Attention
Auditory Streaming Task
Also called: auditory stream segregation, ABA triplets, ABA- paradigm, streaming paradigm, van Noorden paradigm, galloping rhythm task, auditory bistability
Measure auditory scene analysis: listeners report whether repeating low-high-low tone triplets sound like one galloping stream or two separate streams.
9-11 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
- 70
- Evidence grade
- C
- Registered clinical studies
- 2
What it measures
A measure of auditory scene analysis. Participants hear repeating low-high-low tone triplets and press an arrow key whenever what they hear changes between one galloping stream and two separate streams, high and low. Scores are the share of time heard as two streams at each frequency separation and tempo, and how it builds up over each sequence. A higher share is not a better score.
Constructs
- Auditory scene analysis
- Auditory stream segregation
- Perceptual bistability
Research uses
- Auditory perception and attention research
- Aging and auditory processing research, at the group level
- Research on perceptual organization across hearing and vision
How the paradigm works
The Auditory Streaming Task measures auditory scene analysis: how hearing groups a run of sounds into one source or splits it into two. A low tone and a high tone repeat in triplets, low, high, low and a silent slot (van Noorden, 1975). With a small frequency separation the triplets are heard as one stream with a galloping rhythm; with a large separation, or at a faster tempo, they split into two streams, a high one and a low one, each with a steady beat (Bregman and Campbell, 1971; van Noorden, 1975). In between, the percept can go either way and changes from moment to moment, as with a reversible figure (Pressnitzer and Hupé, 2006), and the chance of hearing two streams grows over the first seconds of a sequence, the build-up of streaming (Bregman, 1978; Anstis and Saida, 1985).
More about the paradigm
The participant presses the left arrow key for one stream and the right arrow key for two streams as soon as the tones start and at every change, as in the continuous reports of Micheyl and colleagues (2005), Pressnitzer and Hupé (2006) and Deike and colleagues (2012). The tones are 40 ms pure tones, one every 100 ms (fast) or every 200 ms (slow), 1, 4, 7 or 10 semitones apart around a centre near 800 Hz that changes from sequence to sequence. The standard session has 16 sequences of 16 seconds, each separation at each tempo twice, in a seeded order. Two examples, two practice sequences with feedback and a closing check use tones that nearly everyone hears one way: 1 semitone apart at the slow tempo (one stream) and 15 apart at the fast tempo (two streams). The sequence length (12 to 24 seconds) and the sequences of each kind (1 to 3) are settings. The test starts with the sound kit's loudness step and headphone check (Woods and colleagues, 2017); a failed check ends the test unless the study chooses to continue. About 9 to 11 minutes on a desktop or laptop, with headphones. The test needs normal hearing.
The main score is the share of the reported time heard as two streams. The share at each separation and tempo, the build-up from the first to the last quarter of the sequences, the time to the first report of two streams, the switch rate and the closing check's agreement are also reported. Only time after a sequence's first report counts, so a slow first answer is not taken as one stream (Deike and colleagues, 2012). Scores are research measures of auditory perception; they do not measure hearing and do not diagnose.
Scores
- Two-Stream Share%
- Share of the reported time heard as two streams over the test sequences, counted from each sequence's first report.
- Neither higher nor lower is better
- Two Streams at 1 Semitone%
- Share of the reported time heard as two streams with the tones 1 semitone apart, both tempos.
- Usually near 0
- Two Streams at 4 Semitones%
- The same with the tones 4 semitones apart.
- Informational
- Two Streams at 7 Semitones%
- The same with the tones 7 semitones apart.
- Informational
- Two Streams at 10 Semitones%
- The same with the tones 10 semitones apart.
- Usually high
- Separation Effect%
- Two-stream share at 10 semitones minus at 1 semitone.
- Positive values mean wider separations were heard as two streams more often, as expected
- Two Streams at the Fast Tempo%
- Two-stream share with a tone every 100 ms, all separations.
- Informational
- Two Streams at the Slow Tempo%
- Two-stream share with a tone every 200 ms, all separations.
- Informational
- Tempo Effect%
- Two-stream share at the fast tempo minus at the slow tempo.
- Positive values mean the faster tempo was heard as two streams more often, as expected
- Two Streams Early%
- Two-stream share in the first quarter of each sequence.
- Informational
- Two Streams Late%
- Two-stream share in the last quarter of each sequence.
- Informational
- Build-Up%
- Two-stream share in the last quarter of each sequence minus in the first quarter.
- Positive values mean two streams became more likely during a sequence
- First Report Two Streams%
- Share of the test sequences whose first report was two streams.
- Informational
- Time to Two Streamsseconds
- Median time from a sequence's first tone to its first report of two streams, over the sequences with one.
- Informational
- Time to First Reportms
- Median time from a sequence's first tone to its first arrow key.
