Episodic Memory
Complex Figure Copy and Recall
Also called: complex figure test, complex figure drawing
Measure visual memory and drawing as participants copy a detailed line figure and draw it again from memory, scored by a stated rule with every stroke saved for raters.
8-12 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
- 566
- Evidence grade
- B
- Registered clinical studies
- 276
What it measures
A research measure of visual memory and drawing. Participants copy a detailed line figure and later draw it from memory, on figures ConductScience drew for this task. A stated geometric rule gives up to 2 points for each of 18 parts, and every stroke is saved so raters can score the drawings too.
Constructs
- Visual memory
- Visual construction
- Planning and organization
Research uses
- Memory and aging research
- Visual construction and planning research
- Within-study group comparison
How the paradigm works
The complex figure copy and recall test is a research measure of visual memory and drawing. Participants see a detailed line figure and an empty drawing area on one screen, side by side or stacked, and copy the figure with a mouse, a finger or a stylus. Nothing is said about memory before the copy. A number task, odd or even, then fills a delay and shows no figures, and participants draw the figure from memory in the same drawing area. A button undoes the last line, and Done asks for confirmation. A practice with a small figure of three parts comes first, with feedback. A study can choose figure A or B so that a repeat session can use the other one, set the limit per drawing from 120 to 600 seconds (standard 300), set the delay from 60 to 300 seconds (standard 180), and add a later recall after a wait of 20 to 40 minutes (standard off).
More about the paradigm
ConductScience drew both figures for this task. Each has no real-world meaning and the same make-up: an outer frame, 4 internal dividers, 6 attachments outside the frame and 7 small details inside it, 18 scored parts in all. The procedure follows the copy, recall and delayed recall conditions described by Shin and colleagues (2006), where delayed recall comes after 30 minutes. Here the delay before recall is shorter and the later recall is off by default, so a standard run takes about 10 minutes. When the later recall is on, it runs while the page stays open.
The score comes from a stated geometric rule, not a trained model, measured on the stored lines. Each drawing is first aligned to the figure with one size change, one displacement and a rotation of up to 15 degrees, by the best fit between ink and figure. Each part is then tested for ink that follows it in the same direction: 2 points for the right shape in the right place, 1 for the right shape misplaced or a distorted shape in place, 0.5 for a part both distorted and misplaced but still recognizable, and 0 otherwise. This is how the classic 18-unit method awards points (Zhang and colleagues, 2021), with definitions written for these figures, and the total runs from 0 to 36. On a figure about 100 units wide, the right shape means agreeing ink within 3 units of at least 75% of the part's length, and the right place means ink displaced by at most 7 units. Agreeing ink counts only in unbroken runs at least 5 units long, and a drawing whose credited parts use less than 60% of its ink keeps only its 2-point parts, so a scribble scores low. The main score is the recall score. The scores also give the copy score, retention as recall over copy, the points for each part, the share of extra ink, an organization measure (the share of the frame and dividers drawn before the first small detail, after Shin and colleagues, 2006, and Petilli and colleagues, 2021), drawing times, strokes and undos. The scores support research and do not diagnose.
Every pointer sample is saved with its time, stroke and pointer type, so a researcher can replay each drawing and raters can score the drawings too. A line that was undone stays in the record but not in the score. Browsers report a trackpad as a mouse.
Scores
- Recall Scorepoints
- Points for the drawing made from memory after the delay, from 0 to 36: up to 2 for each of the 18 parts, by the stated rule.
- Higher means more of the figure was drawn from memory, in the right shape and place
- Copy Scorepoints
- Points for the copy made with the figure in view, from 0 to 36, by the same rule.
- Higher means a more complete and accurate copy; below 18 recall is not interpreted
- Retention%
- Recall score as a percentage of the copy score. Empty when the copy scored 0.
- Higher means more of what was copied was kept; it can pass 100
- Later Recall Scorepoints
- Points for the later drawing from memory, from 0 to 36. Empty when no later recall was run.
- Higher means more of the figure was kept after the wait
- Copy Points per Partpoints
- Points for each of the 18 parts in the copy, keyed by part.
- Shows which parts earned or lost points
- Recall Points per Partpoints
- Points for each of the 18 parts in the recall drawing, keyed by part.
- Shows which parts earned or lost points
- Later Recall Points per Partpoints
- Points for each of the 18 parts in the later recall drawing, keyed by part.
- Shows which parts earned or lost points
- Recall Extra Ink%
- Share of the recall drawing's ink more than 6 figure units from every credited part.
- Higher means more lines that match no part of the figure
- Copy Extra Ink%
- Share of the copy's ink more than 6 figure units from every credited part.
- Higher means more lines that match no part of the figure; above 0.6 is flagged for review
- Later Recall Extra Ink%
- Share of the later recall drawing's ink more than 6 figure units from every credited part.
- Higher means more lines that match no part of the figure
- Copy Organization%
- Share of the frame and dividers drawn before the first small detail, in the copy.
- 1 when the main structure came first; lower when details came early
- Recall Organization%
- Share of the frame and dividers drawn before the first small detail, in the recall drawing.
