Episodic Memory
Picture Superiority Task
Also called: picture superiority effect, pictorial superiority effect, picture-word recognition task, pictures versus words memory task
Compare memory for pictures and words: a mixed list of pictures and picture names, then a recognition test of each item in its own form, scored as d-prime for pictures minus words.
7-9 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
- 31
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
Compares memory for pictures and for words. Participants study a mixed list of pictures and picture names, then say whether each picture or word was in the list. The main score is how much better pictures are recognized.
Constructs
- Episodic memory
- Recognition memory
- Picture and word encoding
Research uses
- Memory research in aging, mild cognitive impairment and Alzheimer's disease
- Studies of how people encode pictures and words
- Within-study group comparison
How the paradigm works
The Picture Superiority Task measures how much better people remember pictures than words. A list mixes pictures of everyday things with the written names of other things, one at a time: a cross for 500 ms, then a picture or a word for 2 seconds. After 60 seconds of true-or-false arithmetic, adjustable from 0 to 120, a recognition test shows every studied item in the form it was studied in, mixed with as many new pictures and new words, one at a time. Participants answer Seen or New with no time limit.
More about the paradigm
The design follows the recognition studies of the effect (Shepard 1967; Snodgrass, Volvovitz and Walfish 1972), with the mixed list and the test in the form studied that Ensor, Surprenant and Neath (2019) ran online, and 2-second study items as in Ally, Gold and Budson (2009) and Curran and Doyle (2011). Testing each item in its own form gives pictures and words their own false alarm rates, so d′ can be computed for each. A test in words only would make every studied picture depend on the name a participant gives it in their own language. The standard list has 20 pictures and 20 words, so the test has 80 items; a setting gives 12 to 20 in each role. A practice of one picture and one word comes first, with a short test and feedback.
The pictures are 80 emoji from Noto Emoji, everyday objects and animals with one common name in each of the eight languages, and the words are those names, written in the session's language. Each session gives every item one role and one form at random: studied as a picture, studied as a word, new as a picture or new as a word. So no thing is seen both ways in a session, and across sessions every item serves in every role.
The main score is the picture superiority effect: d′ for pictures minus d′ for words, each from its own hits and false alarms with the log-linear correction (Hautus 1995). Scores also give the hit and false alarm rates for each form and the difference in hit rates, corrected recognition (hits minus false alarms) for each form and its difference, the response criterion for each form, overall accuracy and answer times. Online, with color pictures and plain black words, d′ was about 1 higher for pictures (Ensor and colleagues 2019). Scores support research and do not diagnose.
No norms are applied to this task. The effect is a difference between pictures and words with no better direction, so sessions are not ranked against other people. Use the scores for raw analysis and within-study comparison.
Scores
- Picture Superiority Effectd'
- d-prime for pictures minus d-prime for words, each from the hits and false alarms of its own form, with the log-linear correction.
- Positive when pictures were recognized better; no better direction
- d-Prime, Picturesd'
- z(hit rate for studied pictures) minus z(false alarm rate for new pictures), with the log-linear correction.
- Higher means better recognition of pictures
- d-Prime, Wordsd'
- z(hit rate for studied words) minus z(false alarm rate for new words), with the log-linear correction.
- Higher means better recognition of words
- Hit Rate, Pictures%
- Share of studied pictures answered Seen.
- Higher means better memory for pictures
- Hit Rate, Words%
- Share of studied words answered Seen.
- Higher means better memory for words
- False Alarm Rate, Pictures%
- Share of new pictures answered Seen.
- Lower is better
- False Alarm Rate, Words%
- Share of new words answered Seen.
- Lower is better
- Hit Rate Difference%
- Hit rate for pictures minus hit rate for words.
- Positive when more pictures were recognized
- Corrected Recognition, Pictures%
- Hit rate minus false alarm rate for pictures.
