Language
Verb Generation Task
Also called: verb generation paradigm, noun-verb generation task, generate a verb task, selection demand verb generation
Measure how quickly people choose a word when several fit: participants type a verb for each noun, and the score compares nouns with one strong verb against nouns with many.
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
- 61
- Evidence grade
- C
- Registered clinical studies
- 1
What it measures
Measures how quickly people choose a word when several could fit. Participants see an English noun and type a verb that goes with it (ball: throw). The main score is how much slower the answers are for the 20 nouns with many possible verbs than for the 20 nouns with one strong verb.
Constructs
- Controlled word retrieval
- Selection among competing words
- Semantic memory
- Language production
Research uses
- Language production research
- Research on choosing among competing responses
- Within-study group comparison
How the paradigm works
The Verb Generation Task measures how quickly people choose a word when the meaning of a cue brings more than one to mind. Each screen shows an English noun, and the participant types a verb that goes with it: something done with the thing, or something it does (ball: throw). The task comes from early brain imaging of single words (Petersen and colleagues, 1988). Thompson-Schill and colleagues (1997) varied how much choosing it takes: nouns with one dominant verb against nouns with several fitting verbs and none dominant, which drew more activity in the left inferior frontal gyrus. Martin and Cheng (2006) found that answers to nouns of the second kind are slower when their most frequent verb is weakly linked to the noun, and Snyder and Munakata (2008) found separate effects of association strength and of competition among verbs on response times.
More about the paradigm
There are 40 nouns, written for this test: 20 with one strong verb (pen: write) and 20 with many possible verbs (rope: tie, pull or climb), matched for word frequency and length. Each noun's set follows a stated rule, checked by a second reading; no people have rated these nouns yet, so a pilot should confirm them. Each trial shows a fixation cross for 600 to 1,000 ms and then the noun with a text box. The participant types the first fitting verb and presses Enter, within 10 seconds by default. Three practice nouns with feedback come first, and the order is drawn from the session's seed with no more than three nouns of one kind in a row.
Answers are scored by rule against lists of fitting verbs written for each noun. Any form of a listed verb counts (rang for ring), and so does one spelling slip in a verb of four or more letters, unless the slip spells another English word. Verbs that go with almost anything, such as use or get, count only where a noun's list has them. The primary score is the selection cost: the median time to the first key press on nouns with many verbs minus nouns with one strong verb, over accepted answers. The test also reports both medians, the percent of nouns given a fitting verb overall and for each set, how often the strong verb was given, and counts of the other kinds of answer. The nouns and the answers are English; sessions run in other languages are flagged for review.
Scores
- Selection Costms
- Median time to the first key press on nouns with many possible verbs minus nouns with one strong verb, over accepted answers.
- Larger means more slowing when several verbs fit; read across a group
- Median Time, Many Verbsms
- Median time from the noun to the first key press on accepted answers to nouns with many possible verbs, over presses at or after 100 ms. Needs 5 accepted answers.
- Context for the selection cost
- Median Time, One Strong Verbms
- The same median for nouns with one strong verb. Needs 5 accepted answers.
- Context for the selection cost
- Median Time to First Keyms
- The same median over every accepted answer.
- Context for the scores
- Fitting Verbs%
- Percent of nouns shown given a listed verb, in any form or with one spelling slip.
- Context for the scores
- Fitting Verbs, Many Verbs%
- The same percent for nouns with many possible verbs.
- Context for the scores
- Fitting Verbs, One Strong Verb%
- The same percent for nouns with one strong verb.
- Context for the scores
- Strong Verb Given%
- Percent of the nouns with one strong verb that were given that verb, such as write for pen.
- Shows whether the nouns worked as their set expects
- Generic Verbscount
- Answers that are verbs fitting almost anything, such as use or get, where the noun's list does not have them.
- A quarter of the planned nouns or more is flagged for review
- Other Answerscount
- Answers that are neither a listed verb, a generic verb nor the noun itself.
