need a different version of this test?
Devices and languages
- Untimed test: runs on desktops, laptops, tablets and phones. 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 Two-Tone Pictures Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Mooney, 1957). The main score was perceptual closure (z(share of upright pictures seen) minus z(share of controls seen), faces and foods together: a face is seen when the answer is Yes, a food when any food is named).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Mooney, C. M. (1957). Age in the development of closure ability in children. Canadian Journal of Psychology / Revue canadienne de psychologie, 11(4), 219-226. https://doi.org/10.1037/h0083717
- Schwiedrzik, C. M., Melloni, L., & Schurger, A. (2018). Mooney face stimuli for visual perception research. PLOS ONE, 13(7), e0200106. https://doi.org/10.1371/journal.pone.0200106Face or no face, upright and upside down, against scrambled controls: d prime 1.78 upright and 0.83 upside down at 100 ms.
- Reining, L. C., & Wallis, T. S. A. (2024). A psychophysical evaluation of techniques for Mooney image generation. PeerJ, 12, e18059. https://doi.org/10.7717/peerj.18059Two-tone objects named among four choices after a 1,500 ms display; more blur made them harder to name.
- Verhallen, R. J., & Mollon, J. D. (2016). A new Mooney test. Behavior Research Methods, 48(4), 1546-1559. https://doi.org/10.3758/s13428-015-0666-0Two-tone faces made from photographs by a blur and a threshold, with a test-retest correlation of .80.
- Ke, T.-W., Yu, S. X., & Whitney, D. (2017). Mooney face classification and prediction by learning across tone. 2017 IEEE International Conference on Image Processing (ICIP), 2025-2029. https://doi.org/10.1109/ICIP.2017.8296637Two-tone faces are about 40% white.
- Stojanoski, B., & Cusack, R. (2014). Time to wave good-bye to phase scrambling: Creating controlled scrambled images using diffeomorphic transformations. Journal of Vision, 14(12), 6. https://doi.org/10.1167/14.12.6Scrambling by smooth warps, which keeps an image's basic visual properties and removes its meaning.
- Teufel, C., Subramaniam, N., Dobler, V., Perez, J., Finnemann, J., Mehta, P. R., Goodyer, I. M., & Fletcher, P. C. (2015). Shift toward prior knowledge confers a perceptual advantage in early psychosis and psychosis-prone healthy individuals. Proceedings of the National Academy of Sciences, 112(43), 13401-13406. https://doi.org/10.1073/pnas.1503916112Two-tone pictures were seen better after the photographs, more so in early psychosis.
- Uhlhaas, P. J., Linden, D. E. J., Singer, W., Haenschel, C., Lindner, M., Maurer, K., & Rodriguez, E. (2006). Dysfunctional long-range coordination of neural activity during Gestalt perception in schizophrenia. The Journal of Neuroscience, 26(31), 8168-8175. https://doi.org/10.1523/JNEUROSCI.2002-06.2006Upright two-tone faces shown for 200 ms: 74.3% seen by patients with schizophrenia and 83.6% by controls, compared at the group level.
- Hautus, M. J. (1995). Corrections for extreme proportions and their biasing effects on estimated values of d'. Behavior Research Methods, Instruments, & Computers, 27(1), 46-51. https://doi.org/10.3758/BF03203619The log-linear correction behind d prime.
- DeBruine, L., & Jones, B. (2017). Face Research Lab London Set [Data set]. figshare. https://doi.org/10.6084/m9.figshare.5047666.v5
- Toet, A., Kaneko, D., de Kruijf, I., Ushiama, S., van Schaik, M. G., Brouwer, A.-M., Kallen, V., & van Erp, J. B. F. (2019). CROCUFID: A cross-cultural food image database for research on food elicited affective responses. Frontiers in Psychology, 10, 58. https://doi.org/10.3389/fpsyg.2019.00058
Source papers
- Mooney CM (1957). Age in the development of closure ability in children. Canadian Journal of Psychology, 11(4):219-226.Paradigm source, Children and adults
- Schwiedrzik CM, Melloni L, Schurger A (2018). Mooney face stimuli for visual perception research. PLOS ONE, 13(7):e0200106.N 18, Adults
- DeBruine L, Jones B (2017). Face Research Lab London Set. figshare. Dataset. doi:10.6084/m9.figshare.5047666.v5Paradigm source, Adults
- Toet A, Kaneko D, de Kruijf I, Ushiama S, van Schaik MG, Brouwer AM, Kallen V, van Erp JBF (2019). CROCUFID: A cross-cultural food image database for research on food elicited affective responses. Frontiers in Psychology, 10:58.Paradigm source, Adults
Questions
How long does the Two-Tone Pictures Task take?
About 6 to 8 minutes, as listed in the test library.
What device does a participant need?
Untimed test: runs on desktops, laptops, tablets and phones. The browser must be Chrome, Edge, Firefox or Safari.
What scores does it produce?
Perceptual Closure, Closure Upside Down, Inversion Cost, Face Closure, Food Closure, Upright Faces Seen, Upside-Down Faces Seen, Face Controls Answered Yes, Upright Foods Named, Upside-Down Foods Named, Food Controls Named, Accuracy, Median Answer Time, Trials Answered, Planned Trials and Anticipations. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Two-Tone Pictures Task used in research?
54 papers from 2021 to 2025 name it in their title or abstract, about the same as from 2016 to 2020. That ranks it 296 of 495 test families in the ConductCognition test census (use grade D). Source: ConductCognition test census 0.22.0, counted 3 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Two-Tone Pictures 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).
Which languages can participants use?
Participants can take this test in English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese, with its words in the same language. 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.