Social Cognition
Scrambled Sentences Test
Also called: interpretation bias task, scrambled sentence task
Measure how people resolve ambiguous words: participants tap five of six scrambled words to make a positive or a negative sentence, under a time limit and a memory load.
6-8 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
- 148
- Evidence grade
- B
- Registered clinical studies
- 20
What it measures
Measures how people resolve ambiguous words in research. Participants tap five of six scrambled words to make a sentence that can read as positive or negative, while holding a six-digit number and working against a time limit. The main score is the percent of valid sentences that are negative. The words are in English.
Constructs
- Interpretation bias
- Resolving ambiguity
- Mental control under load
- Emotional processing
Research uses
- Interpretation bias research
- Cognitive load and mental control studies
- Emotion research at the group level
- Within-study group comparison
How the paradigm works
The Scrambled Sentences Test measures how people resolve ambiguous words in research. Each item is six words in scrambled order. The participant taps five of them, in order, to make a sentence, and one word is left out. Every item can make a positive sentence or a negative one, and nothing says so. The instruction is to make the first sentence that comes to mind and to work quickly. The task comes from Wenzlaff and Bates (1998), and a six-digit number held during the block is the load in the source.
More about the paradigm
By default the block holds 20 items with 240 seconds in all, the limit in Van der Does (2005), and a bar shows the time left. The number appears for 10 seconds before the block and is typed at the end. Undo takes back the last word, because a tap can slip on a touch screen, and Next moves on once five words are chosen. Two neutral practice sets with feedback and no time limit come first. The session's seed draws the items, their order and the number, and all of it is saved. The 30 items were written for this test, on everyday life, work, friends and the future; no published item is used.
An answer is valid when it uses five different words of the item and matches one of its two stored sentences. The main score is the negative share: negative sentences as a percent of valid sentences, as in the source studies. The test also reports the counts of positive and negative sentences, invalid answers, items not reached, mean time per item and whether the number was recalled. Across published versions, a meta-analysis pooled internal consistency at alpha = .79 (Würtz and colleagues, 2022). A session with valid sentences on fewer than half of its items is invalid. A wrong number flags the session and does not void it. The words are English; sessions run in other languages are flagged for review. Scores support research and do not diagnose.
Scores
- Negative Share%
- Negative sentences as a percent of valid sentences.
- Higher means more items were resolved in the negative direction
- Positive Sentencesitems
- Valid sentences that read as positive.
- Informational
- Negative Sentencesitems
- Valid sentences that read as negative.
- Informational
- Valid Sentencesitems
- Answers that match one of the item's two stored sentences.
- The base of the main score
- Invalid Answersitems
- Answers of five words that match neither stored sentence.
- Lower supports interpretation
- Items Not Reacheditems
- Items set with no finished answer when the block ended, counting the item on screen.
- Informational
- Mean Time per Itemms
- Mean time from an item's screen to Next, over answered items.
- Informational
- Number Recalled0/1
- 1 when the six-digit number was typed correctly, 0 when not. Empty with the memory load off.
- 1 supports interpretation
- Digits in Placecount
- Digits of the number typed in the right place, out of six.
- Informational
- Anticipationscount
- Answers under 100 ms.
- Lower supports interpretation
- Items Setitems
- Items the session set out to show.
- Informational
Use and evidence
148 papers from 2021 to 2025 name this task in their title or abstract, up 49% compared with 2016 to 2020. That ranks it 151 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 48 |
| 2024 | 25 |
| 2023 | 32 |
| 2022 | 18 |
| 2021 | 25 |
| 2020 | 19 |
| 2019 | 21 |
| 2018 | 24 |
| 2017 | 16 |
| 2016 | 19 |
| 2015 | 27 |
| 2014 | 26 |
| 2013 | 20 |
| 2012 | 14 |
| 2011 | 12 |
| 2010 | 10 |
| 2009 | 11 |
| 2008 | 4 |
| 2007 | 6 |
| 2006 | 4 |
| 2005 | 3 |
| 2004 | 2 |
| 2003 | 6 |
| 2002 | 3 |
| 2001 | 7 |
| 2000 | 6 |
Use grade C
100 to 299 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
- 12
- Papers on reliability or validity
- 61
- Registered clinical studies using it as an outcome
- 20
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.
- Items
- 30 word sets written for this test, on everyday life, work, friends and the future.
- Stimulus
- Six lower-case words in a fixed scrambled order. Five of them make a positive sentence, and a different five make a negative one. The two sentences differ in one word, in the same place.