- 1 when the main structure came first; lower when details came early
- Copy Timeseconds
- Time from the first to the last point of the copy.
- Informational; speed is not the aim of the task
- Recall Timeseconds
- Time from the first to the last point of the recall drawing.
- Informational; speed is not the aim of the task
- Later Recall Timeseconds
- Time from the first to the last point of the later recall drawing.
- Informational; speed is not the aim of the task
- Copy Strokescount
- Lines drawn in the copy and not undone.
- Informational
- Recall Strokescount
- Lines drawn in the recall drawing and not undone.
- 0 means an empty drawing, which is flagged for review
- Later Recall Strokescount
- Lines drawn in the later recall drawing and not undone.
- Informational
- Copy Undoscount
- Presses of the undo button during the copy.
- Informational
- Recall Undoscount
- Presses of the undo button during the recall drawing.
- Informational
- Later Recall Undoscount
- Presses of the undo button during the later recall drawing.
- Informational
- Copy Inkscore
- Length drawn in the copy, after alignment, as a share of the figure's length.
- About 1 for a full copy; below 0.25 makes the session invalid
- Focus Losses During the Delaycount
- Times the page was hidden or lost focus during the delay.
- Above 0 is flagged for review, because the figure could have been rehearsed or looked up
- Drawings Stopped by the Limitcount
- Drawings that ended at the time limit.
- Above 0 is flagged for review, because the drawing may be unfinished
- Figure B Shown0/1
- 1 when figure B was shown, 0 for figure A.
- Shows which figure the scores refer to
- Copy Alignmentscore
- Size change, rotation in degrees and displacement in figure units that aligned the copy with the figure.
- Informational; a scale near 1 means the drawing was about the size of the figure
- Recall Alignmentscore
- Size change, rotation in degrees and displacement in figure units that aligned the recall drawing with the figure.
- Informational
- Later Recall Alignmentscore
- Size change, rotation in degrees and displacement in figure units that aligned the later recall drawing with the figure.
- Informational
- Pointer Types%
- Share of the pointer samples from a mouse, a pen or touch, keyed by type. Browsers report a trackpad as a mouse.
- More than one type is flagged for review
Use and evidence
566 papers from 2021 to 2025 name this task in their title or abstract, up 4% compared with 2016 to 2020. That ranks it 42 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 118 |
| 2024 | 111 |
| 2023 | 106 |
| 2022 | 114 |
| 2021 | 117 |
| 2020 | 120 |
| 2019 | 120 |
| 2018 | 94 |
| 2017 | 114 |
| 2016 | 97 |
| 2015 | 106 |
| 2014 | 110 |
| 2013 | 94 |
| 2012 | 88 |
| 2011 | 79 |
| 2010 | 88 |
| 2009 | 65 |
| 2008 | 59 |
| 2007 | 48 |
| 2006 | 52 |
| 2005 | 31 |
| 2004 | 46 |
| 2003 | 42 |
| 2002 | 27 |
| 2001 | 28 |
| 2000 | 42 |
Use grade B
300 to 999 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
- 23
- Papers on reliability or validity
- 336
- Registered clinical studies using it as an outcome
- 276
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.
- Figures
- Two line figures ConductScience drew for this task, A (standard) or B for a repeat session, each with 18 scored parts: 1 frame, 4 dividers, 6 attachments and 7 small details.
- Copy
- The figure and the drawing area on one screen, side by side or stacked, drawn with a mouse, a finger or a stylus. Undo removes the last line, and Done asks for confirmation. Nothing is said about memory beforehand.
- Limit
- A drawing ends by itself after 300 seconds (standard, 120 to 600). There is no time pressure.
- Delay
- 180 seconds (standard, 60 to 300) of an unscored odd or even number task that shows no figures.
- Recall
- The figure drawn from memory in the same drawing area. A later recall after a wait of 20 to 40 minutes is off by default and runs with the page left open.
- Practice
- A small figure of 3 parts, then feedback on whether every part was drawn and placed.
- Scoring rule
- The drawing is aligned to the figure, then each part earns 2 points for the right shape in the right place, 1 for the right shape misplaced or a distorted shape in place, 0.5 for both but recognizable, and 0 otherwise, for 36 at most. Ink near no credited part counts as extra ink.
- Data
- Every pointer sample with its time, stroke and pointer type, up to 250 per second, with every undo, so each drawing can be replayed.
The score comes from a stated rule, and the strokes are saved so raters can score the drawings too. Mouse, finger and stylus drawings differ, so each result records the pointer used and no percentiles are given. 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: complex_figure_copy, complex_figure_recall and complex_figure_delayed_recall from the scored run, one per drawing. A drawing has no correct option, so accuracy is not judged against chance. A session with no copy drawing, or a copy with less ink than 25% of the figure's length, is invalid. A copy score below 18 of 36, a page that lost focus during the delay, an empty recall, a drawing stopped by the time limit, a drawing with more than 60% extra ink and drawings made with more than one kind of pointer are flagged for review.
