Episodic Memory
Encoding Specificity Task
Also called: encoding specificity principle, cue-dependent recall, weak and strong cues, context-dependent memory
Study English words beside cue words, then recall them from the cue you studied, a related word you never saw, or no cue. Measures how much a cue helps when it matches learning.
7-9 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
- 86
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
Measures encoding specificity. Participants study words in capitals, each beside a related cue word, then recall them from the cue they studied, from a related word they never saw, or with no cue. The main score is how much more the studied cue helps. The words are in English.
Constructs
- Encoding specificity
- Cued recall
- Retrieval cues
- Episodic memory
Research uses
- Memory and learning research
- Studies of retrieval cues and context
- Aging and memory research
How the paradigm works
The Encoding Specificity Task measures how far recalling a word depends on the cue that was there when it was learned. Thomson and Tulving (1970) showed people English words in capitals, each beside a weakly related cue word in small letters, and tested recall with that cue, with a strong associate never shown, or with no cue. A weak cue helped only when it had been studied with its word, and a strong associate given only at test helped far less: what is stored is the word as it was encoded, and a cue works when it matches that encoding. They called this the encoding specificity hypothesis.
More about the paradigm
In this version, the participant studies English words in capitals, each shown for 3 seconds beside a cue word in small letters on its left, as in Thomson and Tulving's study: soak, BATHTUB. Each target has two cues of its own, each related to it in a different way (soak and porcelain for bathtub), and the session's seed decides which one is shown at study. After 30 seconds of true-or-false arithmetic, a third of the targets are cued with the word studied beside them and a third with their other cue, never seen before, one cue at a time with the two kinds mixed and 12 seconds for each answer; the participant types the word in capitals that goes with each cue. Last, the participant types the other words in capitals they remember, with no cue, for up to 60 seconds. The standard list has 24 targets, 8 in each condition, as Thomson and Tulving's lists had 24; a setting allows 6 to 10 in each. The words were written for this task in English, and each session gives every target and cue its role at random.
The main score is the cue match advantage: the share of targets recalled with their studied cue minus the share recalled with their other cue. Each cue is the studied one in some sessions and the new one in others, so the two kinds of cue are on average equally related to their targets, and the advantage measures what a cue gains by having been there at study. Recall with no cue, each cued condition's difference from it, answers that named another target and practice accuracy are also reported. In Thomson and Tulving's third experiment, with weak and strong cues mixed in one list, 73% of targets were recalled with their studied weak cue and 33% with a strong associate seen only at test. Newman and Frith (1977) found strong associates helped as much as the studied weak cues, so this version uses a second weak cue, not a strong associate, as the new cue. One person's advantage over 8 targets of each kind is noisy, so it is best used to compare groups. The words are English in every session, and sessions run in another language are flagged. Scores support research and do not diagnose.
Scores
- Cue Match Advantage%
- Recall with the studied cue minus recall with the target's other cue, never seen at study.
- Positive means the cue helped more when it was there at study
- Recall, Studied Cue%
- Share of targets typed for the cue studied beside them.
- Higher is better
- Recall, New Cue%
- Share of targets typed for their other cue, never seen at study.
- Higher is better
- Recall, No Cue%
- Share of the uncued targets typed in the final recall, or earlier as an answer to another cue.
- Higher is better
- Studied Cue over No Cue%
- Recall with the studied cue minus recall with no cue. The final recall comes last.
- Positive means the studied cue helped
- New Cue over No Cue%
- Recall with the new cue minus recall with no cue. The final recall comes last.
- Context for the advantage
- Other Targets Typedcount
- Cued answers that named another target of the list.
- Context for the scores
- Wrong Answerscount
- Cued answers that named no target.
- Context for the scores
- Blank Answerscount
- Cues answered with an empty box.
- Context for the scores
- Timeoutscount
- Cues not answered within the time limit.
- Context for the scores
- Final Recall Entriescount
- Entries typed in the final recall.
- Context for the scores
- Final Recall Intrusionscount
- Final recall entries that named no target.
- Lower is better
- Practice Accuracy%
- Share of the practice cues answered right.
- Shows the task was understood
- Cues Showncount
- Cues shown in the test.
- Context for the scores
- Cues Plannedcount
- Cues the test planned to show.
- Context for the scores
- Anticipationscount
- Cued answers sent under 100 ms.
- Lower is better
- Paste Attemptscount
- Blocked attempts to paste text into an answer box.
- Should be 0
- Delayseconds
- Seconds of arithmetic between the list and the test.
- Context for the scores
Use and evidence
86 papers from 2021 to 2025 name this task in their title or abstract, up 46% compared with 2016 to 2020. That ranks it 233 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 28 |
| 2024 | 17 |
| 2023 | 15 |
| 2022 | 11 |
| 2021 | 15 |
| 2020 | 17 |
| 2019 | 14 |
| 2018 | 9 |
| 2017 | 10 |
| 2016 | 9 |
| 2015 | 18 |
| 2014 | 9 |
| 2013 | 9 |
| 2012 | 4 |
| 2011 | 10 |
| 2010 | 16 |
| 2009 | 6 |
| 2008 | 10 |
| 2007 | 6 |
| 2006 | 12 |
| 2005 | 7 |
| 2004 | 2 |
| 2003 | 11 |
| 2002 | 12 |
| 2001 | 6 |
| 2000 | 2 |
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
- 10
- Registered clinical studies using it as an outcome
- 0
Note on the counts: includes animal and virtual reality studies.
