Episodic Memory
Memory Conformity Task
Also called: social contagion of memory task, social contagion paradigm, memory conformity paradigm, co-witness misinformation task, collaborative memory contagion task
Measure how a partner's report changes memory: study scenes, read a simulated partner's lists with some things wrong and some left out, then recognize what was there.
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
- 36
- Evidence grade
- D
- Registered clinical studies
- 0
Was this in the kitchen?What it measures
Measures how a partner's report changes memory for scenes. Participants study six scenes of everyday things, read a simulated partner's list for each, which names some things that were not there and leaves others out, then say whether pictures were in each scene. The main score is how much more often the partner's wrong names are called seen than new things. The last screen says the partner was a program and names its errors.
Constructs
- Social contagion of memory
- Memory conformity
- Source monitoring
- Recognition memory
Research uses
- False memory and suggestibility research
- Social influence on memory research
- Within-study comparison of groups and conditions
How the paradigm works
The Memory Conformity Task measures how another person's report changes memory for what was seen, in the social contagion form of Roediger, Meade and Bergman (2001): a participant and a partner study scenes of familiar places, the partner names some things that were not there, and the participant later reports those things as seen. Here the partner is a computer program, disclosed by default. Participants study six scenes (a kitchen, a desk, a workshop, a beach, a garden and a wardrobe), each with eight emoji pictures and shown for 15 seconds (8 to 30 as a setting). They then read the partner's list for each scene, which names four things that were shown and two that were not and leaves out the other four that were shown, much as Carnero-Sierra and Menor (2021) gave a partner's answers on written cards with no partner present.
More about the paradigm
After 60 seconds of true-or-false arithmetic (0 to 180), participants see eight pictures from each scene, one at a time, and answer Yes or No: was it in that scene? There is no time limit, and the instructions ask for Yes only for things seen in the scene. Each scene's eight are two of each kind: shown and listed, shown and left off, listed but not shown, and neither. Each place has twelve things that belong there, and the session's seed chooses which eight are shown, which are listed and which are asked about, so every thing takes every role equally often across sessions. A practice scene with two questions and feedback comes first. The standard wording says the partner is simulated by a computer program; a neutral wording leaves that line out, and the last screen always says the partner was a program whose lists were set in advance and names the listed things that were not in the scenes.
The primary score is the contagion effect: the share of listed-but-not-shown pictures called Yes minus the share of new pictures called Yes. Above 0 means the partner's wrong names were taken for things seen. Scores also give the two shares, the hit rates for shown things the partner listed and left off and their difference, the hit rate over all shown pictures, and corrected recognition (hits minus new pictures called Yes). Roediger, Meade and Bergman (2001) found the effect in recall, larger after 15 than after 60 seconds of study and for things that fit the scene. Meade and Roediger (2002) found that it persisted after warnings and on recognition and source tests, and that it was larger with an actual confederate than with a simulated one; Huang, Cheng and Rajaram (2024) found that a robot partner spread false memories as a human one did. A partner known to be a program may still influence less than a person, so a study comparing its results with confederate studies should say so. A second session would show the scenes again after the partner's errors were named, so the test is meant to be taken once by each person, and a repeat session is flagged. Scores describe how a partner's lists change recognition in this task, a research measure; they do not diagnose.
Scores
- Contagion Effect%
- Share of pictures the partner listed but the scene did not show that were called Yes, minus the share of new pictures called Yes.
- Above 0 means the partner's wrong names were taken for things seen
- Listed but Not Shown, Called Yes%
- Share of pictures the partner listed, but the scene did not show, that were called Yes.
- Compare with new pictures
- New, Called Yes%
- Share of pictures neither shown nor listed that were called Yes.
- The baseline
- Shown and Listed, Called Yes%
- Share of shown pictures the partner listed that were called Yes.
- Compare with shown pictures left off
- Shown and Left Off, Called Yes%
- Share of shown pictures the partner left off its list that were called Yes.
- Compare with shown pictures listed
- Listed Minus Left Off%
- Hit rate for shown pictures the partner listed minus that for shown pictures it left off.
