Attention
Metronome Response Task
Also called: MRT, metronome task, thought probe task, rhythmic response time task, mind-wandering task
Measure how steadily people keep their attention on a task as participants press with a tone every 1.3 seconds and say, when it stops, where their mind was.
17-19 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
- 13
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
Measures how steadily people keep their attention on a simple task. Participants press the space bar in time with a tone that sounds every 1.3 seconds for about 13 minutes; now and then the tones stop and a question asks whether their mind was on the task. The main score is the variability of the presses' timing.
Constructs
- Sustained attention
- Mind wandering
- Attention lapses
- Rhythmic timing
Research uses
- Sustained attention research
- Mind-wandering research
- Developmental and aging research
- Within-study group comparison
How the paradigm works
The Metronome Response Task measures how steadily a person keeps attention on a simple, continuous task (Seli, Cheyne and Smilek, 2013). A short tone sounds every 1,300 ms, and the participant presses the space bar exactly when each tone sounds; each tone's trial is the 650 ms before it and the 650 ms after it. The test plays only tones made by the browser, and it needs headphones and normal hearing. Before it starts, the participant sets a comfortable volume and passes a headphone check (Woods, Siegel, Traer and McDermott, 2017).
More about the paradigm
The standard session is 600 trials, about 13 minutes, the length of the browser version of Seli, Carriere, Levene and Smilek (2013); a study can choose 300 to 900, 900 being the laboratory version's. In each block of 50 trials a thought probe stops the tones at a place drawn from the session's seed, at least 5 trials after the one before, and asks where the participant's mind was just before: on the task, or on something else. Only thoughts unrelated to the task count as something else. The participant answers, then starts the tones again. An 18-trial practice comes first and says only whether the presses came with the tones, or too late or too early, as the browser version's practice did.
The rhythmic response time (RRT) of a trial is its first press minus its tone, against the tone as the device reports it heard. The main score is the variability of the RRTs: the natural log of their variance in a window of five trials, the current and the four before it, averaged over the test, leaving out the first five trials and the five after each probe (Seli, Carriere, Levene and Smilek, 2013). Lower is steadier. The same measure over the five trials before probes answered on the task and off the task, and the share of probes answered off the task, follow it. The mean RRT says whether the presses came before or after the tones. The probe answers are the participant's own reports. Scores support research and do not diagnose.
Scores
- RRT Variabilityscore
- Natural log of the variance (ms squared) of press minus tone in five-trial windows, averaged over the test, leaving out the first five trials and the five after each probe; a window needs presses on 3 of its 5 trials.
- Lower is steadier
- RRT Variability Before On-Task Reportsscore
- The same over the five trials before each probe answered on the task, averaged.
- Lower is steadier
- RRT Variability Before Off-Task Reportsscore
- The same over the five trials before each probe answered on something else, averaged.
- Usually higher than before on-task reports
- Off-Task Reports%
- Share of answered probes answered on something else.
- The participant's own report
- Mean RRTms
- Mean of press minus tone over the trials the windows use, against each tone as the device reports it heard.
- Negative means the presses came before the tones
- Missed Tones%
- Share of trials with no press.
- Above 0.1 makes the session invalid
- Extra Pressescount
- Presses that answered no trial: a second press in a trial, or one outside every trial.
- Informational
- Anticipationscount
- Presses less than 100 ms after the press before them, left out of the scores.
- More than max(2, 20%) of presses makes the session invalid
- Windows Scoredcount
- Five-trial windows with enough presses.
- Informational
- Probes Answeredcount
- Thought probes the participant answered.
- Informational
- Runs with Focus Losscount
- Runs of tones during which the page lost focus, left out of the timing scores.
- Above 0 is flagged for review
- Output Delayms
- Median delay between the browser playing a run's first tone and the device reporting it heard.
- Set by the device; wireless headphones can add delay the device does not report
- Change in Reported Delayms
- Largest change in the output latency the device reported, over the sound setup and the start and end of every run of tones.
- Above 20 is flagged and withholds the percentile
- Output Latencyms
- Output latency the device reported in the sound setup.
- Informational
- Base Latencyms
- Base latency the device reported in the sound setup.
- Informational
- Trialscount
- Trials played in the scored run.
- Informational
- Planned Trialscount
- Trials the settings planned.
- Informational
- Headphone Check Passedflag
- 1 when the headphone check before the test passed, 0 when it did not and the study's setting ran the test anyway.
- 0 is flagged for review
Use and evidence
13 papers from 2021 to 2025 name this task in their title or abstract, up 30% compared with 2016 to 2020. That ranks it 435 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 4 |
| 2024 | 3 |
| 2023 | 4 |
| 2022 | 0 |
| 2021 | 2 |
| 2020 | 2 |
| 2019 | 2 |
| 2018 | 3 |
| 2017 | 3 |
| 2016 | 0 |
| 2015 | 0 |
| 2014 | 2 |
| 2013 | 3 |
| 2012 | 1 |
| 2011 | 0 |
| 2010 | 0 |
| 2009 | 0 |
| 2008 | 0 |
| 2007 | 0 |
| 2006 | 0 |
| 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 D
Little published measurement evidence found.
- Papers on test-retest reliability
- 0
- Papers on reliability or validity
- 8
- 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 2 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Before the test
- A volume step on a steady 1 kHz tone, then the headphone check of Woods and colleagues (2017): 6 rounds of three 200 Hz tones, passed with at least 5 right. A failed check ends the test unless the study chooses to continue.
