Learning
Alternating Serial Reaction Time Task
Also called: ASRT, probabilistic sequence learning task, triplet learning task
Measure implicit statistical learning: participants press the key for each circle that fills in while a hidden pattern alternates with random trials.
9-12 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. Preview any test free after you sign up.
- Papers, 2021 to 2025
- 23
- Evidence grade
- C
- Registered clinical studies
- 1
What it measures
Measures implicit statistical learning. One of four circles fills in and participants press its key. Every other circle follows a hidden repeating pattern and the rest are random, so some runs of three come up more often than others; faster responses to the frequent runs show learning.
Constructs
- Implicit statistical learning
- Procedural learning
- Visuomotor speed
Research uses
- Procedural and statistical learning research
- Aging and development research
- Within-study group comparison
How the paradigm works
The alternating serial reaction time task (ASRT) measures implicit statistical learning: picking up a regularity nobody points out. Four circles sit in a row above the keys D, F, J and K. One circle at a time fills in, and participants press the key under it. The circle stays filled until the right key is pressed, and the next one fills in 120 ms later.
More about the paradigm
Pattern trials and random trials alternate. Each session follows one of four orders of the four places, for example 1-3-2-4, so the trials run 1, random, 3, random, 2, random, 4, random and so on, where each random place is any of the four. Participants are not told about the pattern. Each block has 5 random warm-up trials and then the eight-trial cycle ten times, 85 trials in all, with a short rest and a display of accuracy and speed after each block. The standard session is 8 blocks after 16 practice trials; researchers can set 5 to 25 blocks.
Because of the alternation, some runs of three trials come up more often than others: a run whose first and third places follow each other in the pattern makes up 62.5% of trials, whatever comes between them. People respond faster to the last trial of these high-frequency triplets as they learn. The main score is triplet learning: the median correct response time on low-frequency minus high-frequency triplets, computed in each block and averaged over blocks. Repetitions (1-1-1), trills (1-2-1) and the first seven trials of each block are left out, as in the source studies. Triplet learning is also given for the first and second half of the blocks and in accuracy, and the pattern advantage compares high-frequency triplets that end on a pattern trial with those that end on a random trial (Song, Howard & Howard 2007). Responses under 100 ms count as anticipations.
Scores
- Triplet Learningms
- Median correct response time on low- minus high-frequency triplets, in each block, averaged over the blocks.
- Higher positive values suggest learning
- High-Frequency Triplet RTms
- Median correct response time on trials that end a high-frequency triplet, averaged over the blocks scored.
- Lower is faster
- Low-Frequency Triplet RTms
- Median correct response time on trials that end a low-frequency triplet, averaged over the blocks scored.
- Lower is faster
- Triplet Learning, First Halfms
- Triplet learning averaged over the first half of the blocks.
- Higher positive values suggest learning
- Triplet Learning, Second Halfms
- Triplet learning averaged over the second half of the blocks.
- Higher positive values suggest learning
- Pattern Advantagems
- Median correct response time on high-frequency triplets that end on a random trial minus those that end on a pattern trial, in each block, averaged.
- Higher positive values suggest knowledge of the pattern itself
- Triplet Learning, Accuracy%
- Accuracy on high-frequency minus low-frequency triplets.
- Higher positive values suggest learning
- High-Frequency Triplet Accuracy%
- Share of high-frequency triplets whose first key press was right.
- Higher is better
- Low-Frequency Triplet Accuracy%
- Share of low-frequency triplets whose first key press was right.
- Higher is better
- General Speed-Upms
- Median correct response time on scored trials in the first block minus the last block, whatever the triplet.
- Higher positive values mean faster responses by the end
- Accuracy%
- Share of targets whose first key press was right, presses under 100 ms aside.
- Higher is better
- Median Correct Response Timems
- Median response time of first presses that were right, on all test trials.
- Lower is faster
- Blocks Runcount
- Blocks in the scored run.
- Informational
- Blocks Scoredcount
- Blocks with at least five correct trials of each triplet kind, the blocks triplet learning averages.
- Fewer than half the planned blocks is flagged
- Planned Blockscount
- Blocks the session was set to run.
- Informational
- High-Frequency Trials Usedtrials
- Correct high-frequency triplet trials at or after 100 ms.
- Informational
- Low-Frequency Trials Usedtrials
- Correct low-frequency triplet trials at or after 100 ms.
- Informational
- Anticipationscount
- First presses under 100 ms from the circle filling in.
- Lower is better
- Total Trialstrials
- Targets in the scored run.
- Informational
Use and evidence
23 papers from 2021 to 2025 name this task in their title or abstract, down 21% compared with 2016 to 2020. That ranks it 320 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 4 |
| 2024 | 3 |
| 2023 | 2 |
| 2022 | 8 |
| 2021 | 6 |
| 2020 | 4 |
| 2019 | 3 |
| 2018 | 7 |
| 2017 | 8 |
| 2016 | 7 |
| 2015 | 1 |
| 2014 | 7 |
| 2013 | 3 |
| 2012 | 0 |
| 2011 | 6 |
| 2010 | 8 |
| 2009 | 4 |
| 2008 | 1 |
| 2007 | 2 |
| 2006 | 0 |
| 2005 | 0 |
| 2004 | 0 |
| 2003 | 2 |
| 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
- 2
- Papers on reliability or validity
- 10
- Registered clinical studies using it as an outcome
- 1
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.