Why these rules
Every score comes from the drawn lines, so a copy with almost no ink has nothing to compare recall with. A copy below 18 points suggests that the figure was not fully taken in, so a low recall score cannot be read as forgetting. A hidden tab lets the study, delay or response timers run without the participant, so a percentile is withheld after one. Mouse, finger and pen drawings differ, so a session that mixes them is flagged.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Figure | Figure A (standard) | Figure A (standard), Figure B |
| Limit per drawing (seconds) | 300 | 120 to 600 |
| Delay before recall (seconds) | 180 | 60 to 300 |
| Later recall (minutes) | Off (standard) | Off (standard), 20 minutes, 25 minutes, 30 minutes, 35 minutes, 40 minutes |
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 and tablets. 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 Complex Figure Copy and Recall task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Shin, Park, Park, Seol & Kwon, 2006; Zhang et al., 2021; Petilli et al., 2021), with the standard settings (figure: Figure A; limit per drawing (seconds): 300; delay before recall (seconds): 180; later recall (minutes): Off). The main score was recall score (points for the drawing made from memory after the delay, from 0 to 36: up to 2 for each of the 18 parts, by the stated rule).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Shin, M.-S., Park, S.-Y., Park, S.-R., Seol, S.-H., & Kwon, J. S. (2006). Nature Protocols, 1(2), 892-899. https://doi.org/10.1038/nprot.2006.115The copy, immediate recall and delayed recall conditions, and scoring of location, accuracy and organization.
- Zhang, X., Lv, L., Min, G., Wang, Q., Zhao, Y., & Li, Y. (2021). Overview of the complex figure test and its clinical application in neuropsychiatric disorders, including copying and recall. Frontiers in Neurology, 12, 680474. https://doi.org/10.3389/fneur.2021.680474The point scale of 18 units with up to 2 points each, and equivalent figures for a repeat session.
- Petilli, M. A., Daini, R., Saibene, F. L., & Rabuffetti, M. (2021). Scientific Reports, 11(1), 14895. https://doi.org/10.1038/s41598-021-94247-9Tablet recording, global alignment of a drawing to a template, and measures of whether main structure comes before details.
- Webb, S. S., Moore, M. J., Yamshchikova, A., Kozik, V., Duta, M. D., Voiculescu, I., & Demeyere, N. (2021). Validation of an automated scoring program for a digital complex figure copy task within healthy aging and stroke. Neuropsychology, 35(8), 847-862. https://doi.org/10.1037/neu0000748An automated program compared with two trained raters, which shows how a rater comparison is reported.
- Langer, N., Weber, M., Hebling Vieira, B., et al. (2024). eLife, 13, RP96017. https://doi.org/10.7554/eLife.96017Human ratings of 18 elements scored 0 to 2, and the disagreement among raters.
Source papers
- Shin MS, Park SY, Park SR, Seol SH, Kwon JS (2006). Nature Protocols, 1(2):892-899.Paradigm source
- Zhang X, Lv L, Min G, Wang Q, Zhao Y, Li Y (2021). Overview of the complex figure test and its clinical application in neuropsychiatric disorders, including copying and recall. Frontiers in Neurology, 12:680474.Paradigm source
- Petilli MA, Daini R, Saibene FL, Rabuffetti M (2021). Scientific Reports, 11(1):14895.Paradigm source, Adults
- Webb SS, Moore MJ, Yamshchikova A, Kozik V, Duta MD, Voiculescu I, Demeyere N (2021). Validation of an automated scoring program for a digital complex figure copy task within healthy aging and stroke. Neuropsychology, 35(8):847-862.Paradigm source, Adults
- Langer N, Weber M, Hebling Vieira B, et al. (2024). eLife, 13:RP96017.Paradigm source
Questions
How long does the Complex Figure Copy and Recall take?
About 8 to 12 minutes, as listed in the test library.
What device does a participant need?
Untimed test: runs on desktops, laptops and tablets. The browser must be Chrome, Edge, Firefox or Safari.
What scores does it produce?
Recall Score, Copy Score, Retention, Later Recall Score, Copy Points per Part, Recall Points per Part, Later Recall Points per Part, Recall Extra Ink, Copy Extra Ink, Later Recall Extra Ink, Copy Organization, Recall Organization, Copy Time, Recall Time, Later Recall Time, Copy Strokes, Recall Strokes, Later Recall Strokes, Copy Undos, Recall Undos, Later Recall Undos, Copy Ink, Focus Losses During the Delay, Drawings Stopped by the Limit, Figure B Shown, Copy Alignment, Recall Alignment, Later Recall Alignment and Pointer Types. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Complex Figure Copy and Recall used in research?
566 papers from 2021 to 2025 name it in their title or abstract, up 4% compared with 2016 to 2020. That ranks it 42 of 395 test families in the ConductCognition test census (use grade B). Source: ConductCognition test census 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Complex Figure Copy and Recall have?
In the published research, 23 papers report its test-retest reliability and 276 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?
Figure (Figure A (standard), Figure B), limit per drawing (seconds) (120 to 600), delay before recall (seconds) (60 to 300) and later recall (minutes) (Off (standard), 20 minutes, 25 minutes, 30 minutes, 35 minutes, 40 minutes). 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.