Reasoning
Shape Analogies Task
Also called: geometric analogies, figural analogies, visual analogies, A:B::C:D analogies, proportional analogies, analogical reasoning task
Measure analogical reasoning: a figure changes into a second one, and participants choose what a third figure becomes after the same changes.
7-10 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
- 83
- Evidence grade
- C
- Registered clinical studies
- 2
What it measures
Measures analogical reasoning with figures generated for each session. A figure changes into a second one; participants choose the figure a third one becomes after the same changes. Puzzles have one, two or three changes in size, number, fill, turning, flipping or the position of a dot.
Constructs
- Analogical reasoning
- Fluid reasoning
- Relational mapping
- Visuospatial transformation
Research uses
- Fluid reasoning research
- Development and aging studies
- Education research
- Within-study group comparison
How the paradigm works
The Shape Analogies Task measures analogical reasoning with figures: A is to B as C is to which? Each puzzle shows a figure that changes into a second figure, and a third figure beside an empty panel. The participant works out the change and chooses, from eight figures, the one the third figure becomes after the same change.
More about the paradigm
Puzzles are generated for each session from a seed saved with the trial data. A figure is one to three copies of a shape, with a dot in one corner of its panel. Between the first two figures the shape can turn a quarter or half turn, flip left to right, grow or shrink, gain or lose a copy or take another fill (outline, solid or striped), or the dot can move to another corner. The third figure has a different shape, so the answer is never a copy of the second. Puzzles have one, two or three changes and come in that order, 8 of each in the standard session of 24, after three practice puzzles that show the right answer. There is no time limit.
The eight choices combine a right or a wrong value of each changed feature (and, in one-change puzzles, of one feature that does not change), so every wrong choice gets at least one feature wrong and the answer cannot be found from the choices alone. Accuracy is the main score. Accuracy and the median time to a right answer are also reported for each number of changes, with the share of the changes that the chosen figures got right. Scores support research and do not diagnose.
Scores
- Accuracy%
- Share of answered test puzzles solved.
- Higher is better
- One-Change Accuracy%
- Accuracy on puzzles with one change.
- Higher is better
- Two-Change Accuracy%
- Accuracy on puzzles with two changes.
- Higher is better
- Three-Change Accuracy%
- Accuracy on puzzles with three changes.
- Higher is better
- Median Correct Timems
- Median time from a puzzle appearing to a right answer. Answers under 100 ms are left out.
- Informational
- Median Correct Time, Three Changesms
- The same for puzzles with three changes; one- and two-change times are also reported.
- Informational
- Changes Applied%
- Share of all the puzzles' changes that the chosen figures got right, so a choice that makes some of the changes gets part credit.
- Higher is better
- Total Correctcount
- Test puzzles solved.
- Higher is better
- Puzzles Answeredcount
- Test puzzles answered.
- Informational
- Answers Under 100 mscount
- Answers given less than 100 ms after the puzzle appeared.
- Lower is better
- Practice Accuracy%
- Share of the three practice puzzles solved.
- Informational
Use and evidence
83 papers from 2021 to 2025 name this task in their title or abstract, down 22% compared with 2016 to 2020. That ranks it 202 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 21 |
| 2024 | 11 |
| 2023 | 13 |
| 2022 | 19 |
| 2021 | 19 |
| 2020 | 25 |
| 2019 | 12 |
| 2018 | 24 |
| 2017 | 15 |
| 2016 | 30 |
| 2015 | 19 |
| 2014 | 14 |
| 2013 | 18 |
| 2012 | 13 |
| 2011 | 16 |
| 2010 | 11 |
| 2009 | 17 |
| 2008 | 13 |
| 2007 | 12 |
| 2006 | 11 |
| 2005 | 9 |
| 2004 | 10 |
| 2003 | 6 |
| 2002 | 3 |
| 2001 | 4 |
| 2000 | 5 |
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
- 5
- Papers on reliability or validity
- 65
- Registered clinical studies using it as an outcome
- 2
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.
- Puzzle
- A figure changes into a second figure; the participant chooses, from 8 figures, the one a third figure becomes after the same changes.