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.
- Study
- One list of 24 English words in capitals (standard; 18 to 30), each shown for 3 seconds (standard; 2 to 8) beside one of its two cue words in small letters, with a 250 ms gap after.
- Delay
- 30 seconds (standard; 0 to 120) of true-or-false arithmetic.
- Cued test
- 16 cues, one at a time: for 8 targets the cue studied beside them and for 8 their other cue, never seen, mixed with no more than 3 of a kind in a row. Each word in capitals is typed within 12 seconds (standard; 6 to 30).
- Final recall
- The other 8 targets, with no cue: the participant types the words in capitals remembered and not yet typed, for up to 60 seconds (standard; 30 to 180), and can finish sooner with Done.
- Words
- 30 targets and their cue words, written for this task in English. Each target has two cues, related to it in different ways, and the session's seed decides which one is studied, so across sessions each cue serves as the studied one and the new one. The words are English in every session.
- Scores
- The cue match advantage: recall with the studied cue minus recall with the new cue. Recall with each kind of cue and with none, each cued condition against no cue, and answers that named another target.
- Answer
- Typed with autocorrect off and submitted with Enter or a button. A plural counts as the word, and one spelling slip is accepted in a word of 4 or more letters unless the typed word is another English word.
- Practice
- 3 pairs of their own, studied and then cued, 2 with the studied cue and 1 with a new one, each followed by feedback.
Thomson and Tulving (1970) found that a cue word helps recall of a studied word when it was studied with it, and that a related word given only at test helps far less. Here the new cue is a second related word rather than a strong associate, because studies disagree about strong associates (Newman and Frith, 1977). The words were written for this task; no published list or association norm is used. One person's score is noisy, so compare groups. 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: encoding_specificity_cued from the scored run, one typed word per cue within its time limit, with the final recall and practice rows of the same run. Typed recall has no chance level, so accuracy is not judged against chance. Fewer than 90% of the planned cues shown invalidates, and so does a word typed for fewer than 20% of the cues. Typed answers under 200 ms on at least half of 4 or more answers invalidate. A blocked paste or a session run in another language is flagged for review. A hidden tab withholds the percentile.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. Forgetting never invalidates, but with almost nothing typed a low score cannot be told from no effort, and a score over a much shorter test cannot be compared with a full one. A blocked paste can mean answers were being looked up. 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 word lists are English. A session run in another language shows English words to a participant who may not read English fluently, so it is flagged for review. 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 |
|---|---|---|
| Words per condition | 8 | 6 to 10 |
| Study time per pair (seconds) | 3 | 2 to 8 |
| Time per cue (seconds) | 12 | 6 to 30 |
| Delay before the test (seconds) | 30 | 0 to 120 |
| Final recall time (seconds) | 60 | 30 to 180 |
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
- Instructions and screens: English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese. Words and sentences: English.
- 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 Encoding Specificity Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Thomson & Tulving, 1970), with the standard settings (words per condition: 8; study time per pair (seconds): 3; time per cue (seconds): 12; delay before the test (seconds): 30; final recall time (seconds): 60). The main score was cue match advantage (recall with the studied cue minus recall with the target's other cue, never seen at study).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Thomson, D. M., & Tulving, E. (1970). Associative encoding and retrieval: Weak and strong cues. Journal of Experimental Psychology, 86(2), 255-262. https://doi.org/10.1037/h0029997The paradigm: weak cues at study, then recall with the same cue, a strong associate or no cue, in separate and mixed lists.
- Tulving, E., & Osler, S. (1968). Effectiveness of retrieval cues in memory for words. Journal of Experimental Psychology, 77(4), 593-601. https://doi.org/10.1037/h0026069Cues present or absent at study.
- Newman, S. E., & Frith, U. (1977). Encoding specificity vs associative continuity. Bulletin of the Psychonomic Society, 10(1), 73-75. https://doi.org/10.3758/BF03333551A replication in which strong associates helped as much as studied weak cues, the reason this version's new cue is a second weak cue.
Source papers
- Thomson DM, Tulving E (1970). Associative encoding and retrieval: Weak and strong cues. Journal of Experimental Psychology, 86(2):255-262.Paradigm source
- Tulving E, Osler S (1968). Effectiveness of retrieval cues in memory for words. Journal of Experimental Psychology, 77(4):593-601.Paradigm source
Questions
How long does the Encoding Specificity Task take?
About 7 to 9 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?
Cue Match Advantage, Recall, Studied Cue, Recall, New Cue, Recall, No Cue, Studied Cue over No Cue, New Cue over No Cue, Other Targets Typed, Wrong Answers, Blank Answers, Timeouts, Final Recall Entries, Final Recall Intrusions, Practice Accuracy, Cues Shown, Cues Planned, Anticipations, Paste Attempts and Delay. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Encoding Specificity Task used in research?
86 papers from 2021 to 2025 name it in their title or abstract, up 46% compared with 2016 to 2020. That ranks it 233 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 Encoding Specificity 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?
Words per condition (6 to 10), study time per pair (seconds) (2 to 8), time per cue (seconds) (6 to 30), delay before the test (seconds) (0 to 120) and final recall time (seconds) (30 to 180). Settings can be changed on yearly plans; free and pay-as-you-go studies use the standard settings.
Which languages can participants use?
The instructions and screens are in English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese. The words and sentences stay in English, so sessions in other languages are flagged for review. 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.