- Context for the scores
- No Answerscount
- Nouns with nothing typed before the time ran out, or a phrase saying no verb came.
- Answers on fewer than half the planned nouns invalidate the session
- Noun Repeatedcount
- Answers that repeat the noun itself.
- Context for the scores
- Spelling Slipscount
- Accepted answers with one letter added, missing, changed or swapped, in verbs of four or more letters.
- Context for the scores
- Anticipationscount
- First key presses under 100 ms. They leave the medians.
- More than max(2, 20%) of the nouns invalidates the session
- Paste Attemptscount
- Times text was pasted or dropped into the box. Pasting is blocked.
- Context for the scores
- Nounscount
- Nouns recorded in the scored run.
- Fewer than planned is flagged for review
- Planned Nounscount
- Nouns the session planned to show.
- Informational
- Session in Another Language0/1
- 1 when the session ran in a language other than English, the language of the nouns.
- 1 is flagged for review
Use and evidence
61 papers from 2021 to 2025 name this task in their title or abstract, down 42% compared with 2016 to 2020. That ranks it 282 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 12 |
| 2024 | 6 |
| 2023 | 13 |
| 2022 | 16 |
| 2021 | 14 |
| 2020 | 22 |
| 2019 | 18 |
| 2018 | 15 |
| 2017 | 21 |
| 2016 | 29 |
| 2015 | 27 |
| 2014 | 23 |
| 2013 | 26 |
| 2012 | 23 |
| 2011 | 21 |
| 2010 | 25 |
| 2009 | 16 |
| 2008 | 19 |
| 2007 | 17 |
| 2006 | 33 |
| 2005 | 18 |
| 2004 | 22 |
| 2003 | 12 |
| 2002 | 12 |
| 2001 | 8 |
| 2000 | 12 |
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
- 5
- Papers on reliability or validity
- 47
- Registered clinical studies using it as an outcome
- 1
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.
- Nouns
- 40 concrete English nouns written for this test: 20 with one strong verb, such as pen (write), and 20 with many possible verbs, such as rope (tie, pull, climb). The two sets have mean word frequencies of Zipf 4.35 and 4.38 and mean lengths of 4.90 and 4.95 letters.
- Sets
- Each noun's set follows a stated rule after Thompson-Schill and colleagues (1997), checked by a second reading. No people have rated these nouns yet, so a pilot should confirm them.
- Display
- A fixation cross for 600 to 1,000 ms, then the noun with a text box until the answer or 10 seconds (standard; 5 to 20). Whatever has been typed at the limit counts as the answer.
- Order
- All 40 nouns, in an order drawn from the session's seed with no more than 3 of one kind in a row.
- Answers
- Typed in a text box with word correction off. Enter or a button submits. Spoken answers are not recorded.
- Scoring
- A listed verb in any form, or with one spelling slip, counts. Verbs that fit almost anything, such as use or get, count only where a noun's list has them, and the noun itself never counts.
- Practice
- 3 nouns outside the test set, each followed by a screen that says whether the verb fits. The test gives no feedback.
