Executive Function
Hat Stand Game
Also called: planning game for children, touch tower puzzle game
Measure how children aged 5 to 12 plan moves ahead: three hats stand on three posts, and the child moves one at a time to make the stand look like a picture, in as few moves as possible.
6-9 minUntimed
Free to start: 10 free participants per account, then $0.99 per participant for academic research, or any yearly plan. Try it here first, with no account.
- Papers, 2021 to 2025
- 495
- Evidence grade
- A
- Registered clinical studies
- 247
Figures for the Tower Planning test, the published test this game is compared with.
How these are countedWhat it measures
A planning game for research with children aged 5 to 12, with an adult reading the screens aloud to children under 8. Three hats stand on three posts. The child moves one hat at a time to make the stand look like a picture, in as few moves as possible. The score is the share of problems solved in the fewest moves. It is a raw research measure with no percentile.
Constructs
- Planning
- Planning moves ahead
- Problem solving
Research uses
- Research on planning moves ahead in young and school-age children
How the paradigm works
The Hat Stand Game is a planning game for research with children aged 5 to 12. Three hats, a cap, a bobble hat and a tall hat, stand on three posts that hold 3, 2 and 1 hats, the world of the Tower of London (Shallice, 1982). A picture shows how the stand should look. The child moves one hat at a time by touching it and then touching the post it goes to, until the stand looks like the picture, in as few moves as possible. Only the top hat of a post can move, and a post holds only so many hats. Each problem shows one dot for each of the fewest moves, and the dots fill as moves are made. A problem ends at the goal or after the fewest moves plus 4, and there is no time limit.
More about the paradigm
The game has 16 problems, four each at 2, 3, 4 and 5 fewest moves in that order, found by a breadth-first search over all 36 arrangements of the three hats. None of them is in the class of a Tower Planning puzzle under any relabelling of the hats. At 4 and 5 moves, half of the problems need the counter-intuitive move: a hat has to leave the post the goal puts it on. In four-year-olds, problems that needed search ahead were the hard ones, while five-year-olds solved both kinds equally well (Kaller and colleagues, 2008), and the two towers were not equivalent in children of about 5 (Bull, Espy and Senn, 2004). A demonstration with an animated hand and two practice problems come first, and the hats that differ from the picture glow after a practice problem that ends unsolved; the practice runs once more unless both problems are solved. Two forms serve studies, a list and its mirror in a second order, and a third form with the hats relabelled serves Try it now. The hats are drawn for this game, and no ball, disc or peg board is shown.
The paper counts and grades on this page are those of the Tower Planning test, the published test the game is compared with, not the game's own. The validation plan also compares the game with the Tower of Hanoi. The main score is the share of problems solved in the fewest moves, read within one device class and one input. The share on the 4- and 5-move problems is where a child who plans nothing shows, since a child who moves each hat to its goal place when it can still solves the short ones.
Scores
- Solved in the Fewest Moves%
- The share of scored problems whose moves reach the goal in exactly the fewest moves. Each problem is checked by replaying its recorded moves, so a problem counts only if its moves really reach the goal. Empty below 12 scored problems.
- Higher means more problems solved in the fewest moves; compared within a study and within one device and input
- Fewest Moves on 4- and 5-Move Problems%
- The same share over the 8 problems that take 4 and 5 moves.
- Secondary: this is where planning shows, since a child who moves each hat to its goal place when it can solves few of them
- Problems Solvedcount
- Scored problems whose moves reach the goal within the move limit, the fewest moves plus 4, with or without extra moves.
- Higher means more problems brought to the goal
- Excess Movesmoves
- Moves beyond the fewest, averaged over the scored problems. A problem that ends unsolved counts 4 excess moves, its move limit.
- Lower is better
- Median Time to the First Movems
- Median time from a problem appearing to its first move, over the scored problems.
- Informational
- Refused Tapscount
- Taps on a hat that is not on top and on a post that is full, which do nothing.
- Quality check
- Problems Scoredcount
- Test problems scored, out of 16. A problem with a second finger or a palm, a hidden tab or a resized window is left out.
- Quality check: the score needs 12
- Share of Key Moves%
- Share of the scored problems whose first move was made with a key.
- Informational: the touch-versus-keyboard comparison reads it
- Share of Touch Moves%
- Share of the scored problems whose first move was made by touch.
