Attention
Attribute Amnesia Task
Also called: attribute amnesia paradigm, surprise question paradigm, surprise memory test, unexpected question task
Measure whether participants can report a digit they just used to find a target, from one surprise question compared with expected questions, read across a group.
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. Preview any test free after you sign up.
- Papers, 2021 to 2025
- 27
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
Measures whether people can report a feature they just used to find a target. Four digits flash around a cross, and participants click where the one larger than 5 was; then a surprise question asks which digit it was, and control trials ask again once the question is expected. The main score is whether the surprise answer was right, a measure read across a group. The surprise works once, so each person takes this test once.
Constructs
- Attribute amnesia
- Selective attention
- Working memory encoding
- Expectation
Research uses
- Attention research
- Memory research
How the paradigm works
Attribute amnesia is the failure to report a feature of a target that was just used to find it, when the question is not expected (Chen and Wyble, 2015). This is the digit version (Fu et al., 2024; Li et al., 2024). Four digits flash for 250 ms at the corners of a square around a cross: one larger than 5 (6, 7, 8 or 9) and three smaller. The participant clicks where the digit larger than 5 was. After 4 practice trials with feedback and 28 of these location trials (12 to 60 as a setting), a surprise question on the next trial asks which digit larger than 5 it was, choosing from 6, 7, 8 and 9, before the place. Four control trials then ask both questions, now expected.
More about the paradigm
The main score is whether the surprise answer was right. Each person gives one surprise answer, so the share right is read across a group and compared with the share right on the first control trial; in the sources about half of adults are right on the surprise question and about nine in ten on the first control trial (Fu et al., 2024; Li et al., 2024). The location answers, before the surprise and after it, are reported too. The surprise answer gets worse as more location trials come before it (Wyble et al., 2019), so the number of trials should stay the same within a study. Scores support research and do not diagnose.
Scores
- Surprise Answer Right0/1
- 1 when the digit named on the surprise question was the digit larger than 5 just shown, 0 when not.
- Read across a group against the first control trial; chance is 1 in 4
- First Control Answer Right0/1
- The same question on the first trial after the surprise, now expected.
- The comparison for the surprise answer
- Control Accuracy%
- Share of the four control trials with the digit named right.
- Under 0.5 makes the session invalid
- Surprise Trial Place Right0/1
- 1 when the place was right on the surprise trial, answered after the surprise question.
- Informational
- Place Accuracy Before the Surprise%
- Share of location trials before the surprise with the place right.
- Under 0.6 makes the session invalid
- Place Accuracy on Control Trials%
- Share of control trials with the place right.
- Informational
- Surprise Answer Timems
- Time from the surprise question to the answer.
- Informational
- Did a Task Like This Before0/1
- 1 when the participant said they had done this task or one like it before, 0 when not; empty for not sure.
- 1 is flagged for review
- Earlier Sessionscount
- This participant's sessions of this test begun before this one.
- More than 0 is flagged: only a first session measures attribute amnesia
- Anticipationscount
- Answers given under 100 ms after the question appeared, left out of the scores.
- More than max(2, 20%) of the location answers makes the session invalid
- Location Trials Answeredtrials
- Location trials before the surprise with a usable answer.
- Informational
- Planned Trialstrials
- Test trials planned by the settings: the location trials, the surprise trial and the four control trials.
- Informational
Use and evidence
27 papers from 2021 to 2025 name this task in their title or abstract, up 42% compared with 2016 to 2020. That ranks it 309 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 6 |
| 2024 | 2 |
| 2023 | 9 |
| 2022 | 3 |
| 2021 | 7 |
| 2020 | 2 |
| 2019 | 7 |
| 2018 | 3 |
| 2017 | 5 |
| 2016 | 2 |
| 2015 | 3 |
| 2014 | 0 |
| 2013 | 0 |
| 2012 | 0 |
| 2011 | 0 |
| 2010 | 0 |
| 2009 | 0 |
| 2008 | 0 |
| 2007 | 0 |
| 2006 | 0 |
| 2005 | 0 |
| 2004 | 0 |
| 2003 | 0 |
| 2002 | 0 |
| 2001 | 0 |
| 2000 | 0 |
Use grade E
Fewer than 30 papers from 2021 to 2025.
Evidence grade D
Little published measurement evidence found.
- Papers on test-retest reliability
- 0
- Papers on reliability or validity
- 0
- 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 0.14.0, counted 2 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Display
- A dark panel of 400 by 400 px with a cross in the middle and four rings at the corners of a 240 px square.
- Trial
- The cross and rings for 800 to 1,800 ms, the four digits for 250 ms, the cross and rings for 500 ms, then the questions. One digit is larger than 5 (6, 7, 8 or 9) and three are smaller (1, 2, 3 or 4).
