Language
Cross-Modal Priming Task
Also called: cross-modal lexical priming task, cross-modal semantic priming task, auditory-visual priming task
Measure how a spoken word speeds recognition of a related written word, with lexical decision targets timed from the word on the audio clock.
4-6 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
- 73
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
Measures how a spoken word speeds recognition of a related written word. Participants hear an English word, then decide as quickly and accurately as they can whether the letter string on the screen is an English word. Word targets follow a related or an unrelated spoken word. The voice is synthetic. Plays sound through speakers or headphones.
Constructs
- Spoken word recognition
- Semantic memory
- Visual word recognition
Research uses
- Language comprehension and semantic memory research
- Research on how listening activates word meanings
- Group comparisons of priming across conditions
How the paradigm works
The Cross-Modal Priming Task measures how hearing a word speeds recognition of a related written word. Each trial plays a spoken English word; when it ends, a string of lower-case letters appears, and the participant decides as quickly and accurately as possible whether it is an English word, pressing F or J. Word targets follow a related spoken word (bread, then butter) or an unrelated one; nonword targets follow other spoken words. The method is the cross-modal priming of Swinney (1979), who had people decide whether a written string was a word while they listened to sentences. Here, as in Swinney, Onifer, Prather and Hirshkowitz (1979), the spoken words come one at a time, each followed by its written target at the word's end; in their study, a related spoken word sped the decision by 83 ms.
More about the paradigm
The spoken words come from ConductCognition's speech kit: our own English words spoken by an open synthetic voice, all at one speech level. The 38 related pairs are the kit's words with a common associate as the written target; the kit's two words that its speech recognizer misheard are left out. Unrelated trials re-pair the spoken words of one half of the pairs with that half's other targets, and two lists swap which half is related, so across sessions every target follows both kinds of word and a session hears each word once. The nonwords are made from real words by changing letters, by the rules of the Lexical Decision and Semantic Priming tasks. A session has 6 practice trials with feedback and 72 test trials: 19 related, 19 unrelated and 34 nonword targets, with no more than three of one kind in a row. Which key means word is counterbalanced across sessions and shared with the Lexical Decision and Semantic Priming tasks.
The written string appears at the end of the spoken word, or 1,000 ms later as a setting. Its time is counted from the word's first sound on the browser's audio clock, and it is drawn on the screen frame that reaches the screen then. Each trial records the planned and the measured time from the word's first sound to the string; a session whose strings came more than 20 ms from their planned time on average is flagged. Delays the browser does not report, such as most wireless headphones add, are not measured.
The primary score is the priming effect: mean correct response time after unrelated spoken words minus after related ones. Positive values mean a related spoken word sped recognition of the written word. Accuracy for each kind of target and the nonword response time are reported too. The words are English in every session; sessions run in other languages are flagged for review. Ten of the pairs are pairs of the Semantic Priming Task as well, seven in the same order, so a session that runs both tests repeats them. Cross-modal priming is used in research on aphasia (Howells and Cardell, 2015) and schizophrenia (Surguladze and colleagues, 2002), among others; scores support research and do not diagnose.
Scores
- Priming Effectms
- Unrelated minus related mean correct response time to word targets.
- Positive values mean a related spoken word sped recognition
- Related RTms
- Mean correct response time to word targets after a related spoken word.
- Lower is faster
- Unrelated RTms
- Mean correct response time to word targets after an unrelated spoken word.
- Lower is faster
- Related Accuracy%
- Share of word targets after a related spoken word answered as words.
- Higher is better
- Unrelated Accuracy%
- Share of word targets after an unrelated spoken word answered as words.
- Higher is better
- Nonword Accuracy%
- Share of nonword targets answered as not words.
- Higher is better
- Nonword RTms
- Mean correct response time to nonword targets.
- Lower is faster
- Accuracy%
- Share of all test targets answered right.
- Higher is better
- Anticipationscount
- Responses under 100 ms after the written string.
- Lower supports interpretation
- Missed Responsescount
- Test trials with no response within 2,000 ms.
- Lower supports interpretation
- Test Trialstrials
- Test trials recorded.
- Informational
- Delay After the Wordms
- The setting: time from the end of the spoken word to the written string.
- Informational
- Mean Time From Word to Stringms
- Mean measured time from the spoken word's first sound to the written string, from the audio clock and the screen frames.
- Informational
- Timing Errorms
- Mean difference, either way, between the measured and the planned time from the word's first sound to the string.
- Above 20 ms is flagged for review
- Session in Another Language0/1
- 1 when the session ran in a language other than English, the language of the words.
- 1 is flagged for review
- Reported Output Delayms
- The sound output delay the browser reported for the device, which the timing allows for.