- Session
- 20 sets by default, 10 to 30 as a setting. The session's seed draws them and orders them so that no topic runs for more than 2 sets in a row, and all of it is saved.
- Time limit
- 240 seconds for the block by default, 120 to 480 as a setting, with a bar that runs across the sets. 240 seconds for 20 sets is the limit in Van der Does (2005).
- Memory load
- A six-digit number shown for 10 seconds before the block and typed at the end. On by default; off removes both screens.
- Answer
- Tap five word tiles in order. Undo takes back the last word, and Next moves on once five words are chosen.
- Practice
- 2 neutral sets with feedback and no time limit. The test has no feedback.
The word sets were written for this test; no published item set is used. The words 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: scrambled_sentences_item from the scored run, one row per item reached. Neither sentence is correct, so accuracy is not judged against chance, and the same kind of sentence on every item is accepted. Fewer valid sentences than half of the items set invalidates. Answers under 200 ms on at least half of 4 or more answers invalidate. A wrong six-digit number 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. Making only positive or only negative sentences is the result the test measures, so repeated kinds of answer are not checked. The share of negative sentences rests on the valid sentences, and with fewer than half of the items answered validly, one sentence moves it too far to compare with a full session. A wrong number means the load may not have been held, which changes the condition and does not undo the sentences made, so it flags the session and never voids it. 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 word sets 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 |
|---|---|---|
| Items | 20 | 10 to 30 |
| Time limit (seconds) | 240 | 120 to 480 |
| Memory load | On | On or off |
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.
- 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 Scrambled Sentences Test in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Wenzlaff RM, Bates DE, 1998; Rude SS, Wenzlaff RM, Gibbs B, Vane J, Whitney T, 2002; Van der Does W, 2005), with the standard settings (items: 20; time limit (seconds): 240; memory load: On). The main score was negative share (negative sentences as a percent of valid sentences).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Wenzlaff, R. M., & Bates, D. E. (1998). Journal of Personality and Social Psychology, 75(6), 1559-1571. https://doi.org/10.1037/0022-3514.75.6.1559The task and the held number.
- Rude, S. S., Wenzlaff, R. M., Gibbs, B., Vane, J., & Whitney, T. (2002). Cognition and Emotion, 16(3), 423-440. https://doi.org/10.1080/02699930143000554
- Rude, S. S., Valdez, C. R., Odom, S., & Ebrahimi, A. (2003). Cognitive Therapy and Research, 27(4), 415-429. https://doi.org/10.1023/A:1025472413805
- Van der Does, W. (2005). British Journal of Clinical Psychology, 44(1), 1-14. https://doi.org/10.1348/014466504X19442Sets of 20 items, the 4-minute limit and the score.
- Hallion, L. S., & Ruscio, A. M. (2011). Psychological Bulletin, 137(6), 940-958. https://doi.org/10.1037/a0024355Meta-analysis of the wider research on cognitive biases.
- Bowler, J. O., Mackintosh, B., Dunn, B. D., Mathews, A., Dalgleish, T., & Hoppitt, L. (2012). Journal of Consulting and Clinical Psychology, 80(6), 1021-1033. https://doi.org/10.1037/a0029932A six-digit number viewed for 10 seconds.
- Würtz, F., Zahler, L., Blackwell, S. E., Margraf, J., Bagheri, M., & Woud, M. L. (2022). Clinical Psychology Review, 93, 102133. https://doi.org/10.1016/j.cpr.2022.102133Meta-analysis of published versions.
Source papers
- Wenzlaff and Bates (1998): the task. Van der Does (2005): sets of 20 and the 4-minute limit. Würtz and colleagues (2022): meta-analysis of published versions.Paradigm source, Adults
Questions
How long does the Scrambled Sentences Test take?
About 6 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?
Negative Share, Positive Sentences, Negative Sentences, Valid Sentences, Invalid Answers, Items Not Reached, Mean Time per Item, Number Recalled, Digits in Place, Anticipations and Items Set. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Scrambled Sentences Test used in research?
148 papers from 2021 to 2025 name it in their title or abstract, up 49% compared with 2016 to 2020. That ranks it 151 of 395 test families in the ConductCognition test census (use grade C). Source: ConductCognition test census 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Scrambled Sentences Test have?
In the published research, 12 papers report its test-retest reliability and 20 registered clinical studies use it as an outcome (evidence grade B in the ConductCognition test census).
Does it report a percentile?
No. Scores and quality flags only.
Which settings can a researcher change?
Items (10 to 30), time limit (seconds) (120 to 480) and memory load (On or off). 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.