- Above 0 means the list changed memory for things that were there
- Hit Rate%
- Share of all shown pictures called Yes.
- Memory for the scenes
- Corrected Recognition%
- Hit rate minus the share of new pictures called Yes.
- Higher means shown and new pictures were told apart better
- Median Answer Timems
- Median time to answer a picture, leaving out answers under 100 ms.
- Context for the answers
- Answers Under 100 mscount
- Answers given within 100 ms of the picture. They still count as answers.
- Expected to be 0
- Pictures Answeredcount
- Pictures answered in the scored run.
- Fewer than half of those planned is invalid
- Pictures Plannedcount
- Pictures the session set out to ask about.
- 48 in every session
- Median Time on a Listseconds
- Median time on the partner's list of a scene; Next wakes after 4 seconds.
- Context for the lists
- Time per Sceneseconds
- Seconds each scene showed.
- Setting context
- Delayseconds
- Seconds of arithmetic between the lists and the test.
- Setting context
- Earlier Sessionscount
- This participant's earlier sessions of the task, when scoring counted them.
- Above 0 is flagged for review
Use and evidence
36 papers from 2021 to 2025 name this task in their title or abstract, up 24% compared with 2016 to 2020. That ranks it 333 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 6 |
| 2024 | 15 |
| 2023 | 4 |
| 2022 | 4 |
| 2021 | 7 |
| 2020 | 4 |
| 2019 | 6 |
| 2018 | 2 |
| 2017 | 10 |
| 2016 | 7 |
| 2015 | 6 |
| 2014 | 8 |
| 2013 | 10 |
| 2012 | 13 |
| 2011 | 17 |
| 2010 | 6 |
| 2009 | 5 |
| 2008 | 7 |
| 2007 | 5 |
| 2006 | 3 |
| 2005 | 0 |
| 2004 | 1 |
| 2003 | 1 |
| 2002 | 1 |
| 2001 | 1 |
| 2000 | 1 |
Use grade D
30 to 99 papers from 2021 to 2025.
Evidence grade D
Little published measurement evidence found.
- Papers on test-retest reliability
- 0
- Papers on reliability or validity
- 14
- 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.
- Scenes
- 6 scenes of familiar places (a kitchen, a desk, a workshop, a beach, a garden and a wardrobe), drawn by code with emoji pictures. Each place has 12 things that belong there, and the session's seed shows 8 of them. Each scene shows for 15 seconds (standard; 8 to 30) with a 500 ms blank.
- Partner's lists
- A simulated partner's list for each scene, in the same order: 4 of the things shown and 2 that were not, read at the participant's pace; Next wakes after 4 seconds. The standard wording says the partner is a computer program.
- Delay
- 60 seconds (standard; 0 to 180) of true-or-false arithmetic.
- Test
- For each scene in turn, 8 pictures, each asked Yes or No: was it in that scene? Two of each kind: shown and listed, shown and left off, listed but not shown, and neither, with no more than 3 shown or not shown in a row. No time limit.
- Practice
- A practice scene with 2 questions and feedback.
- Last screen
- Says that the partner was a program whose lists were set in advance, and names the listed things that were not in each scene.
- Score
- Contagion effect: the share of listed-but-not-shown pictures called Yes minus the share of new pictures called Yes.
The scenes, the lists and the partner's rule were written for this task; no published scene set or confederate script is used. The partner is a computer program, disclosed by default, where the original studies used a confederate. A second session would repeat the scenes after the partner's errors were named, so each person takes the task once, and a repeat session is flagged.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: memory_conformity_test from the scored run, a yes or no answer for each picture with no time limit, half of them shown in their scene, with the partner's lists from the same run. Accuracy is not judged against chance over all pictures, because calling the partner's wrong names Yes is what the test measures. Instead, new pictures the partner never named called Yes significantly more often than shown pictures (one-sided Fisher exact test, p < .05) invalidate, and shown pictures not called Yes more often than new ones beyond chance are flagged and withhold the percentile. The same answer on 90% of the pictures invalidates. Answers under 200 ms on at least half of 4 or more pictures invalidate. Fewer than half the pictures answered invalidates, and any other unanswered picture is flagged. An earlier session of this test is flagged. A hidden tab withholds the percentile.
