Executive Function
Numerical Stroop Task
Also called: size congruity task, size congruity effect, number Stroop, numerical size congruency, physical size comparison task, numerical Stroop paradigm
Measure how a digit's value and its print size interfere: how much slower people choose the larger number, or the bigger print, when the two disagree.
7-8 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
- 43
- Evidence grade
- C
- Registered clinical studies
- 9
What it measures
Measures interference between the value of a digit and its print size. Two digits from 1 to 9 appear side by side, often in different print sizes. In one part participants choose the larger number, and in the other the digit printed bigger. The main score is the size congruity effect: how much slower answers are when value and print size disagree than when they agree.
Constructs
- Interference control
- Automatic number processing
- Numerical magnitude
Research uses
- Cognitive control and conflict research
- Math learning and dyscalculia research
- Numerical cognition research
How the paradigm works
The Numerical Stroop Task measures how the value of a digit and its print size interfere with each other. Henik and Tzelgov (1982) showed two digits that differed in value and in print size and asked which was larger: by value in one session and by print size in the other. Answers were faster when the two agreed, such as a large 8 beside a small 3 (congruent), and slower when they disagreed, such as a small 8 beside a large 3 (incongruent). This is the size congruity effect. It appears even when only print size matters, which is read as a sign that a digit's value is processed automatically.
More about the paradigm
The task has two parts, in an order drawn for each session. In the value part participants press F or J for the side of the larger number; in the size part, for the side of the digit printed bigger. On each trial a cross shows for 500 ms, then two digits from 1 to 9 appear until a key or 2,500 ms. The digits are 1 or 5 apart, printed at 64 and 80 px. Neutral pairs differ only in what the part asks about: two values at the same size in the value part, one digit at two sizes in the size part. Each part has 8 practice trials with feedback, then 3 rounds (1 to 4) of 48 trials: 16 congruent, 16 neutral and 16 incongruent. That makes 288 test trials, about 7 to 8 minutes in all. The cross and the digits switch on screen frames, and response times run from the frame that drew the digits. Desktop and laptop only.
The main score is the size congruity effect: median correct response time on incongruent minus congruent pairs, averaged over the two parts. Each part's effect is reported too, split into facilitation (neutral minus congruent) and interference (incongruent minus neutral), with the same effects in errors. Responses under 100 ms count as anticipations.
Scores
- Size Congruity Effectms
- Median correct response time on incongruent minus congruent pairs, averaged over the value part and the size part, counting answers at or after 100 ms.
- Positive values mean answers were slower when value and print size disagreed
- Value Part Congruity Effectms
- Incongruent minus congruent median correct response time when choosing the larger number.
- Positive values mean print size slowed judgments of value
- Size Part Congruity Effectms
- Incongruent minus congruent median correct response time when choosing the digit printed bigger.
- Positive values mean the digits' values slowed judgments of size
- Value Part Facilitationms
- Neutral minus congruent median correct response time in the value part.
- Positive values mean agreeing print sizes sped answers
- Value Part Interferencems
- Incongruent minus neutral median correct response time in the value part.
- Positive values mean conflicting print sizes slowed answers
- Size Part Facilitationms
- Neutral minus congruent median correct response time in the size part.
- Positive values mean agreeing values sped answers
- Size Part Interferencems
- Incongruent minus neutral median correct response time in the size part.
- Positive values mean conflicting values slowed answers
- Error Congruity Effect%
- Share of errors on incongruent minus congruent trials, both parts together; a missed answer counts as an error.
- Positive values mean more errors when value and print size disagreed
- Value Part Error Effect%
- Share of errors on incongruent minus congruent trials in the value part.
- Positive values mean more errors when print size conflicted
- Size Part Error Effect%
- Share of errors on incongruent minus congruent trials in the size part.
- Positive values mean more errors when the values conflicted
- Median Correct Response Timems
- Median response time of correct answers over all test trials.
- Lower is faster
- Accuracy%
- Share of all test trials answered with the key on the side of the right digit.
- Higher is better
- Response Time by Conditionms
- Median correct response time for each part and condition, keyed numerical_congruent to physical_incongruent, where a condition has at least 10 correct answers.
- Informational
- Accuracy by Condition%
- Share answered correctly for each part and condition, keyed numerical_congruent to physical_incongruent.
- Informational
- Anticipationscount
- Answers under 100 ms from the digits' onset.
- Lower is better
- Missed Responsescount
- Trials with no key press within 2,500 ms.
- Lower is better
- Completed Trialstrials
- Trials shown in the scored run.
- Fewer than planned is flagged for review
- Planned Trialstrials
- Trials the run set out to show.
- Informational
Use and evidence
43 papers from 2021 to 2025 name this task in their title or abstract, down 12% compared with 2016 to 2020. That ranks it 264 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 9 |
| 2024 | 8 |
| 2023 | 10 |
| 2022 | 9 |
| 2021 | 7 |
| 2020 | 8 |
| 2019 | 7 |
| 2018 | 12 |
| 2017 | 14 |
| 2016 | 8 |
| 2015 | 9 |
| 2014 | 8 |
| 2013 | 5 |
| 2012 | 4 |
| 2011 | 4 |
| 2010 | 5 |
| 2009 | 6 |
| 2008 | 6 |
| 2007 | 1 |
| 2006 | 6 |
| 2005 | 2 |
| 2004 | 1 |
| 2003 | 1 |
| 2002 | 2 |
| 2001 | 2 |
| 2000 | 0 |
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
- 3
- Papers on reliability or validity
- 18
- Registered clinical studies using it as an outcome
- 9
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
- On a dark panel: a cross for 500 ms, then two digits from 1 to 9 side by side, 120 px left and right of the center, until a key or 2,500 ms.
