Language
Semantic Satiation Task
Also called: verbal satiation task, semantic satiation paradigm, repetition-induced loss of meaning, word repetition and relatedness judgment
Measure how a word's meaning fades with repetition: after a word is shown many times in a row, participants are slower to judge that it is related to another word, with the words written in each of the eight app languages.
8-10 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
- 28
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
Measures how a word's meaning fades when it is seen over and over. A word flashes about 3 or about 22 times in a row, then appears with a second word, and participants press F or J to say whether the two are related in meaning. The score is how much slower related pairs are judged after many showings than after a few. The words are written in each of the eight languages.
Constructs
- Semantic memory
- Access to word meaning
- Repetition and adaptation
Research uses
- Semantic memory and language research
- Research on aging and word meaning
- Within-study group comparison
How the paradigm works
The Semantic Satiation Task measures how a word's meaning fades when the word is seen over and over, the feeling that a much-repeated word has turned into mere letters. A word appears several times in a row, then once more with a second word under it, and the participant decides as quickly and accurately as possible whether the two are related in meaning, pressing F or J. The method follows Smith (1984), whose participants said a category name 3 or 30 times and were slower after 30 to decide that a word belonged to it, and Balota and Black (1997), whose participants judged a repeated word and a second word related or unrelated after up to 24 showings. In Balota and Black's second experiment the word was only watched, not spoken, and the advantage of related pairs fell by 46 ms from the fewest to the most showings in young adults, against 9 ms in older adults; in their fourth experiment, where the repeated word had nothing to do with the pair, it did not fall at all.
More about the paradigm
Each session shows 4 practice trials with feedback, then 40 test trials without, with a pause halfway: 10 of each kind, related or unrelated after a few showings (2 to 4) or many (20 to 24). The number is drawn for each trial, so the pair's moment cannot be counted to, and no more than three trials in a row share an answer or a kind of showing. A cross shows for 500 ms; each showing lasts 500 ms with 200 ms between, and the pair follows 300 ms after the last; a pair not judged within 3 seconds counts as missed. The 40 items come in four quarters, and four lists drawn from the session's seed give each quarter one kind of trial, so across sessions every word is seen in every kind. On an unrelated trial the word appears with the related word of another item of its quarter. Which key means related is drawn from the seed. Participants read the word to themselves at each showing or, as a setting, say it aloud; nothing is recorded. Every pair was written for this library in each of the eight languages, and the words follow the session's language.
The main score, the satiation slowing, is the median time to judge a related pair after many showings minus after a few, over right answers at or after 100 ms. The task also reports the same slowing for unrelated pairs, which tells satiation from general slowing, the advantage of related pairs after a few and after many showings and its change (Balota and Black's measure), accuracy, and the error rates after a few and after many showings. One person's slowing rests on 10 related pairs after each number of showings, so it is a research measure best compared between groups. Scores support research and do not diagnose.
Scores
- Satiation Slowingms
- Median time to judge a related pair after many showings (20 to 24) minus after a few (2 to 4), over right answers.
- Positive values mean the repeated word's meaning came more slowly; not better or worse
- Unrelated Slowingms
- The same difference for unrelated pairs.
- Slowing here too points to tiredness or waiting rather than satiation
- Relatedness Effect, Few Showingsms
- Median time for unrelated pairs minus related pairs after a few showings, over right answers.
- Positive values mean related pairs were judged faster
- Relatedness Effect, Many Showingsms
- Median time for unrelated pairs minus related pairs after many showings, over right answers.
- Positive values mean related pairs were judged faster
- Change in Relatedness Effectms
- The relatedness effect after a few showings minus after many, Balota and Black's measure; it equals the satiation slowing minus the unrelated slowing.
- Positive values mean the related pairs' advantage shrank after many showings
- Accuracy%
- Right answers as a share of right and wrong ones, over all test pairs.
- Below 0.75 is invalid
- Errors After Few Showings%
- Wrong answers as a share of right and wrong ones, after a few showings.
- Lower means fewer errors
- Errors After Many Showings%
- Wrong answers as a share of right and wrong ones, after many showings.
- Lower means fewer errors
- Missed Pairscount
- Test pairs not judged within 3 seconds.
- Expected to be 0
- Anticipationscount
- Answers under 100 ms after the pair appeared, counted neither right nor wrong.
