Reasoning
Tile Row Game
Also called: figural series game, pattern completion game, tile sequence puzzle game, nonverbal reasoning game
Measure how many figural series puzzles adults solve: a row of tiles changes by rules, and players pick the tile that comes next from 4 or 8 answer tiles.
12-16 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
- 1,106
- Evidence grade
- A
- Registered clinical studies
- 272
Figures for the Matrix Reasoning test, the published test this game is compared with.
How these are countedWhat it measures
Measures how many figural series puzzles adults solve: a row of tiles changes by rules, and the player finds the tile that comes next.
Constructs
- Fluid reasoning
- Rule induction
- Figural reasoning
- Reasoning with new problems
Research uses
- Research on reasoning with new problems in adults
How the paradigm works
The Tile Row Game measures how many figural series puzzles an adult solves. A row of three tiles changes from tile to tile by rules, with an empty place after it, and the player finds the tile that comes next among 4 or 8 answer tiles. A tile is one of six drawn shapes, turned by a quarter, a half or three quarters, with 0 to 4 short bars on it and one of three gray shades. Each of these four features follows one rule along the row: it stays the same, steps (a quarter turn, or one bar more or fewer), alternates between two values, or, for the shape, cycles through three. The player taps an answer tile to choose it and then taps Done. The paper counts and grades on this page are those of the Matrix Reasoning test, the published test the game is compared with, not the game's own.
More about the paradigm
A puzzle's level is the number of its four features that change along the row, from 1 at level 1 up to 4 at level 4. Levels 1 and 2 show 4 answer tiles that differ in 2 features, and levels 3 and 4 show 8 that differ in 3, where every combination of the right and a wrong value of those features is one tile, so choosing the answer takes the rule of each of those features: one rule alone leaves 2 of the 4 tiles, or 4 of the 8. There are 32 puzzles, 8 at each level, in order, after three practice puzzles that show the right tile and run once more when fewer than 2 are right. A rest follows puzzle 16. There is no time limit, each puzzle is numbered, and no screen gives a score or a time.
The Matrix Reasoning test fills a grid by row and column rules. This game differs: it shows one row of three tiles that change from tile to tile, with the next tile to find. The validation plan also compares the game with the Letter Pair Reasoning Task. The tiles, the rules, the levels, the answer tiles, the forms and the layout are ours. The design follows the account of Carpenter, Just and Shell (1990), who traced differences in figural reasoning on matrix problems to inducing abstract relations and keeping many goals in working memory, which the rules of a puzzle load, and the rule-based making of figural items that Blum and Holling (2018) studied.
For a study, one session gives the score. For a retest, run a second visit of the same study 7 to 28 days later, which runs the next form. Keep each person on one device class at every visit. In a study that reads change across visits, add a run-in session before baseline, so that the first scored session is not the first time a person meets the puzzles. Try it now shows a seventh form that no study uses.
Scores
- Puzzles Solvedcount
- The test puzzles, of 32, whose confirmed tile is the right one. Empty unless all 32 test puzzles have a usable answer.
- Higher means more puzzles solved; chance is 6 of 32; compared within a study
- Level 1 Puzzles Solvedcount
- The level 1 puzzles solved, of 8: one feature changes along the row and the answer tiles vary in it and in one feature that stays the same.
- Higher is better
- Level 2 Puzzles Solvedcount
- The level 2 puzzles solved, of 8: two features change along the row.
- Higher is better
- Level 3 Puzzles Solvedcount
- The level 3 puzzles solved, of 8: three features change along the row, among 8 answer tiles.
- Higher is better
- Level 4 Puzzles Solvedcount
- The level 4 puzzles solved, of 8: all four features change along the row, and three of them vary across the 8 answer tiles.
- Higher is better
- Median Solve Timems
- The median time from a puzzle appearing to the confirmed choice, over the solved puzzles.
- Informational; no score depends on time
- Answer Changescount
- How many times the selected tile was moved to another before it was confirmed, summed over the usable puzzles.
- Informational
- Interrupted Puzzlescount
- Puzzles in which the tab was hidden, the window changed size, or a second finger or a palm touched the screen. They stay in the score.
