Working Memory
Hebb Repetition Task
Also called: Hebb repetition paradigm, Hebb digits task, Hebb repetition effect, Hebb effect, repeated list learning
Measure learning of a repeated sequence: lists of nine digits recalled in order, with every third list the same list, unannounced.
9-12 minUntimed
Free to start: 10 free participants per account, then $0.99 per participant for academic research, or any yearly plan. Preview any test free after you sign up.
- Papers, 2021 to 2025
- 29
- Evidence grade
- C
- Registered clinical studies
- 0
What it measures
Measures learning of a repeated sequence through immediate recall. Participants see lists of digits, nine as standard, one digit at a time and enter each list in order; every third list is the same list, unannounced.
Constructs
- Sequence learning
- Short-term memory
- Serial order
- Incidental learning
Research uses
- Research on word learning and language disorders such as dyslexia
- Sequence learning studies across development and aging
- Within-study comparison of learning through repetition
How the paradigm works
The Hebb repetition task measures how a sequence is learned through repetition during immediate recall. Lists of nine digits appear one at a time at the center of the screen, one per second (each digit for 800 ms, then a 200 ms blank). After each list the participant enters the digits in the order shown, by typing or on an on-screen keypad on touch screens. Every list is a new order of the same digits, except that every third list is the same list each time. Participants are not told.
More about the paradigm
The task follows Hebb (1961), who found that a nine-digit list repeated on every third trial was recalled better and better, even by participants who did not notice the repetition (Page and Norris 2009). Here there are eight cycles of two new lists followed by the repeated list, 24 lists in all (standard; six to ten cycles), as in Araya, Oberauer and Saito (2022), after two practice lists with feedback. A fixation cross shows for 1,000 ms before each list. Lists of seven or eight digits are a setting for groups with shorter spans; a session then uses the same digits, drawn once, on every list.
Every list is drawn from the session's seed. No list holds a run of three digits counting up or down by one, and no new list shares a run of three neighbors in the same order with the repeated list, because repeating part of an order can itself be learned (Hitch, Fastame and Flude 2005).
A digit counts when it is entered in the place it was shown. The main score is the Hebb gain: the share of digits in place on the repeated list minus that on the new lists, over all cycles. The slope of each list type across cycles and their difference, the share in place overall, lists with every digit in place, and the answer to a closing question on whether any list came back are also given. Melton (1963) found that learning weakens when the repeated list comes back less often. Bogaerts and colleagues (2018) found that individual learning scores in their letter tasks barely held from one session to another about six months later, while overall recall did. Scores support research and do not diagnose.
Scores
- Hebb Gain%
- Share of digits entered in place on the repeated list minus the share on the new lists, over all cycles.
- Positive values mean the repeated list was recalled better
- Repeated List Recall%
- Share of digits entered in the place they were shown on the repeated list.
- Higher is better
- New List Recall%
- Share of digits entered in the place they were shown on the new lists.
- Under 0.2 or at 0.95 and above is flagged for review
- Learning Slope Gainscore
- Slope of the share in place on cycle for the repeated list minus that for the new lists.
- Positive values mean the repeated list improved faster
- Repeated List Slopescore
- Least-squares slope of the repeated list's share in place on cycle number: the change from one cycle to the next.
- Informational
- New List Slopescore
- Least-squares slope of the new lists' share in place on cycle number, which shows general practice.
- Informational
- Digits in Place%
- Share of all digits entered in the place they were shown, over every list.
- Higher is better
- Lists Fully Recalledcount
- Lists with every digit entered in its place.
- Higher is better
- Noticed a Repeat0/1
- 1 when the participant answered that a list came back more than once, 0 when not.
- The gain can appear without awareness
- Median Entry Timems
- Median time from the recall screen to Submit.
- Informational
- Lists Completedtrials
- Test lists recorded in the scored run.
- Under 90% of the planned lists is invalid
- Planned Liststrials
- Test lists the session planned to give.
- Informational
Use and evidence
29 papers from 2021 to 2025 name this task in their title or abstract, down 17% compared with 2016 to 2020. That ranks it 304 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 9 |
| 2024 | 9 |
| 2023 | 3 |
| 2022 | 3 |
| 2021 | 5 |
| 2020 | 5 |
| 2019 | 7 |
| 2018 | 8 |
| 2017 | 9 |
| 2016 | 6 |
| 2015 | 7 |
| 2014 | 6 |
| 2013 | 3 |
| 2012 | 3 |
| 2011 | 3 |
| 2010 | 3 |
| 2009 | 4 |
| 2008 | 4 |
| 2007 | 2 |
| 2006 | 3 |
| 2005 | 0 |
| 2004 | 0 |
| 2003 | 0 |
| 2002 | 0 |
| 2001 | 0 |
| 2000 | 0 |
Use grade E
Fewer than 30 papers from 2021 to 2025.
Evidence grade C
At least 2 test-retest papers, 30 reliability or validity papers or 3 registered clinical studies.