- Higher means better recognition of pictures
- Corrected Recognition, Words%
- Hit rate minus false alarm rate for words.
- Higher means better recognition of words
- Corrected Recognition Difference%
- Corrected recognition for pictures minus that for words.
- Positive when pictures were recognized better
- Criterion, Picturesz
- Minus the mean of z(hit rate) and z(false alarm rate) for pictures, with the log-linear correction.
- Positive means a stricter Seen; context for the rates
- Criterion, Wordsz
- Minus the mean of z(hit rate) and z(false alarm rate) for words, with the log-linear correction.
- Positive means a stricter Seen; context for the rates
- Recognition Accuracy%
- Share of test items answered correctly, Seen for studied and New for new items.
- Higher is better
- Answer Time, Picture Hitsms
- Median time to answer Seen to studied pictures.
- Context for the hit rates
- Answer Time, Word Hitsms
- Median time to answer Seen to studied words.
- Context for the hit rates
- Answer Time, Correct Rejectionsms
- Median time to answer New to new pictures and words.
- Context for the false alarm rates
- Anticipationscount
- Test answers faster than 100 ms, left out of every score.
- Lower is better
- Items Testedcount
- Test items shown in the scored run (80 at the standard settings).
- Informational
- Delay Arithmetic Accuracy%
- Share of true-or-false arithmetic problems in the delay answered correctly.
- Context for the delay
- Delayseconds
- Seconds of arithmetic between the study list and the test.
- Informational
Use and evidence
31 papers from 2021 to 2025 name this task in their title or abstract, about the same as from 2016 to 2020. That ranks it 297 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 12 |
| 2024 | 8 |
| 2023 | 7 |
| 2022 | 2 |
| 2021 | 2 |
| 2020 | 6 |
| 2019 | 8 |
| 2018 | 8 |
| 2017 | 3 |
| 2016 | 6 |
| 2015 | 4 |
| 2014 | 3 |
| 2013 | 6 |
| 2012 | 9 |
| 2011 | 2 |
| 2010 | 5 |
| 2009 | 9 |
| 2008 | 8 |
| 2007 | 7 |
| 2006 | 6 |
| 2005 | 1 |
| 2004 | 3 |
| 2003 | 5 |
| 2002 | 4 |
| 2001 | 3 |
| 2000 | 4 |
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
- 4
- 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 31 October 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Items
- 84 emoji pictures of everyday objects and animals from Noto Emoji, each with its name written in the session's language: 80 for the test and 4 for practice.
- Study list
- 40 items (standard; 12 to 20 in each role), half pictures and half words, each after a 500 ms cross for 2,000 ms. No more than 4 pictures or 4 words in a row.
- Delay
- 60 seconds of true-or-false arithmetic (standard; 0 to 120).
- Test
- Every studied item in the form it was studied in and as many new pictures and words, 80 in all, answered Seen or New with no time limit. No more than 4 old or new items in a row.
- Practice
- 1 picture and 1 word, then a test of those 2 and 2 new ones with 1,500 ms of feedback.
- Roles
- Each session gives every item one role at random: studied as a picture, studied as a word, new as a picture or new as a word.
The pictures are the Directed Forgetting Task's, and most are also shown by the Repetition Priming and Triplet Odd-One-Out tasks; run those in another session. Studies of the effect often use rated picture sets; this task uses emoji named in each language.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: picture_superiority_test from the scored run, one Seen or New answer per picture or word, with the delay arithmetic from the same run. Seen or New accuracy significantly below chance (one-sided exact binomial, p < .05, chance 50%) invalidates; accuracy not above chance over 10 or more answers is flagged. The same answer on 90% of the items invalidates when accuracy is not above chance. Answers under 200 ms on at least half of 4 or more answered items invalidate. No arithmetic answered in the delay, or delay arithmetic at or below chance over at least 10 answers, is flagged. A hidden tab withholds the percentile.
