Language
Token Commands Task
Also called: following commands with tokens, command comprehension task, written token task
Measure how well people understand written commands, from touching one token to moving tokens and following order and conditions, in the participant's own language.
6-9 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
- 99
- Evidence grade
- B
- Registered clinical studies
- 36
What it measures
Measures comprehension of written commands, after the Token Test of De Renzi and Vignolo. Participants see circles and squares in five colors and two sizes and carry out 26 commands in four levels, from "Touch the small blue circle" to commands that move tokens or turn on word order and conditions. The commands are written in the participant's language. Scores the share carried out correctly, overall and by level.
Constructs
- Comprehension of written commands
- Word order, conditions and spatial terms
- Following instructions with several parts
Research uses
- Language comprehension research
- Aphasia and acquired language disorder research
- Within-study group comparison
How the paradigm works
The Token Test of De Renzi and Vignolo (1962) measures how well a person understands commands that give no clues beyond their words. Tokens of two shapes, two sizes and five colors lie in rows, and spoken commands grow harder in five parts: touch one token named by color and shape, then by size as well, then two tokens, then commands that move tokens or turn on word order and conditions. A shortened form of 36 commands followed (De Renzi and Faglioni 1978), and computer versions now show the commands as text on a touch screen (McNeil and colleagues 2015).
More about the paradigm
This task follows that design with commands written for it and shown in the participant's language. The board holds circles and squares, large and small, in red, yellow, blue, white and black. 26 commands come in four levels: one token named by size, color and shape (6); two tokens (6); moving one token on, beside or to touch another (6); and order and conditions (8), with before, after, if and except. The first two levels show all 20 tokens and the last two only the 10 large ones. A click or a tap touches a token; pressing a token, sliding it and letting go moves it; Done ends each command, which has up to 45 seconds. The command stays on the screen, so the task needs reading but not memory for a command heard once. Three practice commands with feedback come first. The order of commands within each level and the order of colors along each row come from the session's random seed. Scores support research and do not diagnose.
The main score is the percentage of commands carried out correctly. In the computer version of McNeil and colleagues (2015), 30 people with aphasia and 30 healthy adults took a listening version and three reading versions twice. In aphasia, retest correlations ran from .89 to .97 and listening and reading scores correlated .79 to .85; the healthy adults scored too close to the maximum for high correlations. Because healthy adults make few errors, the times to the first action and to the last action are reported as well.
Scores
- Percent Correct%
- Test commands carried out correctly, of those shown; a command left undone when its time runs out counts as wrong.
- Higher means more commands understood and carried out
- Commands Correctcount
- Test commands carried out correctly.
- Higher is better; 26 commands at the standard settings
- One Token%
- Percent correct on the 6 commands that name one token by size, color and shape.
- Higher is better
- Two Tokens%
- Percent correct on the 6 commands that name two tokens.
- Higher is better
- Moving Tokens%
- Percent correct on the 6 commands that move one token on, beside or to touch another.
- Higher is better
- Order and Conditions%
- Percent correct on the 8 commands with before, after, if or except.
- Higher is better
- Median Time to First Actionms
- Median time from a command's onset to the first press, on commands carried out correctly, leaving out presses under 100 ms.
- Lower means faster reading and planning; compare within a device type
- Median Completion Timems
- Median time from a command's onset to its last action, on commands carried out correctly.
- Lower means faster; compare within a device type
- Commands Left Undonecount
- Commands whose time ran out with nothing done.
- Expected to be 0
- Commands Timed Outcount
- Commands that reached their time limit, done or not.
- Expected to be 0
- Presses Under 100 mscount
- Commands whose first press came within 100 ms of the command appearing.
- Expected to be 0
- Commands Recordedtrials
- Test commands in the scored run.
- Fewer than 90% of those set is invalid
- Commands Settrials
- Test commands the session set out to show.
- 26
Use and evidence
99 papers from 2021 to 2025 name this task in their title or abstract, down 14% compared with 2016 to 2020. That ranks it 188 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 22 |
| 2024 | 16 |
| 2023 | 13 |
| 2022 | 25 |
| 2021 | 23 |
| 2020 | 14 |
| 2019 | 26 |
| 2018 | 18 |
| 2017 | 38 |
| 2016 | 19 |
| 2015 | 33 |
| 2014 | 27 |
| 2013 | 27 |
| 2012 | 23 |
| 2011 | 18 |
| 2010 | 25 |
| 2009 | 23 |
| 2008 | 25 |
| 2007 | 20 |
| 2006 | 18 |
| 2005 | 17 |
| 2004 | 11 |
| 2003 | 9 |
| 2002 | 7 |
| 2001 | 11 |
| 2000 | 14 |
Use grade D
30 to 99 papers from 2021 to 2025.