- Informational
- Switch Rateswitches/min
- Changes of report per minute of the test sequences.
- Neither higher nor lower is better
- Closing Check Agreement%
- Share of the closing check's reported time heard the way nearly everyone hears those tones: one stream at 1 semitone and the slow tempo, two at 15 semitones and the fast tempo.
- Under 50% is invalid and under 75% is flagged for review
- Sequences Without a Reportcount
- Test sequences with no arrow key.
- Any is flagged; more than max(3, 25%) is invalid
- Repeated Keyscount
- Presses of the key for the percept already reported, which change nothing.
- Informational
- Slipscount
- Reports reversed within 300 ms, which are dropped.
- Informational
- Anticipationscount
- Keys under 100 ms after a sequence's first tone. They are not taken as reports.
- Informational
- Completed Sequencestrials
- Test sequences in the scored run.
- Fewer than planned is flagged for review
- Planned Sequencestrials
- Test sequences the session planned to play.
- Informational
- Headphone Check Passed0/1
- 1 when the sound kit's headphone check passed (at least 5 of 6 right), 0 when it did not.
- 0 with no scores means the check ended the test; 0 with scores is flagged for review
Use and evidence
70 papers from 2021 to 2025 name this task in their title or abstract, down 12% compared with 2016 to 2020. That ranks it 222 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 18 |
| 2024 | 13 |
| 2023 | 16 |
| 2022 | 13 |
| 2021 | 10 |
| 2020 | 15 |
| 2019 | 14 |
| 2018 | 21 |
| 2017 | 13 |
| 2016 | 17 |
| 2015 | 20 |
| 2014 | 24 |
| 2013 | 23 |
| 2012 | 20 |
| 2011 | 14 |
| 2010 | 15 |
| 2009 | 17 |
| 2008 | 14 |
| 2007 | 12 |
| 2006 | 5 |
| 2005 | 5 |
| 2004 | 6 |
| 2003 | 5 |
| 2002 | 5 |
| 2001 | 11 |
| 2000 | 6 |
Use grade D
30 to 99 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
- 14
- Registered clinical studies using it as an outcome
- 2
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 2 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Tones
- Pure tones of 40 ms with 5 ms ramps: low, high, low, then a silent slot, repeating. One tone slot every 100 ms (fast) or every 200 ms (slow). Tones at -6 dB re the loudness tone.
- Separations
- 1, 4, 7 or 10 semitones between the low and high tones, around a centre drawn for each sequence within 3 semitones of 800 Hz.
- Sequences
- 16 sequences of 16 seconds (standard): each separation at each tempo twice, in a seeded order. 12 to 24 seconds and 1 to 3 sequences of each kind as settings. Each starts 1,000 ms after the start key.
- Reports
- The left arrow for one stream and the right arrow for two, at the start of each sequence and at every change. A key under 100 ms after the first tone is an anticipation, a key for the percept already reported changes nothing, and a report reversed within 300 ms is a slip and is dropped.
- Practice
- An example of each percept, then two 12-second practice sequences with feedback on what was reported: 1 semitone at the slow tempo, which nearly everyone hears as one stream, and 15 at the fast tempo, heard as two.
- Check
- A closing check of the same two kinds of sequence, without feedback. Under 50% of its reported time heard the expected way is invalid, and under 75% is flagged.
- Sound
- The loudness step and a headphone check come first; a failed check ends the test unless the study chooses to continue. Needs headphones and normal hearing.
- Score
- The share of the reported time heard as two streams, counted from each sequence's first report, with the share at each separation and tempo and the build-up from the first to the last quarter of each sequence.
van Noorden traced his boundaries by sweeping or adjusting the frequency; this version, like Pressnitzer and Hupé and Deike and colleagues, plays fixed separations and records continuous reports. All tones are made by our code. Scores are research measures and do not measure hearing.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: auditory_streaming_test from the scored run, one per tone sequence, and the closing check's two sequences. What a listener hears has no correct answer, so accuracy is not judged against chance. A sequence with no arrow key at all is a missed response: more than max(3, 25%) such sequences invalidate and any is flagged. In the closing check, one sequence nearly always sounds like one stream and one like two; under 50% of its reported time in those percepts invalidates and under 75% is flagged. A failed headphone check that stopped the test invalidates, and one the study chose to continue past is flagged. A run short of its planned sequences or a closing check with no reports is flagged.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. Each sequence asks for what is heard at its start, so a sequence with no key at all means the participant was not reporting; the timed-test rule for missed responses applies. Below van Noorden's fission boundary nearly everyone hears one stream, and far above his temporal coherence boundary nearly everyone hears two, so reports against both mean the keys were swapped or the task was not followed. Woods and colleagues (2017) let only listeners who passed their headphone check go on to the task. A hidden tab may leave the tones unheard and the keys unreported, so a percentile is withheld after one.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| When the headphone check fails | Stop the test | Stop the test, Continue the test |
| Sequence length (seconds) | 16 | 12 to 24 |
| Sequences of each kind | 2 | 1 to 3 |
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. Needs headphones. 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 Auditory Streaming Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; van Noorden, 1975; Bregman and Campbell, 1971), with the standard settings (when the headphone check fails: Stop the test; sequence length (seconds): 16; sequences of each kind: 2). The main score was two-stream share (share of the reported time heard as two streams over the test sequences, counted from each sequence's first report).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- van Noorden, L. P. A. S. (1975). Temporal coherence in the perception of tone sequences (Doctoral thesis). Eindhoven University of Technology. The triplets of low, high and low tones, the 40 ms tones and the fission and temporal coherence boundaries.