- Trial
- A 1 kHz tone of 75 ms every 1,300 ms; each trial is the 650 ms before its tone and the 650 ms after it. The participant presses the space bar with each tone; the screen does not change while the tones play.
- Thought probes
- One in each block of 50 trials, at a place drawn from the session's seed, at least 5 trials after the one before. The tones stop and a question asks where the participant's mind was just before, on the task or on something else; the participant then starts the tones again.
- Trials
- 12 blocks (standard, 6 to 18): 600 trials and 12 probes, about 13 minutes of tones. First, a practice of 18 trials that says only whether the presses came with the tones.
- Timing
- Each tone is started on the audio clock, on a sample. Each press is timed by its key event and each tone by when the device reports it heard; the delays the device reports are saved in the sound setup and at the start and end of every run of tones.
- Score
- A trial's rhythmic response time is its first press minus its tone; a press under 100 ms after the one before is left out. The main score is the natural log of their variance in windows of 5 trials, averaged over the test, leaving out the first 5 trials and the 5 after each probe.
The test plays tones and needs headphones and normal hearing. Wired headphones are best: wireless headphones play sound late. A delay that stays the same leaves the variability as it is, but one that changes during the session adds to it, so a session whose reported delay changed by more than 20 ms gets no percentile. The probe answers are the participants' own reports. 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 responses: every press of the space bar in the runs of tones (metronome_response_segment) of the scored run. Pressing with a tone has no correct option, so accuracy is not judged against chance. Each press is timed from the press before it, and responses under 100 ms follow the timed-test rule. More than 10% of trials with no press invalidates. A reported output latency that changed by more than 20 ms withholds the percentile; a run during which the page lost focus, no window with enough presses, and a failed headphone check under the continue setting are flagged. A session the headphone check stopped has no results.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. Seli, Cheyne and Smilek (2013) left out participants who missed more than 10% of the tones, and the browser version kept that rule. A second press within 100 ms cannot be a press to a tone 1,300 ms from the next, so presses that fast follow the platform's anticipation rule. A delay that stays the same moves every press against its tone alike and leaves their variability as it is, but a delay that changes during the session adds variability that is not the person's, so a percentile is withheld when the delay the device reports changes. Presses are timed on the page, so a hidden tab or another window can lose presses; a percentile is withheld after one. Woods, Siegel, Traer and McDermott (2017) let only listeners who passed the headphone check go on, because a listener on loudspeakers hears the sounds differently; a session that ran after a failed check keeps its scores but not its percentile.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Blocks | 12 | 6 to 18 |
| When the headphone check fails | Stop the test | Stop the test, Continue the test |
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.
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Metronome Response Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Seli, Cheyne and Smilek, 2013), with the standard settings (blocks: 12; when the headphone check fails: Stop the test). The main score was RRT variability (natural log of the variance (ms squared) of press minus tone in five-trial windows, averaged over the test, leaving out the first five trials and the five after each probe; a window needs presses on 3 of its 5 trials).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Seli, P., Cheyne, J. A., & Smilek, D. (2013). Wandering minds and wavering rhythms: Linking mind wandering and behavioral variability. Journal of Experimental Psychology: Human Perception and Performance, 39(1), 1-5. https://doi.org/10.1037/a0030954The task: pressing with a metronome, thought probes, and the timing variability before them.
- Seli, P., Carriere, J. S. A., Levene, M., & Smilek, D. (2013). How few and far between? Examining the effects of probe rate on self-reported mind wandering. Frontiers in Psychology, 4, 430. https://doi.org/10.3389/fpsyg.2013.00430The task in a web browser, the practice and its advice, the five-trial windows and the limit on missed tones.
- Anderson, T., Petranker, R., Lin, H., & Farb, N. A. S. (2021). The metronome response task for measuring mind wandering: Replication attempt and extension of three studies by Seli et al. Attention, Perception, & Psychophysics, 83(1), 315-330. https://doi.org/10.3758/s13414-020-02131-xThe trial's timing and one probe in each block of 50 trials, in a multisite replication.
- 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.
Source papers
- Seli, Cheyne and Smilek (2013): the paradigm. Seli, Carriere, Levene and Smilek (2013): the browser version and the measures. Anderson, Petranker, Lin and Farb (2021): the trial and the probes in a multisite replication.Paradigm source, Adults
Questions
How long does the Metronome Response Task take?
About 17 to 19 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?
RRT Variability, RRT Variability Before On-Task Reports, RRT Variability Before Off-Task Reports, Off-Task Reports, Mean RRT, Missed Tones, Extra Presses, Anticipations, Windows Scored, Probes Answered, Runs with Focus Loss, Output Delay, Change in Reported Delay, Output Latency, Base Latency, Trials, Planned Trials and Headphone Check Passed. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Metronome Response Task used in research?
13 papers from 2021 to 2025 name it in their title or abstract, up 30% compared with 2016 to 2020. That ranks it 435 of 495 test families in the ConductCognition test census (use grade E). Source: ConductCognition test census 0.22.0, counted 2 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Metronome Response 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).
Which settings can a researcher change?
Blocks (6 to 18) and when the headphone check fails (Stop the test, Continue the test). 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.