- Figures
- Drawn by our code: 1 to 3 copies of one of 7 shapes, none with a mirror line or a turn that maps it onto itself, and a dot in one corner. The third figure always has a different shape than the first.
- Changes
- A quarter turn either way or a half turn, a flip left to right, one size step, one copy more or fewer, another fill (outline, solid or striped), or the dot moving one corner either way or across.
- Puzzles
- 24 (standard; 12 to 36), 8 each with 1, 2 and 3 changes, in that order. Generated for each session from a seed saved with the data.
- Choices
- 8 per puzzle: combinations of right and wrong values of the features that change (and, with one change, of one that does not). Every wrong choice gets at least one feature wrong, and no right value is more common than all the wrong ones, so the answer cannot be found from the choices alone. The answer is in each place equally often.
- Timing
- No time limit; 300 ms between puzzles.
- Practice
- 3 puzzles, each shown again with the answer marked for 2,500 ms.
Every figure and puzzle is generated by ConductScience code; no published test item or figure is used. The format is the geometric analogy of the reasoning literature, with puzzles ordered by their number of changes and choices built as in Blum and Holling (2018). 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: analogies_item test rows from the scored run, eight choices per puzzle (chance 1 in 8), the answer balanced across the eight places. Accuracy below chance invalidates and not above chance is flagged. One place on 90% of puzzles invalidates when accuracy is not above chance. Answers under 200 ms on at least half the puzzles invalidate. Untimed, so nothing can be missed. A hidden tab is flagged.
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.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Number of puzzles | 24 | 12 to 36 |
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.
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Shape Analogies Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Evans TG, 1964; Mulholland TM, Pellegrino JW, Glaser R, 1980; Blum D, Holling H, 2018), with the standard settings (number of puzzles: 24). The main score was accuracy (share of answered test puzzles solved).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Mulholland, T. M., Pellegrino, J. W., & Glaser, R. (1980). Components of geometric analogy solution. Cognitive Psychology, 12(2), 252-284. https://doi.org/10.1016/0010-0285(80)90011-0The number of transformations as a source of difficulty.
- Blum, D., & Holling, H. (2018). Automatic generation of figural analogies with the IMak package. Frontiers in Psychology, 9, 1286. https://doi.org/10.3389/fpsyg.2018.01286Puzzles generated from rules, ordered by the number of rules, with eight choices that combine right and wrong values of each rule.
- Evans, T. G. (1964). A heuristic program to solve geometric-analogy problems. Proceedings of the April 21-23, 1964, Spring Joint Computer Conference (AFIPS '64), 327. https://doi.org/10.1145/1464122.1464156The figure A is to figure B as figure C is to which format.
- Arendasy, M. E., & Sommer, M. (2013). Reducing response elimination strategies enhances the construct validity of figural matrices. Intelligence, 41(4), 234-243. https://doi.org/10.1016/j.intell.2013.03.006Choices that cannot be ruled out without solving the puzzle.
Source papers
- Mulholland TM, Pellegrino JW, Glaser R (1980). Components of geometric analogy solution. Cognitive Psychology, 12(2):252-284.Paradigm source, Adults
- Blum D, Holling H (2018). Automatic generation of figural analogies with the IMak package. Frontiers in Psychology, 9:1286.N 307, Adults, mean age 28.6
Questions
How long does the Shape Analogies Task take?
About 7 to 10 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?
Accuracy, One-Change Accuracy, Two-Change Accuracy, Three-Change Accuracy, Median Correct Time, Median Correct Time, Three Changes, Changes Applied, Total Correct, Puzzles Answered, Answers Under 100 ms and Practice Accuracy. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Shape Analogies Task used in research?
83 papers from 2021 to 2025 name it in their title or abstract, down 22% compared with 2016 to 2020. That ranks it 202 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 Shape Analogies Task have?
In the published research, 5 papers report its test-retest reliability and 2 registered clinical studies use it as an outcome (evidence grade C in the ConductCognition test census).
Which settings can a researcher change?
Number of puzzles (12 to 36). 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.