Why these rules
Half the test items were studied and half are new, in each form, so guessing gives 50%: 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). 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. Accuracy is judged over pictures and words together; the picture superiority effect itself, which may be zero or negative, is never judged. Skipping the delay arithmetic means that part of the procedure may not have been followed. 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 |
|---|---|---|
| Pictures and words in each role | 20 | 12 to 20 |
| Arithmetic delay (seconds) | 60 | 0 to 120 |
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
- 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 Picture Superiority Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Shepard, 1967; Snodgrass, Volvovitz & Walfish, 1972), with the standard settings (pictures and words in each role: 20; arithmetic delay (seconds): 60). The main score was picture superiority effect (d-prime for pictures minus d-prime for words, each from the hits and false alarms of its own form, with the log-linear correction).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Shepard, R. N. (1967). Recognition memory for words, sentences, and pictures. Journal of Verbal Learning and Verbal Behavior, 6(1), 156-163. https://doi.org/10.1016/S0022-5371(67)80067-7
- Snodgrass, J. G., Volvovitz, R., & Walfish, E. R. (1972). Recognition memory for words, pictures, and words + pictures. Psychonomic Science, 27(6), 345-347. https://doi.org/10.3758/BF03328986A mixed study list and a test in the form studied, scored with d-prime.
- Ally, B. A., Gold, C. A., & Budson, A. E. (2009). The picture superiority effect in patients with Alzheimer's disease and mild cognitive impairment. Neuropsychologia, 47(2), 595-598. https://doi.org/10.1016/j.neuropsychologia.2008.10.0102-second study items and corrected recognition for pictures and words.
- Curran, T., & Doyle, J. (2011). Picture superiority doubly dissociates the ERP correlates of recollection and familiarity. Journal of Cognitive Neuroscience, 23(5), 1247-1262. https://doi.org/10.1162/jocn.2010.21464Mixed study lists of pictures and words, 2 seconds per item.
- Ensor, T. M., Surprenant, A. M., & Neath, I. (2019). Increasing word distinctiveness eliminates the picture superiority effect in recognition: Evidence for the physical-distinctiveness account. Memory & Cognition, 47(1), 182-193. https://doi.org/10.3758/s13421-018-0858-9The design run online, with color pictures and plain words.
- Hautus, M. J. (1995). Corrections for extreme proportions and their biasing effects on estimated values of d'. Behavior Research Methods, Instruments, & Computers, 27(1), 46-51. https://doi.org/10.3758/BF03203619The log-linear correction for d-prime.
Source papers
- Shepard RN (1967). Recognition memory for words, sentences, and pictures. Journal of Verbal Learning and Verbal Behavior, 6(1):156-163.Paradigm source, Adults
- Ensor TM, Surprenant AM, Neath I (2019). Increasing word distinctiveness eliminates the picture superiority effect in recognition: evidence for the physical-distinctiveness account. Memory & Cognition, 47(1):182-193.Paradigm source, Adults 19 to 39
Questions
How long does the Picture Superiority Task take?
About 7 to 9 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?
Picture Superiority Effect, d-Prime, Pictures, d-Prime, Words, Hit Rate, Pictures, Hit Rate, Words, False Alarm Rate, Pictures, False Alarm Rate, Words, Hit Rate Difference, Corrected Recognition, Pictures, Corrected Recognition, Words, Corrected Recognition Difference, Criterion, Pictures, Criterion, Words, Recognition Accuracy, Answer Time, Picture Hits, Answer Time, Word Hits, Answer Time, Correct Rejections, Anticipations, Items Tested, Delay Arithmetic Accuracy and Delay. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Picture Superiority Task used in research?
31 papers from 2021 to 2025 name it in their title or abstract, about the same as from 2016 to 2020. That ranks it 297 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 Picture Superiority 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?
Pictures and words in each role (12 to 20) and arithmetic delay (seconds) (0 to 120). 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, with its words in the same language. 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.