Episodic Memory
Picture Partners Game
Also called: picture pair memory game, paired associates game, associative memory game, repeated memory test for older adults
Measure memory for pairs of pictures in adults aged 50 and over: the player studies 8 pairs, answers simple sums for 30 seconds, then picks each picture's partner from four cards, scored as the share of pairs recalled.
2-4 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
- 332
- Evidence grade
- B
- Registered clinical studies
- 204
Figures for the Paired Associate Learning test, the published test this game is compared with.
How these are counted
What it measures
A short picture-pair memory game for repeated sessions in research with adults aged 50 and over. The player sees 8 pairs of pictures, answers simple sums for 30 seconds, then picks each picture's partner from four cards. Each session takes the next of 12 alternate sets of pairs. Scores report the share of pairs recalled and the median choice time. They are raw research measures with no percentile.
Constructs
- Episodic memory
- Associative memory
- Paired associate learning
Research uses
- Research on memory in adults aged 50 and over, with short sessions repeated over days
How the paradigm works
The Picture Partners Game is a short picture-pair memory game for repeated sessions in research with adults aged 50 and over. The player sees 8 pairs of pictures, one pair at a time for 4 seconds each, answers simple true-or-false sums for 30 seconds, and then sees one picture of each pair and picks its partner from four cards, with no time limit and no feedback. The paper counts and grades on this page are those of the Paired Associate Learning test, the published test the game is compared with, not the game's own.
More about the paradigm
Each session takes one of 12 alternate sets of 8 pairs. The order of the sets is drawn once for each participant, and each session takes the next set after the sessions that participant began before it, so a reload keeps its set, a missed visit skips none, and no set comes back until all 12 are used. Every set holds one picture of each of 16 kinds, and the sets pair the kinds differently: two sets share at most one pairing of kinds, so the kinds paired in an earlier session answer at most one pair of a later one. No pair shares a theme. A set's scores are pooled with the other sets' only after first sessions show the sets are equally hard. A practice with two pairs comes in the first session only; later sessions open with a reminder. The pictures and the sets are our own.
Brief tests repeated over about a week kept 80% to 93% of scheduled sessions done in older adults (Nicosia and colleagues 2023; Thompson and colleagues 2022), and averaging 9 or more sessions gave reliable scores on the speed and working memory tasks of Thompson and colleagues (2022). A suggested schedule is 9 sessions: a first visit, then 8 visits that open on days 1 to 8, each open for 1 day. The score for repeated use is the mean of the sessions, taken from the scores export's visit_number column. For a retest 7 to 28 days later, add a visit to the same study rather than a second study: a second study gives the participant a new order of sets, so a set can come back.
Scores
- Pairs Recalled%
- Share of the scored pairs whose partner was chosen.
- Higher means more pairs remembered; compared within a study
- Median Choice Timems
- Median time from the cards appearing to the choice, over the scored pairs.
- Informational
- Usable Pairstrials
- Pairs counted in the scores. Left out: a pair whose study or test came while the screen was left, the window changed size, or a second finger or a palm touched the screen.
- Quality check
Use and evidence
332 papers from 2021 to 2025 name this task in their title or abstract, down 9% compared with 2016 to 2020. That ranks it 86 of 582 test families in our census. These counts belong to the Paired Associate Learning test, the published test this game is compared with. This game is a new measure, so its own reliability and norms come from use. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 70 |
| 2024 | 60 |
| 2023 | 62 |
| 2022 | 61 |
| 2021 | 79 |
| 2020 | 79 |
| 2019 | 63 |
| 2018 | 84 |
| 2017 | 69 |
| 2016 | 71 |
| 2015 | 67 |
| 2014 | 56 |
| 2013 | 59 |
| 2012 | 58 |
| 2011 | 50 |
| 2010 | 42 |
| 2009 | 33 |
| 2008 | 34 |
| 2007 | 29 |
| 2006 | 40 |
| 2005 | 27 |
| 2004 | 27 |
| 2003 | 27 |
| 2002 | 18 |
| 2001 | 16 |
| 2000 | 22 |
Use grade B
300 to 999 papers from 2021 to 2025.
Evidence grade B
At least 8 test-retest papers, and either 5 registered clinical studies or 150 reliability or validity papers.
- Papers on test-retest reliability
- 24
- Papers on reliability or validity
- 171
- Registered clinical studies using it as an outcome
- 204
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.
Note on the counts: includes verbal paired associates.
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 the Paired Associate Learning test, the published test this game is compared with, not the game's own.
Source: ConductCognition test census 0.39.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Pictures
- 198 color pictures from Noto Emoji (Apache 2.0): animals, nature and sky, vehicles, clothing, tools, machines, books and writing things, toys, and music and clocks. None is food or drink, and none comes from the Grocery Trip Game's six categories (fruit, vegetables, bakery, dairy, drinks and household things).
- Sets
- 12 alternate sets of 8 pairs, with no picture in two sets. Every set pairs 8 kinds of picture (animals, nature and vehicles) with 8 kinds of made thing, one picture of each kind, and the pairings of kinds turn from set to set: two sets share at most one, so the kinds paired in an earlier session answer at most one pair of a later one. No pair shares a theme. Each participant gets the sets in an order drawn once for them, and each session takes the next set after the sessions they began before it.
