Working Memory
Partial Report Task
Also called: Sperling task, partial report paradigm, iconic memory task, whole and partial report
Measure iconic memory: twelve letters flash for 50 ms, and participants type all of them or only the row that arrows point to, up to a second later.
8-11 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
- 76
- Evidence grade
- C
- Registered clinical studies
- 0
What it measures
A measure of iconic memory, the brief visual store that holds more letters than a person can report. Twelve consonants flash for about 50 ms in three rows of four. Participants first type all the letters, then, on most trials, only the row that arrows point to; the arrows come at once or up to a second after the flash. The score is the number of letters available just after the flash.
Constructs
- Iconic memory
- Visual sensory memory
- Span of immediate memory
- Visual attention
Research uses
- Visual memory research
- Aging research
- Within-study group comparison
How the paradigm works
The partial report task measures iconic memory: for a fraction of a second after a brief display, people can report more of it than they can hold in mind. Twelve consonants flash for about 50 ms in three rows of four on a dark panel, with a dim dot at the center that stays on throughout. In a whole report, participants type all twelve letters into a grid of boxes, guessing where unsure. In a partial report, two arrows then appear beside one row, and participants type only that row's four letters. Because the row is cued only after the letters are gone, the share right in the cued row samples what was still available when the cue came (Sperling, 1960). The arrows take the place of Sperling's tones, after the bar marker of Averbach and Coriell (1961), and stay clear of the letters.
More about the paradigm
The standard session has 2 practice whole reports, 8 whole reports, 4 practice partial reports and 60 partial reports: 12 at each of five cue delays, 0, 150, 300, 500 and 1,000 ms after the letters go off, with each row cued equally often at each delay, in an order drawn from the session's seed (8 to 16 at each delay as a setting). Every display time is fixed before the session and timed to screen frames, and the measured times are saved with every trial; nothing a participant answers changes a later display.
A letter counts when it is typed in its right place, Sperling's score. The main score is the number of letters available just after the letters go off: the share of cued letters right at the 0 ms delay times the 12 letters of the array. The same estimate at each later delay shows how much of the display fades over a second, and the letters right in a whole report give the span of immediate memory. The partial report advantage, available letters at 0 ms less the whole report, and the letters lost by 1,000 ms are also reported. Sperling's trained observers had about 9.8 of 12 letters available with the cue just before the letters, about 4.6 with the cue a second after, and a whole report of about 4.3. Letter keys before the cue, or within 100 ms of it, are counted as anticipations and not taken. Scores support research and do not diagnose.
Scores
- Letters Available at 0 msitems
- Share of cued letters typed in their right place with the cue as the letters go off, times the 12 letters of the array.
- Higher is better
- Letters Available at 150 msitems
- The same with the cue 150 ms after the letters go off.
- Higher is better
- Letters Available at 300 msitems
- The same with the cue 300 ms after the letters go off.
- Higher is better
- Letters Available at 500 msitems
- The same with the cue 500 ms after the letters go off.
- Higher is better
- Letters Available at 1,000 msitems
- The same with the cue 1,000 ms after the letters go off.
- Higher is better
- Whole Reportitems
- Mean letters typed in their right place when all 12 are asked for.
- Higher is better
- Partial Report Advantageitems
- Letters available at 0 ms less the whole report.
- Higher means more is briefly held than can be reported
- Letters Lost by 1,000 msitems
- Letters available at 0 ms less those available at 1,000 ms.
- Informational; how much of the display fades over a second
- Median Answer Timems
- Median time from the cue to Enter on partial reports.
- Informational; there is no time limit
- Anticipationscount
- Letter keys pressed before the cue or under 100 ms after it. They are not taken.
- Lower is better
- Frame Intervalms
- Median time between screen frames on the participant's device.
- Informational
- Displays on Time%
- Share of trials whose letters and cue delay ran within one frame of plan.
- Below 0.9 is flagged for review
- Planned Trialstrials
- Whole and partial reports the settings planned.
- Informational
- Total Trialstrials
- Whole and partial reports answered in the scored run.
- Fewer than planned is flagged for review
Use and evidence
76 papers from 2021 to 2025 name this task in their title or abstract, down 19% compared with 2016 to 2020. That ranks it 214 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 14 |
| 2024 | 14 |
| 2023 | 23 |
| 2022 | 12 |
| 2021 | 13 |
| 2020 | 14 |
| 2019 | 16 |
| 2018 | 17 |
| 2017 | 22 |
| 2016 | 25 |
| 2015 | 28 |
| 2014 | 21 |
| 2013 | 23 |
| 2012 | 32 |
| 2011 | 26 |
| 2010 | 29 |
| 2009 | 15 |
| 2008 | 7 |
| 2007 | 6 |
| 2006 | 4 |
| 2005 | 13 |
| 2004 | 8 |
| 2003 | 4 |
| 2002 | 4 |
| 2001 | 2 |
| 2000 | 3 |
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
- 3
- Papers on reliability or validity
- 18
- 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.