- Context for the scores
Use and evidence
73 papers from 2021 to 2025 name this task in their title or abstract, down 32% compared with 2016 to 2020. That ranks it 253 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 20 |
| 2024 | 10 |
| 2023 | 12 |
| 2022 | 12 |
| 2021 | 19 |
| 2020 | 19 |
| 2019 | 24 |
| 2018 | 20 |
| 2017 | 25 |
| 2016 | 19 |
| 2015 | 26 |
| 2014 | 22 |
| 2013 | 24 |
| 2012 | 22 |
| 2011 | 32 |
| 2010 | 17 |
| 2009 | 31 |
| 2008 | 16 |
| 2007 | 22 |
| 2006 | 15 |
| 2005 | 11 |
| 2004 | 9 |
| 2003 | 11 |
| 2002 | 19 |
| 2001 | 15 |
| 2000 | 11 |
Use grade D
30 to 99 papers from 2021 to 2025.
Evidence grade D
Little published measurement evidence found.
- Papers on test-retest reliability
- 0
- Papers on reliability or validity
- 7
- 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.
- Trial
- A 500 ms cross, a spoken word, then the written string at the end of the word (standard) or 1,000 ms after it, until a response or 2,000 ms.
- Timing
- The string's time is counted from the word's first sound on the audio clock, and it is drawn on the screen frame that shows it then. Each trial saves the planned and the measured time.
- Items
- 72 targets: 19 after a related spoken word, 19 after an unrelated one and 34 nonwords.
- Lists
- Two lists swap which targets follow related words; each session's list comes from its seed.
- Voice
- Our own English words in an open synthetic voice, all at one speech level, through speakers or headphones.
- Keys
- F and J for word and nonword, counterbalanced across sessions.
- Practice
- 6 trials with feedback.
The voice is synthetic, and the words are English; sessions run in other languages are flagged for review. Ten pairs are also Semantic Priming Task pairs, so a session running both repeats them. 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 rows: cross_modal_priming_test from the scored run, F or J for a word or nonword target (chance 50%; 38 of the 72 targets are words) within 2,000 ms, correctness from the key for the target. 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 mean timing error of the written string above 20 ms, or a session run in a language other than English, 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. Priming depends on when the written string comes after the spoken word, so a device that could not show it at its planned time changes the measure; the time is measured from the audio clock and the screen frames as the browser reports them. The word lists are English. A session run in another language shows English words to a participant who may not read English fluently, so it is flagged for review.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Time from the end of the word to the letters | At the end of the word (standard) | At the end of the word (standard), 1,000 ms after the end of the word |
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
- Instructions and screens: English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese. Words and sentences: English.
- 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 Cross-Modal Priming Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Swinney, 1979; Swinney, Onifer, Prather & Hirshkowitz, 1979), with the standard settings (time from the end of the word to the letters: At the end of the word). The main score was priming effect (unrelated minus related mean correct response time to word targets).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Swinney, D. A. (1979). Lexical access during sentence comprehension: (Re)consideration of context effects. Journal of Verbal Learning and Verbal Behavior, 18(6), 645-659. https://doi.org/10.1016/S0022-5371(79)90355-4
- Swinney, D. A., Onifer, W., Prather, P., & Hirshkowitz, M. (1979). Semantic facilitation across sensory modalities in the processing of individual words and sentences. Memory & Cognition, 7(3), 159-165. https://doi.org/10.3758/BF03197534Single spoken words, each followed by a written associate at the word's end.
- Surguladze, S., Rossell, S., Rabe-Hesketh, S., & David, A. S. (2002). Cross-modal semantic priming in schizophrenia. Journal of the International Neuropsychological Society, 8(7), 884-892. https://doi.org/10.1017/S1355617702870023Use in schizophrenia research.
- Howells, S. R., & Cardell, E. A. (2015). Semantic priming in anomic aphasia: A focused investigation using cross-modal methodology. Aphasiology, 29(6), 744-761. https://doi.org/10.1080/02687038.2014.985184Use in aphasia research.
Source papers
- Swinney DA (1979). Lexical access during sentence comprehension: (Re)consideration of context effects. Journal of Verbal Learning and Verbal Behavior, 18(6):645-659.Paradigm source, Adults
- Swinney DA, Onifer W, Prather P, Hirshkowitz M (1979). Semantic facilitation across sensory modalities in the processing of individual words and sentences. Memory & Cognition, 7(3):159-165.Paradigm source, Adults
Questions
How long does the Cross-Modal Priming Task take?
About 4 to 6 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?
Priming Effect, Related RT, Unrelated RT, Related Accuracy, Unrelated Accuracy, Nonword Accuracy, Nonword RT, Accuracy, Anticipations, Missed Responses, Test Trials, Delay After the Word, Mean Time From Word to String, Timing Error, Session in Another Language and Reported Output Delay. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Cross-Modal Priming Task used in research?
73 papers from 2021 to 2025 name it in their title or abstract, down 32% compared with 2016 to 2020. That ranks it 253 of 495 test families in the ConductCognition test census (use grade D). Source: ConductCognition test census 0.22.0, counted 2 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Cross-Modal Priming 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).
Does it report a percentile?
No. Scores and quality flags only.
Which settings can a researcher change?
Time from the end of the word to the letters (At the end of the word (standard), 1,000 ms after the end of the word). 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 words and sentences stay in English, so sessions in other languages are flagged for review. 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.