- Parts
- A value part (F or J for the side of the larger number) and a size part (the side of the digit printed bigger), in an order drawn for each session (standard) or fixed in the settings. F is the left digit and J the right one.
- Pairs
- 8 pairs, 4 that are 1 apart and 4 that are 5 apart, printed at 64 and 80 px. Congruent: the larger number printed bigger. Incongruent: the larger number printed smaller. Neutral: two values at 72 px in the value part, one digit at 64 and 80 px in the size part.
- Trials
- 48 trials per part in each round: 16 congruent, 16 neutral and 16 incongruent, with the right answer on each side in half. 3 rounds (standard; 1 to 4): 144 trials per part, 288 in all, each round in its own seeded order.
- Practice
- 8 trials before each part, with feedback for 800 ms.
- Score
- The size congruity effect: median correct response time on incongruent minus congruent pairs, averaged over the two parts, counting answers at or after 100 ms. Each part's facilitation (neutral minus congruent), interference (incongruent minus neutral) and effect in errors are reported too.
Henik and Tzelgov (1982) ran the two judgments on different days, with feedback after every trial. This test runs both in one session, gives feedback only in practice and draws which part comes first, because the first judgment changes the second.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: numerical_stroop_test from the scored run, F or J within 2,500 ms (chance 50%; the right answer is on each side in half of every condition), graded from the saved digits, print sizes and part. Below chance invalidates and not above chance is flagged. One key on 90% of answers invalidates when accuracy is not above chance. Missed responses and responses under 100 ms follow the timed-test rule. A run short of its planned trials, too few correct answers to compute the size congruity effect, or fewer than half the incongruent trials of a part answered correctly is flagged for review.
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.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Rounds of each part's 48 trials | 3 | 1 to 4 |
| Order of the two parts | Drawn for each session (standard) | Drawn for each session (standard), Value part first, Size part first |
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.
- 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 Numerical Stroop Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Henik and Tzelgov, 1982; Besner and Coltheart, 1979), with the standard settings (rounds of each part's 48 trials: 3; order of the two parts: Drawn for each session). The main score was size congruity effect (median correct response time on incongruent minus congruent pairs, averaged over the value part and the size part, counting answers at or after 100 ms).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Henik, A., & Tzelgov, J. (1982). Is three greater than five: The relation between physical and semantic size in comparison tasks. Memory & Cognition, 10(4), 389-395. https://doi.org/10.3758/BF03202431Judging two digits by value or by print size, the congruent, neutral and incongruent pairs, and the size congruity effect.
- Besner, D., & Coltheart, M. (1979). Ideographic and alphabetic processing in skilled reading of English. Neuropsychologia, 17(5), 467-472. https://doi.org/10.1016/0028-3932(79)90053-8
- Szűcs, D., Soltész, F., Jármi, É., & Csépe, V. (2007). The speed of magnitude processing and executive functions in controlled and automatic number comparison in children: An electro-encephalography study. Behavioral and Brain Functions, 3, 23. https://doi.org/10.1186/1744-9081-3-23Digits 1 or 5 apart, a 1.25 size ratio, neutral pairs at a middle size and the two parts in counterbalanced order.
- Szűcs, D., & Soltész, F. (2007). Event-related potentials dissociate facilitation and interference effects in the numerical Stroop paradigm. Neuropsychologia, 45(14), 3190-3202. https://doi.org/10.1016/j.neuropsychologia.2007.06.013Facilitation and interference, each measured against neutral pairs.
- Rubinsten, O., & Henik, A. (2005). Automatic activation of internal magnitudes: A study of developmental dyscalculia. Neuropsychology, 19(5), 641-648. https://doi.org/10.1037/0894-4105.19.5.641A smaller congruity effect in size judgments in adults with developmental dyscalculia.
Source papers
- Henik A, Tzelgov J (1982). Is three greater than five: The relation between physical and semantic size in comparison tasks. Memory & Cognition, 10(4):389-395.N 26, University students
Questions
How long does the Numerical Stroop Task take?
About 7 to 8 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?
Size Congruity Effect, Value Part Congruity Effect, Size Part Congruity Effect, Value Part Facilitation, Value Part Interference, Size Part Facilitation, Size Part Interference, Error Congruity Effect, Value Part Error Effect, Size Part Error Effect, Median Correct Response Time, Accuracy, Response Time by Condition, Accuracy by Condition, Anticipations, Missed Responses, Completed Trials and Planned Trials. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Numerical Stroop Task used in research?
43 papers from 2021 to 2025 name it in their title or abstract, down 12% compared with 2016 to 2020. That ranks it 264 of 395 test families in the ConductCognition test census (use grade D). Source: ConductCognition test census 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Numerical Stroop Task have?
In the published research, 3 papers report its test-retest reliability and 9 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?
Rounds of each part's 48 trials (1 to 4) and order of the two parts (Drawn for each session (standard), Value part first, Size part first). 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.