Executive Function
Sentence Completion Inhibition Test
Also called: sentence completion suppression task, verbal response suppression task, response initiation and suppression task
Measure verbal response suppression: participants complete sentences with an expected word, then type words unrelated to other sentences, graded by a stated rule.
5-8 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
- 48
- Evidence grade
- C
- Registered clinical studies
- 21
What it measures
Measures the ability to hold back a strongly expected word. Participants read sentences missing their last word and type one word: in part 1 a word that completes each of 15 sentences, and in part 2 a word with no link to each of 15 others. The main score counts part 2 answers that complete or relate to the sentence. The sentences are in English.
Constructs
- Response inhibition
- Verbal response suppression
- Response initiation
- Strategy use
Research uses
- Executive function research
- Aging and neuropsychology research
- Within-study group comparison
How the paradigm works
The sentence completion inhibition test measures the ability to hold back a strongly expected word, after the method of Burgess and Shallice (1996). Each sentence is missing its last word, which the rest of the sentence makes highly predictable. In part 1 (initiation) participants type a word that completes each of 15 sentences, as quickly as they can. In part 2 (suppression) they type a word with no link to each of 15 other sentences, a different word each time. Two practice sentences with feedback come before each part. The 34 sentences, their word lists and the scoring rule were written for this test; the sentences are English, and the instructions follow the session's language. Every session sees the same sentences in each part, in an order drawn from its seed. Answers are typed, and each sentence waits 30 seconds by default.
More about the paradigm
Each part 2 answer is graded by a stated rule. It is a completion error when it is one of the sentence's listed completions: the expected word or another word that completes the sentence sensibly. It is a related error when it repeats a word of the sentence or is one of its listed related words: associates of the expected word, its opposite and words of the same scene. Any other word counts as unrelated. A plural counts as the word, and part 1 also accepts one spelling slip. Burgess and Shallice (1996) found that people with frontal lobe lesions gave more words related to the sentence in the suppression condition, and that initiation and suppression were unrelated. Volle and colleagues (2012) linked initiation slowing, suppression slowing and suppression errors to partly different prefrontal regions. Martyr, Boycheva and Kudlicka (2019) found that completion errors, but not part 2 time adjusted for part 1 time, separated people with early Alzheimer's or Parkinson's disease from older controls; the test reports both.
The main score counts part 2 errors of both kinds, and each kind is also reported. Times run from the sentence's appearance to the first typed character, a typed stand-in for the spoken response latency of the original; the suppression cost is the part 2 median time minus the part 1 median. Phone keyboards are slower, so compare times within a device type. The word lists cannot hold every related word, so the rule can count a related answer as unrelated; every typed answer is saved so a study can score it again. Scores support research and do not diagnose.
Scores
- Part 2 Errorscount
- Part 2 answers that complete the sentence or are related to it, out of 15.
- Lower means the expected word was held back more often
- Part 2 Completion Errorscount
- Part 2 answers that complete the sentence sensibly.
- Lower is better; the clearest failure to hold back the expected word
- Part 2 Related Errorscount
- Part 2 answers that repeat a word of the sentence or are on its list of related words.
- Lower is better
- Part 2 Median Timems
- Median time from a part 2 sentence's appearance to the first typed character, over answered sentences, leaving out starts under 100 ms.
- Lower is faster
- Part 1 Median Timems
- Median time from a part 1 sentence's appearance to the first typed character, over answered sentences, leaving out starts under 100 ms.
- Lower is faster
- Suppression Costms
- Part 2 median time minus part 1 median time.
- Higher means finding an unrelated word slowed answers more
- Part 1 Completionscount
- Part 1 answers that complete their sentence by the word lists, out of 15.
- Near 15 expected; fewer than 8 is flagged for review
- No Answercount
- Sentences in either part with no answer before the time ran out.
- 0 expected
- Repeated Answerscount
- Part 2 answers that repeat an earlier part 2 answer.
- 5 or more is flagged for review
- Starts Under 100 mscount
- Answers whose first character came less than 100 ms after the sentence appeared.
- Expected to be 0
- Sentences Presentedcount
- Sentences shown in both parts.
- 30 in a full session
- Paste Attemptscount
- Blocked attempts to paste or drop text into the answer box.
- Any is flagged for review
Use and evidence
48 papers from 2021 to 2025 name this task in their title or abstract, down 51% compared with 2016 to 2020. That ranks it 258 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 14 |
| 2024 | 11 |
| 2023 | 2 |
| 2022 | 12 |
| 2021 | 9 |
| 2020 | 15 |
| 2019 | 12 |
| 2018 | 13 |
| 2017 | 31 |
| 2016 | 26 |
| 2015 | 22 |
| 2014 | 14 |
| 2013 | 14 |
| 2012 | 20 |
| 2011 | 15 |
| 2010 | 19 |
| 2009 | 16 |
| 2008 | 19 |
| 2007 | 11 |
| 2006 | 7 |
| 2005 | 8 |
| 2004 | 7 |
| 2003 | 6 |
| 2002 | 5 |
| 2001 | 3 |
| 2000 | 9 |
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
- 31
- Registered clinical studies using it as an outcome
- 21
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.