- Display
- Four outlined circles in a row, with the keys D, F, J and K under them. The target circle fills in white.
- Timing
- The circle stays filled until the right key is pressed, and the next one fills in 120 ms later.
- Pattern
- Pattern and random trials alternate: 1, random, 3, random, 2, random, 4, random for the pattern 1-3-2-4. Each session draws one of four patterns: 1-2-4-3, 1-3-2-4, 1-3-4-2 or 1-4-2-3. Participants are not told about it.
- Blocks
- 8 blocks (standard; 5 to 25) of 85 trials: 5 random warm-up trials, then the eight-trial cycle ten times. A short rest after each block shows accuracy and speed.
- Practice
- 16 random trials; a wrong key brings a reminder.
- Score
- Low- minus high-frequency triplet median correct response time, in each block, averaged over the blocks. Repetitions, trills and the first 7 trials of each block are left out.
The source studies ran 20 to 45 blocks, often over several sessions. The standard 8 blocks keep the test near 10 minutes; Farkas and colleagues (2024) found that it takes about 25 blocks for the score's internal consistency to reach .65.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: alternating_srt_test from the scored run, the first key pressed for each target (D, F, J or K; chance 25%); the target stays until the right key, so no response is missed. Below chance invalidates and not above chance is flagged. One key on 90% of first presses invalidates when accuracy is not above chance. Responses under 100 ms follow the timed-test rule. Fewer than half the planned blocks with a triplet learning value is flagged and withholds the percentile.
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 triplet learning score averages one difference per block, so it needs enough blocks with five correct trials of each triplet kind; that limit and the half of the blocks are ConductScience choices.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Number of blocks | 8 | 5 to 25 |
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.
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Alternating Serial Reaction Time task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Howard & Howard, 1997; Song, Howard & Howard, 2007), with the standard settings (number of blocks: 8). The main score was triplet learning (median correct response time on low- minus high-frequency triplets, in each block, averaged over the blocks).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Howard, J. H., Jr., & Howard, D. V. (1997). Age differences in implicit learning of higher order dependencies in serial patterns. Psychology and Aging, 12(4), 634-656. https://doi.org/10.1037/0882-7974.12.4.634
- Song, S., Howard, J. H., Jr., & Howard, D. V. (2007). Implicit probabilistic sequence learning is independent of explicit awareness. Learning & Memory, 14(3), 167-176. https://doi.org/10.1101/lm.437407The display, the 120 ms interval, triplet scores by block and the pattern advantage (their event type effect).
- Howard, D. V., Howard, J. H., Jr., Japikse, K., DiYanni, C., Thompson, A., & Somberg, R. (2004). Implicit sequence learning: Effects of level of structure, adult age, and extended practice. Psychology and Aging, 19(1), 79-92. https://doi.org/10.1037/0882-7974.19.1.79Leaving out repetitions and trills.
- Janacsek, K., Fiser, J., & Nemeth, D. (2012). The best time to acquire new skills: Age-related differences in implicit sequence learning across the human lifespan. Developmental Science, 15(4), 496-505. https://doi.org/10.1111/j.1467-7687.2012.01150.xBlocks of 85 trials and triplet learning from childhood to old age.
- Farkas, B. C., Krajcsi, A., Janacsek, K., & Nemeth, D. (2024). The complexity of measuring reliability in learning tasks: An illustration using the Alternating Serial Reaction Time Task. Behavior Research Methods, 56(1), 301-317. https://doi.org/10.3758/s13428-022-02038-5Reliability of the triplet learning score by number of blocks.
Source papers
- Howard JH Jr, Howard DV (1997). Age differences in implicit learning of higher order dependencies in serial patterns. Psychology and Aging, 12(4):634-656.Paradigm source, Adults
- Janacsek K, Fiser J, Nemeth D (2012). The best time to acquire new skills: age-related differences in implicit sequence learning across the human lifespan. Developmental Science, 15(4):496-505.Paradigm source, 4 to 85 years
- Farkas BC, Krajcsi A, Janacsek K, Nemeth D (2024). The complexity of measuring reliability in learning tasks: an illustration using the Alternating Serial Reaction Time Task. Behavior Research Methods, 56(1):301-317.Paradigm source, Young adults
Questions
How long does the Alternating Serial Reaction Time Task take?
About 9 to 12 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?
Triplet Learning, High-Frequency Triplet RT, Low-Frequency Triplet RT, Triplet Learning, First Half, Triplet Learning, Second Half, Pattern Advantage, Triplet Learning, Accuracy, High-Frequency Triplet Accuracy, Low-Frequency Triplet Accuracy, General Speed-Up, Accuracy, Median Correct Response Time, Blocks Run, Blocks Scored, Planned Blocks, High-Frequency Trials Used, Low-Frequency Trials Used, Anticipations and Total Trials. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Alternating Serial Reaction Time Task used in research?
23 papers from 2021 to 2025 name it in their title or abstract, down 21% compared with 2016 to 2020. That ranks it 320 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 Alternating Serial Reaction Time Task have?
In the published research, 2 papers report its test-retest reliability and 1 registered clinical study uses it as an outcome (evidence grade C in the ConductCognition test census).
Which settings can a researcher change?
Number of blocks (5 to 25). 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.