Evidence grade B
At least 8 test-retest papers, and either 5 registered clinical studies or 150 reliability or validity papers.
- Papers on test-retest reliability
- 21
- Papers on reliability or validity
- 123
- Registered clinical studies using it as an outcome
- 36
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.
- Tokens
- 20 circles and squares, large and small, in red, yellow, blue, white and black, in rows on a gray board. Green names the token that is not there in the false condition.
- Commands
- 26 commands in four levels: one token named by size, color and shape (6), two tokens (6), moving one token on, beside or to touch another (6), and order and conditions with before, after, if and except (8). The first two levels show all 20 tokens, the last two the 10 large ones.
- Languages
- The commands are written in each of the eight participant languages, so the whole task runs in the participant's language.
- Answer
- A click or tap touches a token; pressing, sliding and letting go moves it. Done, which waits for the first action, ends each command, and each has up to 45 seconds.
- Session
- 3 practice commands with feedback, then the 26 commands with none. The order within each level and the order of colors in each row come from the session's seed.
- Scoring
- The percent of commands carried out correctly is the main score, with the percent in each level and the median times to the first and the last action.
The commands were written for this task from the design of De Renzi and Vignolo (1962); none is taken from a published Token Test. The original commands are spoken. Here they are written and stay on the screen, so the task also needs reading.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: token_commands_test from the scored run, each command judged from its stored touches and moves. A command can be carried out in many ways, so accuracy is not judged against chance. A command whose time ran out with nothing done is a missed response: more than max(3, 25%) invalidate and any is flagged. First presses under 200 ms on at least half the commands invalidate. A run with fewer than 90% of its planned commands is invalid.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. Every command needs at least one action, and Done waits for the first, so a command with nothing done when its time runs out signals non-engagement rather than a wrong answer. An answer under 200 ms is faster than a simple visual reaction, so it cannot reflect the item. Following response-time effort screening for untimed tests (Wise and Kong 2005), the session is invalid only when at least half the answers are that fast. A share of correct commands over a much shorter run cannot be compared with one over the full run.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Time per command (seconds) | 45 | 20 to 90 |
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.
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Token Commands Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; De Renzi & Vignolo, 1962; De Renzi & Faglioni, 1978), with the standard settings (time per command (seconds): 45). The main score was percent correct (test commands carried out correctly, of those shown; a command left undone when its time runs out counts as wrong).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- De Renzi, E., & Vignolo, L. A. (1962). The Token Test: A sensitive test to detect receptive disturbances in aphasics. Brain, 85(4), 665-678. https://doi.org/10.1093/brain/85.4.665
- De Renzi, E., & Faglioni, P. (1978). Normative data and screening power of a shortened version of the Token Test. Cortex, 14(1), 41-49. https://doi.org/10.1016/S0010-9452(78)80006-9
- McNeil, M. R., Pratt, S. R., Szuminsky, N., Sung, J. E., Fossett, T. R. D., Fassbinder, W., & Lim, K. Y. (2015). Reliability and validity of the Computerized Revised Token Test: Comparison of reading and listening versions in persons with and without aphasia. Journal of Speech, Language, and Hearing Research, 58(2), 311-324. https://doi.org/10.1044/2015_JSLHR-L-13-0030
Source papers
- De Renzi E, Vignolo LA (1962). The Token Test: A sensitive test to detect receptive disturbances in aphasics. Brain, 85(4):665-678.Paradigm source, Adults
- McNeil MR, Pratt SR, Szuminsky N, Sung JE, Fossett TRD, Fassbinder W, Lim KY (2015). Reliability and validity of the Computerized Revised Token Test: Comparison of reading and listening versions in persons with and without aphasia. Journal of Speech, Language, and Hearing Research, 58(2):311-324.Paradigm source, Adults
Questions
How long does the Token Commands Task take?
About 6 to 9 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?
Percent Correct, Commands Correct, One Token, Two Tokens, Moving Tokens, Order and Conditions, Median Time to First Action, Median Completion Time, Commands Left Undone, Commands Timed Out, Presses Under 100 ms, Commands Recorded and Commands Set. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Token Commands Task used in research?
99 papers from 2021 to 2025 name it in their title or abstract, down 14% compared with 2016 to 2020. That ranks it 188 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 Token Commands Task have?
In the published research, 21 papers report its test-retest reliability and 36 registered clinical studies use it as an outcome (evidence grade B in the ConductCognition test census).
Which settings can a researcher change?
Time per command (seconds) (20 to 90). 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, with its words in the same language. 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.