- Answer
- Click the ring where the digit larger than 5 was, with no time limit.
- Trials
- 28 location trials (standard, 12 to 60), with each place and each target digit equally often, in an order drawn from the session's seed; then the surprise trial and 4 control trials.
- Surprise question
- On the surprise trial, before the place, a question asks which digit larger than 5 was just seen, from 6, 7, 8 and 9. The 4 control trials ask both questions, now expected.
- Practice
- 4 trials with the digits shown for 1,000 ms; after each, a green ring marks the right place for 1,200 ms.
- Score
- Whether the surprise answer was right. One answer per person, so the share right is read across a group and compared with the first control trial.
- Once per person
- The question is a surprise only the first time, so each person takes the test once. Scoring flags a later session of the same participant and a reported earlier task like this one.
The last screen explains the surprise question and asks people not to tell others; studies debrief participants as their ethics review requires. More location trials make the surprise answer less often right, so keep one setting within a study. Results are read across a group and do not diagnose; no percentile is computed.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: attribute_amnesia_location from the scored run, one answer per trial, practice left out. The place has 4 choices, so chance is 25%, and the target took each place equally often before the surprise: accuracy below chance invalidates, not above chance is flagged, and one place on 90% of the answers invalidates when accuracy is not above chance. Responses under 100 ms follow the timed-test rule. Under 60% right on the places before the surprise, or under half the numbers named right on the control trials, invalidates. No surprise answer, a reported earlier task like this one, and an earlier session of this test are 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). Finding the number is what fixes attention on it, so a session that did not find it does not test memory for it; the studies leave out people under 60% on the places (Fu et al. 2024; O'Donnell, Chen and Wyble 2021) or with poor control answers (Li et al. 2024). The question is a surprise only once, so an earlier encounter with the task is flagged.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Location trials before the surprise question | 28 | 12 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
- 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 Attribute Amnesia Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Chen & Wyble, 2015), with the standard settings (location trials before the surprise question: 28). The main score was surprise answer right (1 when the digit named on the surprise question was the digit larger than 5 just shown, 0 when not).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Chen, H., & Wyble, B. (2015). Amnesia for object attributes: Failure to report attended information that had just reached conscious awareness. Psychological Science, 26(2), 203-210. https://doi.org/10.1177/0956797614560648The task: locating a target many times, then one surprise question about its identity, then control trials.
- Fu, Y., Guo, T., Zheng, J., He, J., Shen, M., & Chen, H. (2024). Children exhibit superior memory for attended but outdated information compared to adults. Nature Communications, 15, 4058. https://doi.org/10.1038/s41467-024-48457-0The digit version: a digit larger than 5 among digits smaller than 5, with 4 control trials.
- Li, J., Chen, L., Zhou, D., Tang, E., Zheng, J., Huang, X., Zhong, B.-L., Guan, C., Liu, H., Shen, M., & Chen, H. (2024). Flexibility retained: Unimpaired updating of expectations in schizophrenia. Behavioral Sciences, 14(1), 41. https://doi.org/10.3390/bs14010041The digits shown for 250 ms, and adult results on the surprise and control trials.
- Chen, H., & Wyble, B. (2016). Attribute amnesia reflects a lack of memory consolidation for attended information. Journal of Experimental Psychology: Human Perception and Performance, 42(2), 225-234. https://doi.org/10.1037/xhp0000133
- Wyble, B., Hess, M., O'Donnell, R. E., Chen, H., & Eitam, B. (2019). Learning how to exploit sources of information. Memory & Cognition, 47(4), 696-705. https://doi.org/10.3758/s13421-018-0881-xThe surprise answer gets worse as more location trials come before it.
- Chen, W., & Howe, P. D. L. (2017). Attribute amnesia is greatly reduced with novel stimuli. PeerJ, 5, e4016. https://doi.org/10.7717/peerj.4016
Source papers
- Chen and Wyble (2015): the task. Fu et al. (2024) and Li et al. (2024): the digit version and its timing. Wyble et al. (2019): more trials before the surprise make its answer worse.Paradigm source, Adults
Questions
How long does the Attribute Amnesia 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?
Surprise Answer Right, First Control Answer Right, Control Accuracy, Surprise Trial Place Right, Place Accuracy Before the Surprise, Place Accuracy on Control Trials, Surprise Answer Time, Did a Task Like This Before, Earlier Sessions, Anticipations, Location Trials Answered and Planned Trials. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Attribute Amnesia Task used in research?
27 papers from 2021 to 2025 name it in their title or abstract, up 42% compared with 2016 to 2020. That ranks it 309 of 395 test families in the ConductCognition test census (use grade E). Source: ConductCognition test census 0.14.0, counted 2 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Attribute Amnesia 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?
Location trials before the surprise question (12 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?
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.