The nouns, the verb lists and the scoring rule were written for this test. The selection cost is a research measure read across a group: one person's difference rests on 20 nouns in each set, and there are no percentiles. The nouns and answers are English; sessions run in other languages are flagged for review.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: verb_generation_test from the scored run, one typed answer for each noun within its time limit. A typed verb has no chance level, so accuracy is not judged against chance. Answers on fewer than half the planned nouns invalidate. The same verb on 90% of at least 12 answers invalidates. First key presses under 100 ms on more than max(2, 20%) of the nouns invalidate, and any is flagged. Fewer nouns than planned withhold the percentile. Verbs that fit almost anything on a quarter of the planned nouns are flagged. A session run in another language is flagged. A hidden tab withholds the percentile.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. Not finding a verb in time can be a real result, so a noun with no answer is not an error of its own, but times read from fewer than half the nouns cannot be compared with a full session. Every noun is different, so one verb for nearly every noun cannot come from the nouns. A key press under 100 ms after the noun appears cannot be a reaction to it. A verb that fits almost anything, such as use or get, avoids choosing among the verbs a noun brings to mind, the process the selection cost measures. 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. A hidden tab lets the study, delay or response timers run without the participant, so a percentile is withheld after one.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Time per noun (seconds) | 10 | 5 to 20 |
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 Verb Generation Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Petersen SE and colleagues, 1988; Thompson-Schill SL and colleagues, 1997; Martin RC, Cheng Y, 2006), with the standard settings (time per noun (seconds): 10). The main score was selection cost (median time to the first key press on nouns with many possible verbs minus nouns with one strong verb, over accepted answers).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Petersen, S. E., Fox, P. T., Posner, M. I., Mintun, M., & Raichle, M. E. (1988). Positron emission tomographic studies of the cortical anatomy of single-word processing. Nature, 331(6157), 585-589. https://doi.org/10.1038/331585a0Verb generation in early brain imaging of single words.
- Thompson-Schill, S. L., D'Esposito, M., Aguirre, G. K., & Farah, M. J. (1997). Role of left inferior prefrontal cortex in retrieval of semantic knowledge: A reevaluation. Proceedings of the National Academy of Sciences, 94(26), 14792-14797. https://doi.org/10.1073/pnas.94.26.14792Nouns with one strong verb against nouns with many.
- Thompson-Schill, S. L., Swick, D., Farah, M. J., D'Esposito, M., Kan, I. P., & Knight, R. T. (1998). Verb generation in patients with focal frontal lesions: A neuropsychological test of neuroimaging findings. Proceedings of the National Academy of Sciences, 95(26), 15855-15860. https://doi.org/10.1073/pnas.95.26.15855The same two kinds of noun in a behavioral study.
- Martin, R. C., & Cheng, Y. (2006). Selection demands versus association strength in the verb generation task. Psychonomic Bulletin & Review, 13(3), 396-401. https://doi.org/10.3758/BF03193859The selection ratio and association strength.
- Snyder, H. R., & Munakata, Y. (2008). So many options, so little time: The roles of association and competition in underdetermined responding. Psychonomic Bulletin & Review, 15(6), 1083-1088. https://doi.org/10.3758/PBR.15.6.1083Separate effects of association and competition on response times.
Source papers
- Petersen SE, Fox PT, Posner MI, Mintun M, Raichle ME (1988). Positron emission tomographic studies of the cortical anatomy of single-word processing. Nature, 331(6157):585-589.Paradigm source, Adults
- Thompson-Schill SL, D'Esposito M, Aguirre GK, Farah MJ (1997). Role of left inferior prefrontal cortex in retrieval of semantic knowledge: A reevaluation. Proceedings of the National Academy of Sciences, 94(26):14792-14797.Paradigm source, Adults
- Martin RC, Cheng Y (2006). Selection demands versus association strength in the verb generation task. Psychonomic Bulletin and Review, 13(3):396-401.Paradigm source, Adults
Questions
How long does the Verb Generation 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?
Selection Cost, Median Time, Many Verbs, Median Time, One Strong Verb, Median Time to First Key, Fitting Verbs, Fitting Verbs, Many Verbs, Fitting Verbs, One Strong Verb, Strong Verb Given, Generic Verbs, Other Answers, No Answers, Noun Repeated, Spelling Slips, Anticipations, Paste Attempts, Nouns, Planned Nouns and Session in Another Language. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Verb Generation Task used in research?
61 papers from 2021 to 2025 name it in their title or abstract, down 42% compared with 2016 to 2020. That ranks it 282 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 Verb Generation Task have?
In the published research, 5 papers report its test-retest reliability and 1 registered clinical study uses 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?
Time per noun (seconds) (5 to 20). 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.