- Informational: the touch-versus-keyboard comparison reads it
Use and evidence
495 papers from 2021 to 2025 name this task in their title or abstract, down 20% compared with 2016 to 2020. That ranks it 60 of 711 test families in our census. These counts belong to the Tower Planning test, the published test this game is compared with. This game is a new measure, so its own reliability and norms come from use. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 96 |
| 2024 | 93 |
| 2023 | 100 |
| 2022 | 106 |
| 2021 | 100 |
| 2020 | 96 |
| 2019 | 169 |
| 2018 | 113 |
| 2017 | 122 |
| 2016 | 116 |
| 2015 | 117 |
| 2014 | 102 |
| 2013 | 79 |
| 2012 | 94 |
| 2011 | 93 |
| 2010 | 78 |
| 2009 | 73 |
| 2008 | 63 |
| 2007 | 103 |
| 2006 | 61 |
| 2005 | 57 |
| 2004 | 66 |
| 2003 | 46 |
| 2002 | 40 |
| 2001 | 35 |
| 2000 | 37 |
Use grade B
300 to 999 papers from 2021 to 2025.
Evidence grade A
At least 25 papers report its test-retest reliability and at least 25 registered clinical studies use it as an outcome.
- Papers on test-retest reliability
- 30
- Papers on reliability or validity
- 190
- Registered clinical studies using it as an outcome
- 247
Measured on ConductCognition
Pooled across the studies that feed our platform norms, by device group. A figure needs 50 or more people.
No pooled figure yet.
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 the Tower Planning test, the published test this game is compared with, not the game's own.
Source: ConductCognition test census 0.58.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Hats and posts
- Three hats drawn in code, a cap with a visor, a bobble hat and a tall hat, stand on three posts that hold 3, 2 and 1 hats, the world of the Tower of London. The hats differ by outline and by fill (a slate grey, a violet and a cream), at least 20 apart in CIEDE2000 in normal, protan, deutan and tritan vision, so no colour vision is needed. The game look adds a soft hall behind everything, the same for every problem; the plain look shows the same hats on a plain ground. A hat is at least 72 px wide, a post's tap area at least 96 px, and the picture and the three posts fit a 360 by 640 screen.
- Problems
- 16 problems a form, 4 each at 2, 3, 4 and 5 fewest moves, in that order, found by a breadth-first search over the 36 arrangements of the hats. At 4 and at 5 moves, 2 of the 4 need a move that takes a hat off the post the goal puts it on, and 2 do not. From 3 moves up each count holds a tower goal and a flat goal. None is in the class of a Tower Planning puzzle under any relabelling of the hats, and no two share a class.
- Trial
- A picture shows how the stand should look, the board shows the hats now, and one dot for each of the fewest moves fills as moves are made. The child taps a top hat, which lifts it, then a post with room, which puts it there and counts a move. A tap on the lifted hat puts it back with no move. A problem ends when the stand matches the picture, with a soft glow and 1,000 ms, or after the fewest moves plus 4, so every problem ends with a move. There is no time limit. A demonstration with an animated hand and 2 practice problems come first; a practice problem that ends unsolved makes the hats that differ from the picture glow, and the practice runs once more unless both problems are solved. A rest follows problem 8.
- Answers
- Only the top hat of a post can move. A tap on a hat that is not on top, or on a post that is full, does nothing and is counted. A second tap on the same post within 300 ms changes nothing and is counted. A second finger or a palm, a hidden tab or a window that changed size leaves that problem out of the scores. A keyboard works by Tab to a post, then Enter or Space.
- Forms
- 2 alternate forms for studies: a list and its mirror, with the posts read right to left, in a fixed second order within each count. A third form, with the hats relabelled, serves Try it now. Each participant gets the forms in an order drawn once for them, and each session takes the next form. Two shortcuts are bounded by construction: a mover who moves a hat to its goal place whenever it can, and otherwise at random, is expected to solve at most 0.45 of the 16 in the fewest moves, and a mover at random solves under 5% of the 4- and 5-move problems.
- Scores
- The share of scored problems solved in the fewest moves (the main score), checked by replaying every problem's moves on the board. The same share over the 8 problems of 4 and 5 moves, the problems solved, the excess moves (an unsolved problem counts the fewest plus 4), the median time to the first move, refused taps, the problems scored and the shares of first moves made by key and by touch come with it. A run with fewer than 12 scored problems has no score.
- Devices
- Computers, tablets and phones of at least 360 by 640 px. No answer is timed, so no timing check applies. Use a tablet or a computer for a child under 7. Keep each child on one device class and one input at every visit. For a child under 8, an adult sits beside the child, reads each screen aloud and does not point to hats or move them.
