Episodic Memory
Fan Effect Task
Also called: fan effect paradigm, fact retrieval paradigm, person-location paradigm, fan effect test, associative interference task
Measure how memory retrieval slows as facts share a concept: participants learn made-up facts about people and places, then recognize them among re-paired sentences, with the facts written in each of the eight app languages.
13-16 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
- 23
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
Measures how recognizing a fact slows as more facts share its person or place. Participants learn 18 made-up facts such as "The painter is on the island", then decide quickly whether each sentence is a learned fact or a re-paired one. The main score is the fan slope.
Constructs
- Episodic memory retrieval
- Associative interference
- Memory for facts
Research uses
- Memory and interference research
- Research on aging and memory retrieval
- Within-study group comparison
How the paradigm works
The Fan Effect Task measures how retrieving a fact from memory slows as more facts share its parts. Participants learn 18 made-up facts that each say where a person is, such as "The painter is on the island". Two people are in 3 places each, three in 2 and six in 1, and the places likewise, so each fact's person and place carry 1 to 3 facts: its fans. Participants then decide as quickly and accurately as possible whether each sentence is one of the facts, pressing F or J. The paradigm comes from Anderson (1974), and the learning phase follows Schneider and Anderson (2012): recognition takes longer the more facts are attached to a sentence's person or place, the fan effect.
More about the paradigm
Each fact is shown for 5 seconds with 500 ms between. Then come 2 rounds (1 to 3 as a setting) of 22 questions, where each person is and who is in each place, answered by choosing among the 11 places or the 11 people. A wrong answer shows the right ones for 3 seconds and comes back at the end of the round, up to 4 tries, and a round ends when every question has been answered right once, the dropout procedure of Schneider and Anderson. After 8 trials that practice the keys, the test shows 3 blocks (2 to 4) of the 18 facts and 18 re-paired sentences, which pair the same people and places, with the same numbers of facts, in ways never studied, as Radvansky, O'Rear and Fisher (2017) built theirs. No person or place comes twice in a row and no more than 3 sentences in a row share an answer. A blank of 500 ms and a cross of 1,000 ms come before each sentence, which stays until a key or 5 seconds, and a wrong or missed answer brings a message for 1 second. Which word fills which place of the design is drawn from the session's seed, so across sessions every person and place carries every fan. Every fact was written for this library in each of the eight languages, and the facts follow the session's language.
The main score, the fan slope, is the slope of the median time to accept a learned fact on its fan (1, 2 or 3 facts), with the person's fan and the place's fan counted alike, over right answers at or after 100 ms: how much longer recognition takes for each more fact attached to a concept. The task also reports the slopes over the person's and the place's fans alone, the same slope for rejecting re-paired sentences, the median times, accuracy, the hit and false alarm rates, and how the learning rounds went. One person's slope rests on 18 answers to facts at each fan of a person or a place in the standard design, so it is a research measure best compared between groups. Scores support research and do not diagnose.
Scores
- Fan Slopems/item
- Slope of the line fitted to the median time to accept a learned fact at each fan (1, 2 and 3 facts), the person's and the place's fans counted alike, over right answers.
- The time each more fact on a person or place adds to recognition; not better or worse
- Fan Slope, Peoplems/item
- The same slope over the fan of the fact's person only.
- Context for the fan slope
- Fan Slope, Placesms/item
- The same slope over the fan of the fact's place only.
- Context for the fan slope
- Fan Slope, Re-paired Sentencesms/item
- The same slope for rejecting re-paired sentences.
- Context for the fan slope; often steeper than for facts
- Median Time, Factsms
- Median time to accept a learned fact, over right answers.
- Lower means faster recognition
- Median Time, Re-paired Sentencesms
- Median time to reject a re-paired sentence, over right answers.
- Lower means faster rejection
- Accuracy%
- Right answers as a share of right and wrong ones, over all test sentences.
- Below 0.80 is flagged for review
- Hit Rate%
- Facts answered learned, as a share of the facts answered.
- Higher means more facts recognized
- False Alarm Rate%
- Re-paired sentences answered learned, as a share of those answered.
- Lower means fewer re-pairings accepted
- Learning Roundscount
- Rounds of questions run before the test.
- The study's setting
- Learning Answerscount
- Answers given to the learning questions, repeats included.
- Fewer means the facts were learned sooner
- Learning Errorscount
- Wrong answers to the learning questions.
- Fewer means the facts were learned sooner
- Learning Criterion Met0/1
- 1 when every question of every round was answered right before its tries ran out, otherwise 0.
- 0 is flagged for review
- Fewest Usable Trialscount
- Usable times to facts at the fan of a person or a place with the fewest.
- Under 8 is flagged for review
- Missed Sentencescount
- Test sentences not answered within 5 seconds.
- Expected to be 0
- Anticipationscount
- Answers under 100 ms after the sentence appeared, counted neither right nor wrong.
- Expected to be 0
Use and evidence
23 papers from 2021 to 2025 name this task in their title or abstract, up 44% compared with 2016 to 2020. That ranks it 390 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 11 |
| 2024 | 5 |
| 2023 | 2 |
| 2022 | 3 |
| 2021 | 2 |
| 2020 | 2 |
| 2019 | 4 |
| 2018 | 2 |
| 2017 | 4 |
| 2016 | 4 |
| 2015 | 2 |
| 2014 | 5 |
| 2013 | 2 |
| 2012 | 1 |
| 2011 | 5 |
| 2010 | 4 |
| 2009 | 4 |
| 2008 | 3 |
| 2007 | 6 |
| 2006 | 2 |
| 2005 | 3 |
| 2004 | 4 |
| 2003 | 2 |
| 2002 | 1 |
| 2001 | 1 |
| 2000 | 3 |
Use grade E
Fewer than 30 papers from 2021 to 2025.
