Language
Lexical Tone Task
Also called: lexical tone discrimination task, Mandarin tone discrimination task, AX tone discrimination
Measure Mandarin tone perception with same-or-different judgments of spoken syllable pairs.
4-6 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
- 78
- Evidence grade
- C
- Registered clinical studies
- 4
What it measures
Measures how accurately listeners distinguish Mandarin speech tones. Participants hear two syllables from a synthetic voice and choose whether their tones are the same or different. The main score is accuracy across the four tones. Use headphones.
Constructs
- Lexical tone perception
- Speech sound discrimination
- Auditory comparison
Research uses
- Native and non-native speech perception research
- Language experience studies
- Within-study comparison
How the paradigm works
The Lexical Tone Task measures discrimination of the pitch patterns that distinguish Mandarin syllables. Listeners compare two sounds with the same consonant and vowel and choose same or different. This is an AX discrimination task, with no need to name the tones or know the words.
More about the paradigm
The speech kit generates our 20 Mandarin clips with one synthetic voice: four syllables in four tones for the test and a fifth syllable for practice. Each block has 48 pairs, half same and half different, with all six contrasts represented equally and both orders balanced. The session seed sets the sequence. Eight practice pairs give feedback; test pairs do not. The listener presses Play, hears both syllables 500 ms apart, then answers without a deadline. Same pairs replay one token; different pairs keep the natural duration and voice-quality cues of the synthesized tones.
The primary score is percent correct, with same-pair and different-pair accuracy, each tone contrast, and the share of different responses. Answer times are descriptive, not the primary measure. The task uses headphones and the shared headphone check.
Scores
- Tone Discrimination Accuracy%
- Correct same or different answers among answered test pairs. Chance is 50%.
- Higher is better
- Same-Tone Accuracy%
- Correct judgments among pairs with the same tone.
- Higher is better
- Different-Tone Accuracy%
- Correct judgments among pairs with different tones.
- Higher is better
- Accuracy by Tone Contrast%
- Percent correct for each unordered different-tone contrast, keyed 1-2 through 3-4. Unmeasured contrasts are absent.
- Higher is better
- Different Answers%
- Share of answered pairs called different.
- Describes response preference
- Median Correct RTms
- Median answer time after both sounds end, excluding answers under 100 ms.
- Context for the scores
- Quick Answerscount
- Answers under 100 ms, retained in accuracy.
- Half or more invalidates the session
- Answered Pairstrials
- Test pairs answered.
- Fewer than 24 invalidates the session
- Planned Pairstrials
- Number of planned test pairs.
- Informational
- Headphone Check Passed0/1
- 1 when the headphone check passed, 0 when it failed.
- Context for the scores
- Headphone Check Correctcount
- Correct headphone-check responses, out of six.
- Informational
- Stopped at Headphone Check0/1
- 1 if the failed headphone check stopped the test.
- Informational
- Reported Output Delayms
- Output delay reported by the browser.
- Context for the scores
Use and evidence
78 papers from 2021 to 2025 name this task in their title or abstract, down 26% compared with 2016 to 2020. That ranks it 281 of 582 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 13 |
| 2024 | 12 |
| 2023 | 19 |
| 2022 | 16 |
| 2021 | 18 |
| 2020 | 19 |
| 2019 | 14 |
| 2018 | 30 |
| 2017 | 25 |
| 2016 | 18 |
| 2015 | 26 |
| 2014 | 24 |
| 2013 | 19 |
| 2012 | 22 |
| 2011 | 12 |
| 2010 | 12 |
| 2009 | 14 |
| 2008 | 17 |
| 2007 | 11 |
| 2006 | 5 |
| 2005 | 7 |
| 2004 | 6 |
| 2003 | 7 |
| 2002 | 4 |
| 2001 | 7 |
| 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
- 1
- Papers on reliability or validity
- 8
- Registered clinical studies using it as an outcome
- 4
Measured on ConductCognition
Pooled across the studies that feed our platform norms, by device group. A figure needs 50 or more people.
No pooled figure yet.