Why these rules
Half the pictures were in their scenes and half were not, and the pictures the partner named but did not show are expected to be called Yes, so accuracy over all pictures does not measure engagement. Shown pictures and new pictures the partner never named separate memory from guessing: Responses concentrated on trials that needed none (one-sided Fisher exact test, p < .05) mean the instructions were reversed and invalidate; responses that do not separate the two trial types are flagged. The same answer for nearly every picture means shown and new pictures were not told apart. 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. A second session has seen the scenes and been told which names were wrong, so an earlier session is flagged. 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 |
|---|---|---|
| Time per scene (seconds) | 15 | 8 to 30 |
| Arithmetic before the test (seconds) | 60 | 0 to 180 |
| Partner wording | Simulated by a computer program (standard) | Simulated by a computer program (standard), Neutral: a partner |
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, 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 Memory Conformity Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Roediger, Meade & Bergman, 2001; Meade & Roediger, 2002), with the standard settings (time per scene (seconds): 15; arithmetic before the test (seconds): 60; partner wording: Simulated by a computer program). The main score was contagion effect (share of pictures the partner listed but the scene did not show that were called Yes, minus the share of new pictures called Yes).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Roediger, H. L., Meade, M. L., & Bergman, E. T. (2001). Social contagion of memory. Psychonomic Bulletin & Review, 8(2), 365-371. https://doi.org/10.3758/BF03196174
- Meade, M. L., & Roediger, H. L. (2002). Explorations in the social contagion of memory. Memory & Cognition, 30(7), 995-1009. https://doi.org/10.3758/BF03194318Recognition and source tests after a partner's report, and a simulated confederate.
- Carnero-Sierra, S., & Menor, J. (2021). Aging does not enhance social contagion effect. Frontiers in Psychology, 12, 691229. https://doi.org/10.3389/fpsyg.2021.691229A partner's answers given as written cards, with no partner present.
- Huang, T.-R., Cheng, Y.-L., & Rajaram, S. (2024). Unavoidable social contagion of false memory from robots to humans. American Psychologist, 79(2), 285-298. https://doi.org/10.1037/amp0001230A partner known not to be a person still spreads false memories.
Source papers
- Roediger HL, Meade ML, Bergman ET (2001). Social contagion of memory. Psychonomic Bulletin & Review, 8(2):365-371.Paradigm source, Students
- Meade ML, Roediger HL (2002). Explorations in the social contagion of memory. Memory & Cognition, 30(7):995-1009.Paradigm source, Students
- Carnero-Sierra S, Menor J (2021). Aging does not enhance social contagion effect. Frontiers in Psychology, 12:691229.Paradigm source, Adults 21 to 34 and 60 to 88
- Huang TR, Cheng YL, Rajaram S (2024). Unavoidable social contagion of false memory from robots to humans. American Psychologist, 79(2):285-298.Paradigm source, Adults
Questions
How long does the Memory Conformity 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, tablets and phones. The browser must be Chrome, Edge, Firefox or Safari.
What scores does it produce?
Contagion Effect, Listed but Not Shown, Called Yes, New, Called Yes, Shown and Listed, Called Yes, Shown and Left Off, Called Yes, Listed Minus Left Off, Hit Rate, Corrected Recognition, Median Answer Time, Answers Under 100 ms, Pictures Answered, Pictures Planned, Median Time on a List, Time per Scene, Delay and Earlier Sessions. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Memory Conformity Task used in research?
36 papers from 2021 to 2025 name it in their title or abstract, up 24% compared with 2016 to 2020. That ranks it 333 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 Memory Conformity 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 D in the ConductCognition test census).
Does it report a percentile?
No. Scores and quality flags only.
Which settings can a researcher change?
Time per scene (seconds) (8 to 30), arithmetic before the test (seconds) (0 to 180) and partner wording (Simulated by a computer program (standard), Neutral: a partner). 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, 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.