- Papers on test-retest reliability
- 4
- Papers on reliability or validity
- 4
- 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 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Lists
- Nine digits, 1 to 9 in a new order on each list (standard; seven or eight as a setting, the same digits on every list of a session).
- Repetition
- Cycles of three lists: two new lists, then the repeated list, the same on every cycle and never announced. Eight cycles, 24 lists in all (standard; six to ten).
- List rules
- No run of three digits counting up or down by one, and no new list shares a run of three neighbors in the same order with the repeated list.
- Trial
- A fixation cross for 1,000 ms, then each digit for 800 ms with a 200 ms blank, then recall.
- Recall
- The digits in order, typed, or tapped on a keypad of 1 to 9 on touch screens, with Delete and Submit. Untimed.
- Practice
- Two new lists, each followed by the digits shown, the digits entered and how many were in place.
- Awareness
- After the last list, one question: did any list come back more than once?
- Score
- The Hebb gain: digits in place on the repeated list minus on the new lists. Learning slopes, overall recall and the awareness answer.
Every list is drawn by rule from the session's seed; no published list is used. Percentiles are not computed, because the main score is a difference between list types and individual learning scores have shown test-retest reliability close to zero.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: test-phase recall entries of the scored run, one per list. A whole list in order has no meaningful chance level, so accuracy is not judged against chance; recall is untimed, so nothing can be missed. Fewer than 90% of the planned lists invalidates. New lists recalled under 20% in place, or at 95% or more, flag the session and withhold the percentile, because the gain cannot show at the floor or the ceiling. A hidden tab withholds the percentile.
Why these rules
The Hebb effect is a gain of the repeated list over new lists, so it needs both list types away from the floor and the ceiling. Entering a list's digits in a random order puts 1 in 7 to 1 in 9 of them in place, by list length, and a span well above the list length leaves no room to improve. 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 |
|---|---|---|
| Digits in each list | 9 | 7 to 9 |
| Showings of the repeated list | 8 | 6 to 10 |
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 Hebb Repetition Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Hebb, 1961; Page and Norris, 2009), with the standard settings (digits in each list: 9; showings of the repeated list: 8). The main score was hebb gain (share of digits entered in place on the repeated list minus the share on the new lists, over all cycles).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Hebb, D. O. (1961). Distinctive features of learning in the higher animal. In J. F. Delafresnaye (Ed.), Brain mechanisms and learning (pp. 37-46). Blackwell. The method: nine-digit lists for immediate recall, with every third list the same.
- Page, M. P. A., & Norris, D. (2009). A model linking immediate serial recall, the Hebb repetition effect and the learning of phonological word forms. Philosophical Transactions of the Royal Society B, 364(1536), 3737-3753. https://doi.org/10.1098/rstb.2009.0173The Hebb effect as a model of learning new word forms, and its appearance without awareness.
- Araya, C., Oberauer, K., & Saito, S. (2022). The Hebb repetition effect in complex span tasks: Evidence for a shared learning mechanism with simple span tasks. Memory & Cognition, 50(5), 925-940. https://doi.org/10.3758/s13421-021-01261-3Eight cycles of two new lists and the repeated list, scored by items in their place.
- Bogaerts, L., Siegelman, N., Ben-Porat, T., & Frost, R. (2018). Is the Hebb repetition task a reliable measure of individual differences in sequence learning? Quarterly Journal of Experimental Psychology, 71(4), 892-905. https://doi.org/10.1080/17470218.2017.1307432Low test-retest reliability of individual learning scores.
- Melton, A. W. (1963). Implications of short-term memory for a general theory of memory. Journal of Verbal Learning and Verbal Behavior, 2(1), 1-21. https://doi.org/10.1016/S0022-5371(63)80063-8Learning weakens when the repeated list comes back less often.
Source papers
- Hebb DO (1961). Distinctive features of learning in the higher animal. In: Delafresnaye JF (ed). Brain Mechanisms and Learning, pp 37-46. Oxford: Blackwell.Paradigm source, Adults
- Bogaerts L, Siegelman N, Ben-Porat T, Frost R (2018). Is the Hebb repetition task a reliable measure of individual differences in sequence learning? Quarterly Journal of Experimental Psychology, 71(4):892-905.Paradigm source, Adults
Questions
How long does the Hebb Repetition Task take?
About 9 to 12 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?
Hebb Gain, Repeated List Recall, New List Recall, Learning Slope Gain, Repeated List Slope, New List Slope, Digits in Place, Lists Fully Recalled, Noticed a Repeat, Median Entry Time, Lists Completed and Planned Lists. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Hebb Repetition Task used in research?
29 papers from 2021 to 2025 name it in their title or abstract, down 17% compared with 2016 to 2020. That ranks it 304 of 395 test families in the ConductCognition test census (use grade E). Source: ConductCognition test census 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Hebb Repetition Task have?
In the published research, 4 papers report its test-retest reliability and 0 registered clinical studies use it as an outcome (evidence grade C in the ConductCognition test census).
Does it report a percentile?
No. Scores and quality flags only.
Which settings can a researcher change?
Digits in each list (7 to 9) and showings of the repeated list (6 to 10). 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. 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.