- Expected to be 0
Use and evidence
28 papers from 2021 to 2025 name this task in their title or abstract, up 40% compared with 2016 to 2020. That ranks it 367 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 1 |
| 2024 | 5 |
| 2023 | 11 |
| 2022 | 5 |
| 2021 | 6 |
| 2020 | 6 |
| 2019 | 2 |
| 2018 | 3 |
| 2017 | 5 |
| 2016 | 4 |
| 2015 | 0 |
| 2014 | 5 |
| 2013 | 10 |
| 2012 | 7 |
| 2011 | 3 |
| 2010 | 2 |
| 2009 | 1 |
| 2008 | 2 |
| 2007 | 3 |
| 2006 | 4 |
| 2005 | 1 |
| 2004 | 3 |
| 2003 | 2 |
| 2002 | 0 |
| 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
- 0
- 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.
- Pairs
- 40 pairs of related words written for this test in each of the eight languages, in four quarters of 10, and 4 for practice. On an unrelated trial a word appears with the related word of another item of its quarter. The words follow the session's language.
- Showings
- A few (2 to 4) or many (20 to 24), the number drawn for each trial. Each showing lasts 500 ms with 200 ms between, and the pair follows 300 ms after the last.
- Lists
- Four lists drawn from the session's seed give each quarter one kind of trial: related or unrelated, after a few or many showings, 10 of each.
- Trial
- A cross for 500 ms, the showings, then the word above the second word until F or J or 3,000 ms. Which key means related is drawn from the seed.
- Session
- 4 practice trials with feedback, then 40 with none and a pause after 20, with no more than 3 of one answer or one kind of showing in a row.
- Scoring
- The satiation slowing is the median time to judge a related pair after many showings minus after a few, over right answers at or after 100 ms. More than 25% wrong judgments is invalid.
The pairs were written for this test in each language, not taken from published item sets. Read the slowing beside the unrelated slowing to tell satiation from tiredness: Smith (1984) found the slowing for related pairs only. One person's score rests on 10 related pairs after each number of showings, 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: semantic_satiation_test from the scored run, F or J for related or unrelated within 3 seconds (chance 50%, half the pairs related). Below chance invalidates, not above chance is flagged, and one key on 90% of answers invalidates when accuracy is not above chance. Missed responses and responses under 100 ms follow the timed-test rule. More than 25% wrong judgments invalidates.
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 pairs are plainly related or unrelated, so many wrong judgments mean the task was not followed: Tian and Huber (2010) set aside participants below 90% correct and report the same result with a 75% cutoff. These limits are ConductScience choices.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Read the word aloud | Off | 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
- 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 Semantic Satiation Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Smith, 1984; Balota & Black, 1997), with the standard settings (read the word aloud: Off). The main score was satiation slowing (median time to judge a related pair after many showings (20 to 24) minus after a few (2 to 4), 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.
- Smith, L. C. (1984). Semantic satiation affects category membership decision time but not lexical priming. Memory & Cognition, 12(5), 483-488. https://doi.org/10.3758/BF03198310The paradigm: a category name said 3 or 30 times, then a slower decision after 30 that a word belongs to it.
- Balota, D. A., & Black, S. (1997). Semantic satiation in healthy young and older adults. Memory & Cognition, 25(2), 190-202. https://doi.org/10.3758/BF03201112Relatedness judgments after up to 24 showings, spoken or only watched; the related pairs' advantage fell by 46 ms in young adults, and not when the repeated word had nothing to do with the pair.
- Kounios, J., Kotz, S. A., & Holcomb, P. J. (2000). On the locus of the semantic satiation effect: Evidence from event-related brain potentials. Memory & Cognition, 28(8), 1366-1377. https://doi.org/10.3758/BF03211837Brain responses (the N400) in which many showings shrank the difference between related and unrelated words, evidence that satiation can reach meaning itself, not only perception.
- Tian, X., & Huber, D. E. (2010). Testing an associative account of semantic satiation. Cognitive Psychology, 60(4), 267-290. https://doi.org/10.1016/j.cogpsych.2010.01.003Satiation measured within one speeded matching task, with participants below 90% correct set aside and the same result at a 75% cutoff.
Source papers
- Smith LC (1984). Semantic satiation affects category membership decision time but not lexical priming. Memory & Cognition, 12(5):483-488.Paradigm source, Adults
- Balota DA, Black S (1997). Semantic satiation in healthy young and older adults. Memory & Cognition, 25(2):190-202.Paradigm source, Young and older adults
- No normative data for this task.Paradigm source, N/A
Questions
How long does the Semantic Satiation Task take?
About 8 to 10 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?
Satiation Slowing, Unrelated Slowing, Relatedness Effect, Few Showings, Relatedness Effect, Many Showings, Change in Relatedness Effect, Accuracy, Errors After Few Showings, Errors After Many Showings, Missed Pairs and Anticipations. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Semantic Satiation Task used in research?
28 papers from 2021 to 2025 name it in their title or abstract, up 40% compared with 2016 to 2020. That ranks it 367 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 Semantic Satiation 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?
Read the word aloud (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, 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.