- Quality check
- Usable Puzzlescount
- Test puzzles with a usable answer: a row of a known form, puzzle, level and tile.
- Quality check
- Share of Key Answers%
- The share of the usable puzzles confirmed with a key (Enter or Space on Done) and not a tap or a click.
- Informational; the comparison of touch and keys reads this share within each look
- Share of Touch Answers%
- The share of the usable puzzles confirmed by touch.
- Informational; the comparison of touch and keys reads this share within each look
Use and evidence
1,106 papers from 2021 to 2025 name this task in their title or abstract, down 1% compared with 2016 to 2020. That ranks it 24 of 711 test families in our census. These counts belong to the Matrix Reasoning 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 | 303 |
| 2024 | 216 |
| 2023 | 200 |
| 2022 | 204 |
| 2021 | 183 |
| 2020 | 251 |
| 2019 | 214 |
| 2018 | 215 |
| 2017 | 229 |
| 2016 | 203 |
| 2015 | 219 |
| 2014 | 202 |
| 2013 | 186 |
| 2012 | 198 |
| 2011 | 181 |
| 2010 | 150 |
| 2009 | 123 |
| 2008 | 121 |
| 2007 | 92 |
| 2006 | 91 |
| 2005 | 80 |
| 2004 | 98 |
| 2003 | 74 |
| 2002 | 87 |
| 2001 | 51 |
| 2000 | 41 |
Use grade A
1,000 or more 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
- 89
- Papers on reliability or validity
- 868
- Registered clinical studies using it as an outcome
- 272
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 Matrix Reasoning 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.
- Tiles
- A tile is one of 6 drawn shapes, turned 0, 90, 180 or 270 degrees, with 0 to 4 short bars on it and one of 3 gray shades. Every shape, bar and gray is drawn in code with no picture file. No shape is a letter, a digit or a shape of another test.
- Puzzles
- A row of 3 tiles and an empty place. Each of the four features follows one rule along the row: it stays the same, steps (a quarter turn, or one bar more or fewer), alternates between two values, or, for the shape, cycles through three. The player picks the tile that comes next.
- Levels
- A puzzle's level is the number of its four features that change along the row, from 1 at level 1 up to 4 at level 4. Levels 1 and 2 offer 4 answer tiles that differ in 2 features, and levels 3 and 4 offer 8 that differ in 3, where every combination of the right and a wrong value of those features is one tile, so choosing the answer takes the rule of each of those features: one rule alone leaves 2 of the 4 tiles, or 4 of the 8. There are 32 puzzles, 8 at each level, in order.
- Answering
- A tap or a click on an answer tile selects it with a ring, a tap on another moves the ring, and Done confirms. The next puzzle follows 500 ms later. Tab moves through the tiles and Done, and Enter or Space acts on the one that has the focus: on a tile it selects, on Done it confirms. A second tap on the selected tile within 300 ms changes nothing and is counted. Every tile is at least 72 px, so 8 answer tiles fit a 360 by 640 screen.
- Practice and rest
- Three practice puzzles come first, each at level 1 (the marks, the shade and the shapes). The right tile glows for 1,500 ms after each, with one line that names the change. The practice runs once more when fewer than 2 of the 3 are right. A rest screen follows puzzle 16. Each puzzle is numbered, as in Puzzle 5 of 32, and no screen shows a score or a time.
- Timing
- No time limit. The game takes about 12 to 16 minutes, and no score depends on how fast a person answers. A puzzle with a hidden tab, a resized window or a second finger stays in the score, and its row records it.
- Forms
- 6 forms for studies, built from one plan, so each puzzle has the same level, the same rules and the same kind of answer tiles in every form. A seventh form is for Try it now, and no study uses it. Each person gets the forms in an order drawn once for them, and each session takes the next form.
- Looks
- The game look adds a pale sky, clouds and hills behind the tiles, and the plain look draws the same tiles on a pale ground. Tiles, grays, sizes and places are the same in both. Each gray is at least 20 apart from the others and from the ink on the CIEDE2000 scale, in normal vision and in simulated protan, deutan and tritan vision.