- Bregman, A. S., & Campbell, J. (1971). Primary auditory stream segregation and perception of order in rapid sequences of tones. Journal of Experimental Psychology, 89(2), 244-249. https://doi.org/10.1037/h0031163Rapid alternating high and low tones split into two streams.
- Bregman, A. S. (1978). Auditory streaming is cumulative. Journal of Experimental Psychology: Human Perception and Performance, 4(3), 380-387. https://doi.org/10.1037/0096-1523.4.3.380The tendency to hear two streams builds up over a sequence.
- Anstis, S., & Saida, S. (1985). Adaptation to auditory streaming of frequency-modulated tones. Journal of Experimental Psychology: Human Perception and Performance, 11(3), 257-271. https://doi.org/10.1037/0096-1523.11.3.257Streaming builds up over seconds of listening and adapts.
- Micheyl, C., Tian, B., Carlyon, R. P., & Rauschecker, J. P. (2005). Perceptual organization of tone sequences in the auditory cortex of awake macaques. Neuron, 48(1), 139-148. https://doi.org/10.1016/j.neuron.2005.08.039The build-up of two streams measured in listeners with repeating triplets.
- Pressnitzer, D., & Hupé, J.-M. (2006). Temporal dynamics of auditory and visual bistability reveal common principles of perceptual organization. Current Biology, 16(13), 1351-1357. https://doi.org/10.1016/j.cub.2006.05.054Streaming reported continuously as a bistable percept.
- Deike, S., Heil, P., Böckmann-Barthel, M., & Brechmann, A. (2012). The build-up of auditory stream segregation: A different perspective. Frontiers in Psychology, 3, 461. https://doi.org/10.3389/fpsyg.2012.00461Continuous reports with two buttons, and counting only the time after the first report.
- Snyder, J. S., Carter, O. L., Lee, S.-K., Hannon, E. E., & Alain, C. (2008). Effects of context on auditory stream segregation. Journal of Experimental Psychology: Human Perception and Performance, 34(4), 1007-1016. https://doi.org/10.1037/0096-1523.34.4.1007A wider separation in the previous sequence lowers streaming in the next, so each sequence saves the one before.
- Woods, K. J. P., Siegel, M. H., Traer, J., & McDermott, J. H. (2017). Headphone screening to facilitate web-based auditory experiments. Attention, Perception, & Psychophysics, 79(7), 2064-2072. https://doi.org/10.3758/s13414-017-1361-2The headphone check before the test.
Source papers
- van Noorden LPAS (1975). Temporal coherence in the perception of tone sequences. Doctoral thesis, Eindhoven University of Technology.N 3, Adults
- Deike S, Heil P, Böckmann-Barthel M, Brechmann A (2012). The build-up of auditory stream segregation: a different perspective. Frontiers in Psychology, 3:461.N 22, 19-38
Questions
How long does the Auditory Streaming Task take?
About 9 to 11 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. Needs headphones. The browser must be Chrome, Edge, Firefox or Safari.
What scores does it produce?
Two-Stream Share, Two Streams at 1 Semitone, Two Streams at 4 Semitones, Two Streams at 7 Semitones, Two Streams at 10 Semitones, Separation Effect, Two Streams at the Fast Tempo, Two Streams at the Slow Tempo, Tempo Effect, Two Streams Early, Two Streams Late, Build-Up, First Report Two Streams, Time to Two Streams, Time to First Report, Switch Rate, Closing Check Agreement, Sequences Without a Report, Repeated Keys, Slips, Anticipations, Completed Sequences, Planned Sequences and Headphone Check Passed. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Auditory Streaming Task used in research?
70 papers from 2021 to 2025 name it in their title or abstract, down 12% compared with 2016 to 2020. That ranks it 222 of 395 test families in the ConductCognition test census (use grade D). Source: ConductCognition test census 0.14.0, counted 2 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Auditory Streaming Task have?
In the published research, 4 papers report its test-retest reliability and 2 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?
When the headphone check fails (Stop the test, Continue the test), sequence length (seconds) (12 to 24) and sequences of each kind (1 to 3). 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.