Evidence grade D
Little published measurement evidence found.
- Papers on test-retest reliability
- 0
- Papers on reliability or validity
- 2
- 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.
- Facts
- 18 made-up facts pairing 11 people with 11 places, written for this test in each of the eight languages: 2 people and 2 places in 3 facts each, 3 in 2 and 6 in 1, so 2 facts for each pairing of a person's and a place's number of facts. Which word fills which place is drawn from the session's seed.
- Study
- Each fact for 5,000 ms with 500 ms between, in a seeded order with no person or place twice in a row.
- Learning
- 2 rounds (1 to 3 as a setting) of 22 questions, where each person is and who is in each place, answered by choosing among the 11 places or people. A wrong answer shows the right ones for 3,000 ms and comes back at the end of the round, up to 4 tries; a round ends when every question has been answered right once.
- Sentences
- Each block shows the 18 facts and 18 re-paired sentences, which pair the same people and places with the same numbers of facts but were never studied: 3 blocks (2 to 4 as a setting), with no person or place twice in a row and no more than 3 of one answer in a row.
- Trial
- A blank for 500 ms and a cross for 1,000 ms, then the sentence until F or J or 5,000 ms. A wrong or missed answer shows a message for 1,000 ms. 8 key practice trials come first. Which key means learned is drawn from the session's ID.
- Scoring
- The fan slope is the slope of the median time to accept a learned fact on its fan (1, 2 or 3 facts), with the person's and the place's fans counted alike, over right answers at or after 100 ms. The same slope for re-paired sentences and the slopes over people and places alone are reported too.
The facts were written for this test in each language, not taken from published sets. The study list of 18 facts with fans of 1 to 3 follows Radvansky and colleagues, and the dropout rounds Schneider and Anderson. One person's slope rests on 18 answers to facts at each fan of a person or a place, so compare groups, not single people. 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: fact_recognition_test from the scored run, F or J for a learned fact or a re-paired sentence (chance 50%, half the sentences are facts), correctness from the stored key and answer. Below chance invalidates and not above chance is flagged. One key on 90% of answers invalidates when accuracy is not above chance. Answering fewer than half the planned sentences invalidates, and any other incomplete run is flagged. Accuracy under 80%, a learning round with a question still wrong when its tries ran out, or fewer than 8 usable times to facts at any fan of a person or a place, is flagged for review. Missed responses and responses under 100 ms follow the timed-test rule.
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). Missed responses and responses under 100 ms keep the platform's timed-test rule: more than max(3, 25%) missed responses or more than max(2, 20%) responses under 100 ms invalidate, and any is flagged. A response under 100 ms cannot be a reaction to the stimulus. The fan effect is read from the time to recognize facts that were learned, so a fact not learned to the criterion, or few right answers at a fan, leaves the slope without its basis; Radvansky, O'Rear and Fisher (2017) report error rates of .02 to .08 per cell when tested right after learning to a criterion.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Learning rounds | 2 | 1 to 3 |
| Test blocks | 3 | 2 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 Fan Effect Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Anderson, 1974), with the standard settings (learning rounds: 2; test blocks: 3). The main score was fan slope (slope of the line fitted to the median time to accept a learned fact at each fan (1, 2 and 3 facts), the person's and the place's fans counted alike, over right answers).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Anderson, J. R. (1974). Retrieval of propositional information from long-term memory. Cognitive Psychology, 6(4), 451-474. https://doi.org/10.1016/0010-0285(74)90021-8The paradigm: facts about people in places, recognized more slowly the more facts share the person or the place.
- Schneider, D. W., & Anderson, J. R. (2012). Modeling fan effects on the time course of associative recognition. Cognitive Psychology, 64(3), 127-160. https://doi.org/10.1016/j.cogpsych.2011.11.001Facts studied for 5 s each with 500 ms between, questions about each person and place in a dropout procedure until each was answered right, and foils re-paired with the same fans.
- Radvansky, G. A., O'Rear, A. E., & Fisher, J. S. (2017). Event models and the fan effect. Memory & Cognition, 45(6), 1028-1044. https://doi.org/10.3758/s13421-017-0713-418 sentences with fans of 1 to 3, learned to a criterion, re-paired sentences built within the same fans, and an error message during the test; error rates of .02 to .08 per cell right after learning.
Source papers
- Anderson JR (1974). Retrieval of propositional information from long-term memory. Cognitive Psychology, 6(4):451-474.Paradigm source, Adults
- Radvansky GA, O'Rear AE, Fisher JS (2017). Event models and the fan effect. Memory & Cognition, 45(6):1028-1044.Paradigm source, Adults
- No normative data for this task.Paradigm source, N/A
Questions
How long does the Fan Effect Task take?
About 13 to 16 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?
Fan Slope, Fan Slope, People, Fan Slope, Places, Fan Slope, Re-paired Sentences, Median Time, Facts, Median Time, Re-paired Sentences, Accuracy, Hit Rate, False Alarm Rate, Learning Rounds, Learning Answers, Learning Errors, Learning Criterion Met, Fewest Usable Trials, Missed Sentences and Anticipations. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Fan Effect Task used in research?
23 papers from 2021 to 2025 name it in their title or abstract, up 44% compared with 2016 to 2020. That ranks it 390 of 495 test families in the ConductCognition test census (use grade E). Source: ConductCognition test census 0.22.0, counted 3 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Fan Effect 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?
Learning rounds (1 to 3) and test blocks (2 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, 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.