- Study
- The 8 pairs one at a time, two pictures side by side for 4,000 ms, with 500 ms blank between pairs. No answer is asked.
- Delay
- 30 seconds of simple true-or-false sums, which are not scored.
- Test
- Each pair's first picture comes back in a new order, above the four cards, or beside them on a wide screen: its partner and the partners of 3 other pairs of the set. The partner is in each of the 4 places twice. There is no time limit and no feedback, and the first tap, click or key press on a card chooses.
- Practice
- In the first session only: 2 practice pairs, then a choice for each among four cards, with a tick or a cross and the right partner shown. Later sessions open with a short reminder.
- Cards
- Every card is at least 120 px, laid out from the window before each trial. The game look adds a warm edge to every card and a soft background; the plain look shows the cards on white. Both looks give the pictures and cards the same outline, size and place.
- Schedule
- A suggested 9 sessions: a first visit, then 8 visits that open on days 1 to 8, each open for 1 day. The score for repeated use is the mean of the sessions, from the scores export's visit_number column.
- Scores
- The share of scored pairs whose partner was chosen (the main score), the median choice time and the number of usable pairs. A pair is left out when the screen was left, the window changed size, or a second finger or a palm touched the screen while it was studied or tested.
Brief repeated sessions follow the phone studies of Thompson and colleagues (2022) and Nicosia and colleagues (2023); the pictures, sets, words and looks are our own. This is a new measure for research with adults aged 50 and over: its reliability and norms come from use, the figures published in those studies are not its own, and it computes no percentiles. For a retest 7 to 28 days later, add a visit to the same study.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: picture_partners_game test choices, one of 4 cards per picture, leaving out a pair whose study or test trial had the tab hidden, the window resized, or a second finger or a palm on the screen. Fewer than 6 scored pairs of 8, or 5 or more answers under 200 ms, invalidates. The card in the same place on 7 or more answers is flagged, as is a tab hidden during the delay. A hidden tab withholds the percentile.
Why these rules
Guessing among 4 cards gives 25%, but 8 answers are too few to test accuracy against chance in one session, and recalling few pairs can be a real result in memory impairment, so accuracy is not judged against chance. The partner sits in each of the 4 places twice, so the same place on 7 or more answers means the cards may not have been compared. An answer under 200 ms is faster than a simple visual reaction, so it cannot reflect the cards; following response-time effort screening for untimed tests (Wise and Kong 2005), the session is invalid when 5 or more of the 8 answers are that fast. A pair studied or tested while the tab was hidden never counts as recalled. A tab hidden during the delay only lengthens the time every pair is held, so the pairs still count and the session is flagged for review. The 6-pair floor, the 7-answer place limit and the 5-answer speed limit are ConductScience choices, not published ones. A hidden tab lets the study, delay or response timers run without the participant, so a percentile is withheld after one.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Look | Game | Game, Plain |
Every plan can change the look, which never changes how a result is scored.
Need a setting this test does not have?
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.
- 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 Picture Partners Game in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Thompson and colleagues, 2022), with the standard settings (look: Game). The main score was pairs recalled (share of the scored pairs whose partner was chosen).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Thompson, L. I., Harrington, K. D., Roque, N., et al. (2022). A highly feasible, reliable, and fully remote protocol for mobile app-based cognitive assessment in cognitively healthy older adults. Alzheimer's & Dementia: Diagnosis, Assessment & Disease Monitoring, 14(1), e12283. https://doi.org/10.1002/dad2.12283Brief phone tests three times a day for 8 days in 52 cognitively healthy older adults, who completed 93% of the scheduled sessions. A single session's reliability was ICC .26 to .53, and the mean of 9 sessions reached .76 and .82 on the speed and working memory tasks; the associative memory task had the lowest single-session value (.26). These figures belong to that app's tasks, not to this game.
- Nicosia, J., Aschenbrenner, A. J., Balota, D. A., et al. (2023). Unsupervised high-frequency smartphone-based cognitive assessments are reliable, valid, and feasible in older adults at risk for Alzheimer's disease. Journal of the International Neuropsychological Society, 29(5), 459-471. https://doi.org/10.1017/S135561772200042XBrief associative memory, processing speed and working memory tests on a phone, up to 4 times a day for 7 days and unsupervised, in 268 cognitively normal adults aged 65 to 97 and 22 with very mild dementia, who completed 80% of the scheduled sessions (abstract read).
Source papers
Questions
How long does the Picture Partners Game take?
About 2 to 4 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?
Pairs Recalled, Median Choice Time and Usable Pairs. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Paired Associate Learning test used in research?
These counts belong to the Paired Associate Learning test, the published test this game is compared with. This game is a new measure, so its own reliability and norms come from use. 332 papers from 2021 to 2025 name the task in their title or abstract, down 9% compared with 2016 to 2020. That ranks it 86 of 582 test families in the ConductCognition test census (use grade B). Source: ConductCognition test census 0.39.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Paired Associate Learning test have?
These counts belong to the Paired Associate Learning test, the published test this game is compared with. This game is a new measure, so its own reliability and norms come from use. In the published research, 24 papers report the task's test-retest reliability and 204 registered clinical studies use it as an outcome (evidence grade B in the ConductCognition test census).
Does it report a percentile?
No. Scores and quality flags only.
Which settings can a researcher change?
Look (Game, Plain). Every plan can change the look, which never changes how a result is scored.
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.