Episodic Memory
Face Memory Test
Also called: face recognition memory test, unfamiliar face memory test, three-choice face memory task
Measure memory for unfamiliar faces: 6 faces learned from three angles, then picked out of three faces in the same pictures, new pictures and visual noise.
8-10 minUntimed
Free to start: 10 free participants per account, then $0.99 per participant for academic research, or any yearly plan. Preview any test free after you sign up.
- Papers, 2021 to 2025
- 93
- Evidence grade
- C
- Registered clinical studies
- 4
What it measures
Measures memory for unfamiliar faces. Participants learn 6 faces, then pick each from three faces in the same pictures, new pictures and visual noise.
Constructs
- Face recognition
- Face memory
- Visual episodic memory
Research uses
- Face perception and face memory research
- Individual differences in face recognition
- Within-study comparison of face memory
How the paradigm works
The face memory test measures how well people learn and recognize unfamiliar faces. Participants learn 6 faces, each from three pictures (turned left, front and turned right) shown for 3 seconds each. On each item they see three faces and choose the one they learned, with no time limit. Right after each face is learned, 3 items use its learned pictures (18 items). After the 6 faces are shown together for 20 seconds, 30 items use new pictures of the faces, smiling and in the three views. After a second review, 24 items use new pictures covered in visual noise. Practice uses a drawn face and gives feedback; the test gives none.
More about the paradigm
The design follows Duchaine and Nakayama (2006). The photos come from the Face Research Lab London Set (CC BY 4.0) and are grayscale, cropped to an oval that hides hair and clothing. The 4 distractors for each learned face share its gender and ethnic group and are close in age, and a learned face is never a distractor. The noise stage reuses the smiling pictures with noise at three levels and can be turned off.
The main score is the number of items correct, out of 72 (chance is 24). Scores for each stage and median answer times are also reported. Scores support research and do not diagnose.
Scores
- Items Correctcount
- Items on which the learned face was chosen, out of 72 (48 without the noise stage).
- Higher is better; chance is a third of the items
- Percent Correct%
- Items correct as a percent of the items shown.
- Higher is better; chance is 33%
- Same Pictures Correctcount
- Correct items, of 18, that show the learned pictures.
- Higher is better
- New Pictures Correctcount
- Correct items, of 30, that show new pictures of the learned faces.
- Higher is better
- Noise Correctcount
- Correct items, of 24, that show new pictures covered in visual noise.
- Higher is better
- Median Answer Timems
- Median time to choose a face; answers under 100 ms are left out.
- Context for accuracy
- Median Time, Same Picturesms
- Median answer time on the items with the learned pictures.
- Context for accuracy
- Median Time, New Picturesms
- Median answer time on the items with new pictures.
- Context for accuracy
- Median Time, Noisems
- Median answer time on the items with noise.
- Context for accuracy
Use and evidence
93 papers from 2021 to 2025 name this task in their title or abstract, down 19% compared with 2016 to 2020. That ranks it 194 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 20 |
| 2024 | 13 |
| 2023 | 17 |
| 2022 | 18 |
| 2021 | 25 |
| 2020 | 29 |
| 2019 | 23 |
| 2018 | 18 |
| 2017 | 19 |
| 2016 | 26 |
| 2015 | 20 |
| 2014 | 16 |
| 2013 | 10 |
| 2012 | 16 |
| 2011 | 6 |
| 2010 | 11 |
| 2009 | 3 |
| 2008 | 2 |
| 2007 | 1 |
| 2006 | 1 |
| 2005 | 3 |
| 2004 | 1 |
| 2003 | 0 |
| 2002 | 0 |
| 2001 | 0 |
| 2000 | 0 |
Use grade D
30 to 99 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
- 8
- Papers on reliability or validity
- 65
- Registered clinical studies using it as an outcome
- 4
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 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Faces
- 6 faces to learn, each with 4 distractor faces of the same gender and ethnic group and close in age, from the Face Research Lab London Set: grayscale photos cropped to an oval that hides hair and clothing. A learned face is never a distractor.
- Learning
- Each face in three pictures (turned left, front and turned right, neutral expression), 3 seconds each (standard; 2 to 5).
- Items
- Three faces side by side; the participant picks the learned face by click, tap or the 1, 2 or 3 key, with no time limit and no feedback. The learned face is equally often in each position, and no face comes twice in a row after the first stage.
- Stages
- 18 items with the learned pictures, 3 right after each face is learned; 30 items with new pictures (smiling, in the three views); and 24 items with new pictures covered in visual noise. Before the second and third stages, the 6 faces are shown together for 20 seconds (standard; 10 to 30).
- Noise
- Gaussian noise added to the faces at a standard deviation of 24, 36 or 48 gray levels (of 255), the same for the three faces of an item, each level on 8 items. The noise stage can be turned off.
- Practice
- A drawn face learned from three angles, then 3 items with feedback.
The design follows Duchaine and Nakayama (2006) with photos from the Face Research Lab London Set (CC BY 4.0); none of the published test's faces are used. The set has six pictures per person, so the noise stage reuses the smiling pictures.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: face_memory_test from the scored run, one choice among three faces per item (chance 1 in 3, 24 of 72). Accuracy significantly below chance invalidates and accuracy not above chance is flagged. One position on 90% of at least 12 answers invalidates when accuracy is not above chance. Answers under 200 ms on at least half of the items invalidate. Noise items shown without noise are flagged. A hidden tab withholds the percentile.
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. The learned face is equally often in each position. 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. The noise stage is harder only when the noise is drawn, so noise items shown without it are flagged. A hidden tab lets the study, delay or response timers run without the participant, so a percentile is withheld after one.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Time per learning picture (seconds) | 3 | 2 to 5 |
| Review of all faces (seconds) | 20 | 10 to 30 |
| Noise stage | On | On or off |
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.
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Face Memory Test in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Duchaine & Nakayama, 2006), with the standard settings (time per learning picture (seconds): 3; review of all faces (seconds): 20; noise stage: On). The main score was items correct (items on which the learned face was chosen, out of 72 (48 without the noise stage)).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Duchaine, B., & Nakayama, K. (2006). Neuropsychologia, 44(4), 576-585. https://doi.org/10.1016/j.neuropsychologia.2005.07.001
- DeBruine, L., & Jones, B. (2017). Face Research Lab London Set [Data set]. figshare. https://doi.org/10.6084/m9.figshare.5047666.v5
Source papers
- Duchaine B, Nakayama K (2006). Neuropsychologia, 44(4):576-585.Paradigm source, Adults
- DeBruine L, Jones B (2017). Face Research Lab London Set. figshare. Dataset. doi:10.6084/m9.figshare.5047666.v5Paradigm source, Adults
Questions
How long does the Face Memory Test 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?
Items Correct, Percent Correct, Same Pictures Correct, New Pictures Correct, Noise Correct, Median Answer Time, Median Time, Same Pictures, Median Time, New Pictures and Median Time, Noise. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Face Memory Test used in research?
93 papers from 2021 to 2025 name it in their title or abstract, down 19% compared with 2016 to 2020. That ranks it 194 of 395 test families in the ConductCognition test census (use grade D). Source: ConductCognition test census 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Face Memory Test have?
In the published research, 8 papers report its test-retest reliability and 4 registered clinical studies use it as an outcome (evidence grade C in the ConductCognition test census).
Which settings can a researcher change?
Time per learning picture (seconds) (2 to 5), review of all faces (seconds) (10 to 30) and noise stage (On or off). 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.