The world is the Tower of London's, three posts that hold 3, 2 and 1 (Shallice 1982). A computerized Tower of London gave greatest lower bound reliability of .76 and .80 in 178 typically developing children aged 6 to 13, and its four- and five-move problems carried the performance variability (Unterrainer and colleagues 2019). Problems that needed search ahead were the hard ones for four-year-olds (Kaller and colleagues 2008), and the Tower of London and the Tower of Hanoi were not interchangeable in children of about 5 (Bull, Espy and Senn 2004). The hats, the stand, the 16 problems and their orders, the move limit of the fewest moves plus 4, the dots, the forms, the words and the looks are our own. This is a new measure for research with children aged 5 to 12: its reliability and norms come from use, the figures published in those studies are not its own, and it computes no percentiles. For a retest 7 to 28 days later, add a second visit to the same study.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: hat_stand_game test problems whose moves replay on the board from the problem's start (a problem with a second finger or a palm, a hidden tab or a resized window is left out). Fewer than 12 of the 16 problems scored invalidates; 8 or more unsolved problems whose first move came under 500 ms are flagged. A problem ends only at its goal or its move limit, so none ends without a move.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. A child who plans nothing still solves the short problems, so the score reads the share solved in the fewest moves, and 12 problems keep three at each length. A first move under 500 ms on a problem that then ended unsolved at its move limit is too quick to have read the picture, which is a ConductScience choice, not a published limit. A problem cannot end without a move, so no rule counts missed problems.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Look | Game | Game, Plain |
Every plan can change the look, which never changes how a result is scored.
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 Hat Stand Game in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Shallice, 1982), with the standard settings (look: Game). The main score was solved in the fewest moves (the share of scored problems whose moves reach the goal in exactly the fewest moves).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Shallice, T. (1982). Specific impairments of planning. Philosophical Transactions of the Royal Society of London. B, Biological Sciences, 298(1089), 199-209. https://doi.org/10.1098/rstb.1982.0082An information-processing model of planning, tested with a task from artificial intelligence on which patients with left anterior lesions had a specific deficit (abstract read). The game takes the task's world, three posts that hold 3, 2 and 1, not its patient findings.
- Unterrainer, J. M., Rahm, B., Loosli, S. V., Rauh, R., Schumacher, L. V., Biscaldi, M., & Kaller, C. P. (2019). Psychometric analyses of the Tower of London planning task reveal high reliability and feasibility in typically developing children and child patients with ASD and ADHD. Child Neuropsychology, 26(2), 257-273. https://doi.org/10.1080/09297049.2019.1642317A computerized Tower of London gave greatest lower bound reliability of .76 and .80 in 178 typically developing children aged 6 to 13, and a subset of four- and five-move problems showed enough variability to reflect the whole set (abstract read). The game takes four problems at each of 2, 3, 4 and 5 fewest moves.
- Kaller, C. P., Rahm, B., Spreer, J., Mader, I., & Unterrainer, J. M. (2008). Thinking around the corner: The development of planning abilities. Brain and Cognition, 67(3), 360-370. https://doi.org/10.1016/j.bandc.2008.02.003In four-year-olds, planning accuracy fell in problems that needed search ahead, while five-year-olds solved both kinds of problem equally well, and the order of the subgoals changed the times to the first move (abstract read). The game balances the problems that need search ahead, half of those at 4 and at 5 moves.
- Bull, R., Espy, K. A., & Senn, T. E. (2004). A comparison of performance on the Towers of London and Hanoi in young children. Journal of Child Psychology and Psychiatry, 45(4), 743-754. https://doi.org/10.1111/j.1469-7610.2004.00268.xIn 118 children of about 5, the two towers were not interchangeable, and trials that needed moves in the counter-intuitive direction relative to the end-state goal were the complex ones (abstract read). The game is compared with the Tower of Hanoi without treating the two as the same.
Source papers
Questions
How long does the Hat Stand Game take?
About 6 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?
Solved in the Fewest Moves, Fewest Moves on 4- and 5-Move Problems, Problems Solved, Excess Moves, Median Time to the First Move, Refused Taps, Problems Scored, Share of Key Moves and Share of Touch Moves. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Tower Planning test used in research?
These counts belong to the Tower Planning test, the published test this game is compared with. This game is a new measure, so its own reliability and norms come from use. 495 papers from 2021 to 2025 name the task in their title or abstract, down 20% compared with 2016 to 2020. That ranks it 60 of 711 test families in the ConductCognition test census (use grade B). Source: ConductCognition test census 0.58.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Tower Planning test have?
These counts belong to the Tower Planning test, the published test this game is compared with. This game is a new measure, so its own reliability and norms come from use. In the published research, 30 papers report the task's test-retest reliability and 247 registered clinical studies use it as an outcome (evidence grade A in the ConductCognition test census).
Does it report a percentile?
No. Scores and quality flags only.
Which settings can a researcher change?
Look (Game, Plain). Every plan can change the look, which never changes how a result is scored.
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.
Related tests