- Age and use
- For adults, aged 18 and over. In older-adult mode the words and the buttons are 1.2 times larger; the tiles, the timing and the scoring do not change. Keep each person on one device class at every visit. For a retest, run a second visit of the same study 7 to 28 days later, which runs the next form. In a study that reads change across visits, add a run-in session before baseline.
- Scores
- The puzzles solved, of 32 (the main score), and the puzzles solved at each level, of 8. Also the median time to confirm a solved puzzle, how often the selected tile was changed, the puzzles interrupted and the number of usable puzzles. The share of puzzles confirmed with a key and by touch is also kept, so the two can be compared. No percentile is computed.
The design follows the account of Carpenter, Just and Shell (1990), who traced differences in figural reasoning on matrix problems to inducing abstract relations and keeping many goals in working memory, and the rule-based making of figural items that Blum and Holling (2018) studied, where most of the rules varied added to an item's difficulty. The tiles, the rules, the levels, the answer tiles, the forms and the layout are our own. This game is a new measure for adults, built for repeated visits: its reliability and norms come from use, the figures published for the Matrix Reasoning test are not its own, and it computes no percentiles.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: tile_row_game test puzzles from the scored run, up to 32 (four levels of 8, with 4 answer tiles at levels 1 and 2 and 8 at levels 3 and 4: chance 1 in 4 or 1 in 8 a puzzle, 6 of 32 in all). Below chance invalidates and not above chance is flagged. One answer position on 90% of the puzzles invalidates when accuracy is not above chance. Confirms under 200 ms on at least half the puzzles invalidate. A run without all 32 test puzzles is invalid. Fewer than 5 of the 8 level 1 puzzles solved, 8 or more wrong confirms under 2,000 ms, or 8 or more interrupted puzzles flags it for review. A puzzle with a hidden tab, a resized window or a second finger stays in the score. Untimed, so nothing can be missed.
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. The answer tiles of each form are placed so that every position holds the answer equally often, so one position cannot score above chance. 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. A run that stops before puzzle 32 rests on part of the game and is not read as the puzzles solved (ours). Missing most of the 8 easiest puzzles suggests the game was not understood, and 8 or more wrong answers confirmed in under 2 seconds suggest the puzzles were not worked through (ours; a right answer never counts as rushed). A puzzle waits on screen until it is confirmed, so a hidden tab, a resized window or a second finger leaves no puzzle out, and 8 or more of them are flagged only (ours). A run that holds two forms cannot be read against one table and is not scored.
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 Tile Row Game in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Carpenter, Just and Shell, 1990), with the standard settings (look: Game). The main score was puzzles solved (the test puzzles, of 32, whose confirmed tile is the right one).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Carpenter, P. A., Just, M. A., & Shell, P. (1990). What one intelligence test measures: A theoretical account of the processing in the Raven Progressive Matrices Test. Psychological Review, 97(3), 404-431. https://doi.org/10.1037/0033-295X.97.3.404Differences in figural reasoning traced to inducing abstract relations and keeping many goals in working memory, which the puzzles' rules load (abstract read).
- 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.2018.01286Figural items made from stated rules, where most of the rules varied added to an item's difficulty, as the number of features that change sets a puzzle's level here (abstract read).
Source papers
Questions
How long does the Tile Row Game take?
About 12 to 16 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?
Puzzles Solved, Level 1 Puzzles Solved, Level 2 Puzzles Solved, Level 3 Puzzles Solved, Level 4 Puzzles Solved, Median Solve Time, Answer Changes, Interrupted Puzzles, Usable Puzzles, Share of Key Answers and Share of Touch Answers. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Matrix Reasoning test used in research?
These counts belong to the Matrix Reasoning test, the published test this game is compared with. This game is a new measure, so its own reliability and norms come from use. 1,106 papers from 2021 to 2025 name the task in their title or abstract, down 1% compared with 2016 to 2020. That ranks it 24 of 711 test families in the ConductCognition test census (use grade A). Source: ConductCognition test census 0.58.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Matrix Reasoning test have?
These counts belong to the Matrix Reasoning 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, 89 papers report the task's test-retest reliability and 272 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.