- Display
- 12 different consonants in three rows of four, bold capitals in a 56 px font in columns 80 px apart and rows 76 px apart, on a dark panel of 560 by 320 px with a dim dot at the center.
- Timing
- The dot alone for 500 ms, the letters for 3 frames of a 60 Hz screen (about 50 ms), then the dot until the cue: at once in a whole report, or 0, 150, 300, 500 or 1,000 ms after the letters go off in a partial report. Displays switch on screen frames, and their shown length is saved.
- Cue
- Two white arrows 64 px beyond the first and last letters of the cued row, which stay until the answer.
- Answer
- The letters typed into a grid of boxes under the panel, all 12 or the cued row's 4, every box filled with a guess where unsure, then Enter, with no time limit. Letter keys before the cue or within 100 ms of it are counted and not taken.
- Length
- 8 whole reports, then 12 partial reports at each delay (8 to 16), 60 in all, in a seeded order with each row cued equally often at each delay and a rest after every 30.
- Practice
- 2 whole reports and 4 partial reports, at 0, 150, 500 and 1,000 ms, each followed by the letters again with the typed ones marked right or wrong for 2,500 ms.
- Scores
- Letters available at each delay (the share of cued letters typed in their right place, times 12), the whole report, the partial report advantage and the letters lost by 1,000 ms.
Sperling (1960) cued the row with a high, middle or low tone and trained a few observers over many sessions. This version cues the row with arrows, because the tests have no sound, and runs in one session with fixed delays, so its estimates at each delay are noisier than his.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: partial_report_test from the scored run, one answer for each letter asked, 12 in a whole report and 4 in a partial report. A letter counts when it is typed in its right place; a guess is right one time in 20, the 20 consonants. Fewer right letters than chance invalidates and not more than chance is flagged. One letter typed in 90% of the places invalidates when the letters are not right more often than chance; the letters shown are spread evenly over the consonants. A first letter typed under 200 ms after the cue on at least half the answers invalidates. Letter keys before the cue, or within 100 ms of it, are not taken. A run short of its planned trials, or under 90% of trials with the letters and the cue delay within one frame of plan, is flagged and its percentile withheld.
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). 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 letters show for about 50 ms and the cue delays are what the decay is measured over, so trials whose displays ran late change what was measured. 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 |
|---|---|---|
| Trials at each cue delay | 12 | 8 to 16 |
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.
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Partial Report Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Sperling, 1960; Averbach and Coriell, 1961), with the standard settings (trials at each cue delay: 12). The main score was letters available at 0 ms (share of cued letters typed in their right place with the cue as the letters go off, times the 12 letters of the array).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Sperling, G. (1960). The information available in brief visual presentations. Psychological Monographs: General and Applied, 74(11), 1-29. https://doi.org/10.1037/h0093759
- Averbach, E., & Coriell, A. S. (1961). Short-term memory in vision. Bell System Technical Journal, 40(1), 309-328. https://doi.org/10.1002/j.1538-7305.1961.tb03987.xA visual marker in place of a tone, kept clear of the letters.
- Lu, Z.-L., Neuse, J., Madigan, S., & Dosher, B. A. (2005). Fast decay of iconic memory in observers with mild cognitive impairments. Proceedings of the National Academy of Sciences, 102(5), 1797-1802. https://doi.org/10.1073/pnas.0408402102Iconic memory decay measured by partial report in older adults with and without mild cognitive impairment.
Source papers
- Sperling (1960): whole and partial report of consonant arrays, letters available at each cue delay. Averbach and Coriell (1961): a visual marker in place of a tone. Lu, Neuse, Madigan and Dosher (2005): iconic memory decay in older adults with and without mild cognitive impairment.Paradigm source, Adults
Questions
How long does the Partial Report Task take?
About 8 to 11 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?
Letters Available at 0 ms, Letters Available at 150 ms, Letters Available at 300 ms, Letters Available at 500 ms, Letters Available at 1,000 ms, Whole Report, Partial Report Advantage, Letters Lost by 1,000 ms, Median Answer Time, Anticipations, Frame Interval, Displays on Time, Planned Trials and Total Trials. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Partial Report Task used in research?
76 papers from 2021 to 2025 name it in their title or abstract, down 19% compared with 2016 to 2020. That ranks it 214 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 Partial Report Task have?
In the published research, 3 papers report its test-retest reliability and 0 registered clinical studies use it as an outcome (evidence grade C in the ConductCognition test census).
Which settings can a researcher change?
Trials at each cue delay (8 to 16). 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.