Note on the counts: includes infant and second-language 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 0.39.0, counted 8 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Sounds
- 20 synthetic Mandarin clips: ma, ba, du and po in four tones for the test, and mi in four tones for practice. Our syllable selection uses the speech kit’s Chinese voice.
- Trials
- 48 pairs per block: 24 same and 24 different. Each of the 6 tone contrasts appears 4 times, twice in each order. The standard is 1 block; researchers can choose 1 to 3.
- Listening
- Play starts two clips at the same speech level, 500 ms apart. Same pairs repeat one clip. Different pairs use naturally synthesized tones, with their duration and voice-quality differences. Designed for participants with normal hearing. Headphones and the headphone check come first.
- Answers
- Same or different, after both clips end, with no response deadline. The seed sets the order.
- Practice
- 8 pairs with feedback for 1,000 ms each. The test gives no feedback.
- Scores
- Percent correct overall, on same and different pairs, and for each of the 6 contrasts. The share of different answers describes response preference; answer time is secondary.
The voices are synthetic. The task measures discrimination of these clips; it is not a hearing test or a language proficiency test. Mandarin experience matters, so compare participants with similar language backgrounds. No published norms or percentiles are applied.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scores the most complete run of same or different tone choices, with chance at 50%. Below-chance accuracy invalidates; accuracy not above chance is flagged. One answer on 90% of trials invalidates when performance is not above chance. Answers under 100 ms still count; half or more invalidates. Fewer than 24 answers invalidates; an unfinished run or a failed headphone check is flagged.
Why these rules
The two answers are balanced. Accuracy reflects tone discrimination rather than speed, so a choice prepared during the sounds still counts. Incomplete runs give unequal coverage of the contrasts. A failed headphone check changes the listening conditions.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Blocks | 1 | 1 to 3 |
| 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
- Untimed test: runs on desktops, laptops, tablets and phones. Needs headphones. Participant devices
- Instructions and screens: English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese. Spoken syllables: Mandarin Chinese.
- 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 Lexical Tone Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Gao et al., 2019), with the standard settings (blocks: 1; when the headphone check fails: Stop the test). The main score was tone discrimination accuracy (correct same or different answers among answered test pairs).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Gao, X., Yan, T.-T., Tang, D.-L., Huang, T., Shu, H., Nan, Y., & Zhang, Y.-X. (2019). What Makes Lexical Tone Special: A Reverse Accessing Model for Tonal Speech Perception. Frontiers in Psychology, 10, 2830. https://doi.org/10.3389/fpsyg.2019.02830Tone discrimination as a measure of speech perception; our clip set and untimed protocol are new.
Source papers
- Gao, X., Yan, T.-T., Tang, D.-L., Huang, T., Shu, H., Nan, Y., & Zhang, Y.-X. (2019). What Makes Lexical Tone Special: A Reverse Accessing Model for Tonal Speech Perception. Frontiers in Psychology, 10, 2830.Paradigm source, Adults
Questions
How long does the Lexical Tone Task take?
About 4 to 6 minutes, as listed in the test library.
What device does a participant need?
Untimed test: runs on desktops, laptops, tablets and phones. Needs headphones. The browser must be Chrome, Edge, Firefox or Safari.
What scores does it produce?
Tone Discrimination Accuracy, Same-Tone Accuracy, Different-Tone Accuracy, Accuracy by Tone Contrast, Different Answers, Median Correct RT, Quick Answers, Answered Pairs, Planned Pairs, Headphone Check Passed, Headphone Check Correct, Stopped at Headphone Check and Reported Output Delay. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Lexical Tone Task used in research?
78 papers from 2021 to 2025 name it in their title or abstract, down 26% compared with 2016 to 2020. That ranks it 281 of 582 test families in the ConductCognition test census (use grade D). Source: ConductCognition test census 0.39.0, counted 8 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Lexical Tone Task have?
In the published research, 1 paper reports its test-retest reliability and 4 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?
Blocks (1 to 3) 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?
The instructions and screens are in English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese. The syllables are Mandarin Chinese in every session. Participants judge the tones and do not need to know the words' meanings. 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.