- Sentences
- 34 English sentences written for this test, each missing a strongly expected last word: 15 in part 1, 15 in part 2 and 2 practice sentences before each part. Every session sees the same sentences in each part, in an order drawn from its seed.
- Part 1
- Type one word that completes the sentence, as quickly as possible.
- Part 2
- Type one word that has no link to the sentence, a different word each time, as quickly as possible.
- Answer
- Typed in a box and sent with Enter or a button. Each sentence waits 30 seconds by default, 15 to 60 as a setting, and then counts as unanswered.
- Scoring rule
- A part 2 answer is a completion error if it is one of the sentence's listed completions, a related error if it is a word of the sentence or one of its listed related words, and unrelated otherwise. A plural counts as the word.
- Practice
- 2 sentences before each part, each followed by feedback by the same rule.
- Scores
- Part 2 errors, each kind of error, the median time to the first typed letter in each part and their difference, and part 1 completions.
The method follows Burgess and Shallice (1996), with sentences, word lists and a scoring rule written for this test; no published test's sentences, scores or norms are used. Answers are typed rather than spoken. The lists cannot hold every related word, so the rule can count a related answer as unrelated; every typed answer is saved for rescoring. The sentences are English; sessions run in other languages are flagged for review. The 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: the scored run's typed answers to the part 1 and part 2 sentences, each within its time limit. Typed words have no chance level, so accuracy is not judged against chance, and part 2 errors are the result the test measures. Fewer than two thirds of the part 2 sentences answered invalidates, and 3 or more without an answer is flagged. One answer on 90% of answered sentences invalidates. Answers under 200 ms on at least half the sentences invalidate. Five or more repeated part 2 answers, fewer than 8 part 1 completions, a blocked paste or a session run in another language is flagged for review. 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. A sentence with no answer is neither an error nor a success, so a part 2 score over too few answers cannot be compared with a full one. One word for every sentence does not follow either instruction. Repeating one unrelated word avoids the work of suppression, which Burgess and Shallice (1996) tied to strategy, so it is flagged. Part 1 answers that rarely complete the sentence suggest the instructions were misread. 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. The sentences are English. A session run in another language shows English sentences 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.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Time per sentence (seconds) | 30 | 15 to 60 |
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
- 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.
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Sentence Completion Inhibition Test in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Burgess & Shallice, 1996), with the standard settings (time per sentence (seconds): 30). The main score was part 2 errors (part 2 answers that complete the sentence or are related to it, out of 15).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Burgess, P. W., & Shallice, T. (1996). Response suppression, initiation and strategy use following frontal lobe lesions. Neuropsychologia, 34(4), 263-272. https://doi.org/10.1016/0028-3932(95)00104-2
- Volle, E., de Lacy Costello, A., Coates, L. M., McGuire, C., Towgood, K., Gilbert, S., Kinkingnehun, S., McNeil, J. E., Greenwood, R., Papps, B., van den Broeck, M., & Burgess, P. W. (2012). Dissociation between verbal response initiation and suppression after prefrontal lesions. Cerebral Cortex, 22(10), 2428-2440. https://doi.org/10.1093/cercor/bhr322Initiation and suppression times and suppression errors.
- Martyr, A., Boycheva, E., & Kudlicka, A. (2019). Journal of Neuropsychology, 13(1), 67-81. https://doi.org/10.1111/jnp.12129Part 2 time adjusted for part 1 time, and completion errors.
- Peelle, J. E., Miller, R. L., Rogers, C. S., Spehar, B., Sommers, M. S., & Van Engen, K. J. (2020). Completion norms for 3085 English sentence contexts. Behavior Research Methods, 52(4), 1795-1799. https://doi.org/10.3758/s13428-020-01351-1Published sentence contexts the sentences were checked against.
Source papers
- Burgess PW, Shallice T (1996). Response suppression, initiation and strategy use following frontal lobe lesions. Neuropsychologia, 34(4):263-272.Paradigm source, Adults
Questions
How long does the Sentence Completion Inhibition Test take?
About 5 to 8 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?
Part 2 Errors, Part 2 Completion Errors, Part 2 Related Errors, Part 2 Median Time, Part 1 Median Time, Suppression Cost, Part 1 Completions, No Answer, Repeated Answers, Starts Under 100 ms, Sentences Presented and Paste Attempts. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Sentence Completion Inhibition Test used in research?
48 papers from 2021 to 2025 name it in their title or abstract, down 51% compared with 2016 to 2020. That ranks it 258 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 Sentence Completion Inhibition Test have?
In the published research, 1 paper reports its test-retest reliability and 21 registered clinical studies use it as an outcome (evidence grade C in the ConductCognition test census).
Which settings can a researcher change?
Time per sentence (seconds) (15 to 60). 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.
Related tests