Executive Function
Shortest Route Task
Also called: traveling salesperson problem, traveling salesman task, TSP task, visual traveling salesman problem, Euclidean traveling salesman task, route planning puzzle
Measure visual route planning: participants make the shortest route through the points on a map, and each route is compared with the shortest one, worked out exactly.
7-10 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
- Evidence grade
- D
- Registered clinical studies
- 0
What it measures
Measures visual route planning. Each map shows 9, 12 or 15 points, and participants tap them in the order they would visit them to make the shortest route through all of them, back to the start. The main score is how much longer the routes are than the shortest possible ones, worked out exactly.
Constructs
- Visual problem solving
- Spatial planning
- Executive function
- Perceptual organization
Research uses
- Planning and problem-solving research
- Individual differences in visual problem solving
- Spatial cognition and navigation research
- Within-study group comparison
How the paradigm works
The Shortest Route Task measures visual route planning with the traveling salesperson problem: given points on a map, make the shortest route that visits each point once and returns to the start. People find routes close to the shortest one quickly and without training, an order of magnitude better than simple rules of thumb (MacGregor and Ormerod 1996), and the problem has been used to study visual problem solving, planning and individual differences. Each map shows points our code places at random on a square board, at least 14% of its side apart; the start is a square. The participant taps or clicks the points in the order they would visit them, each tap drawing a line from the point before, and after the last point the route closes back to the start by itself. Undo takes back the last point and Start over clears the route, as in the computer version of Graham, Joshi and Pizlo (2000); Done ends the map. There is no time limit. The standard session has 12 maps, 4 each of 9, 12 and 15 points, with each size once in every three maps, after a practice map of 6 points that is followed by the shortest route drawn beside the participant's. The test maps have no feedback.
More about the paradigm
The main score is how much longer the routes are than the shortest routes through the same points, in percent, averaged over the maps: MacGregor and Ormerod's percentage above optimal. The shortest route of each map is worked out exactly, by the dynamic program of Held and Karp (1962), when the map is made and again by the scorer from the stored points. Also reported are the score for each map size, the number of maps solved exactly, routes with crossing legs (which a shortest route never has), the share of routes that visit the points on the outer edge of the map in their order around it, the median time to the first tap (the planning time of Mueller and colleagues 2015) and to the end of a route, and the use of Undo and Start over. Scores support research and do not diagnose.
Scores
- Percent Above Shortest%
- How much longer each route was than the shortest route through the same points, in percent, averaged over the maps.
- Lower is closer to the shortest routes
- Percent Above Shortest, 9 Points%
- The same over the maps of 9 points.
- Lower is closer to the shortest routes
- Percent Above Shortest, 12 Points%
- The same over the maps of 12 points.
- Lower is closer to the shortest routes
- Percent Above Shortest, 15 Points%
- The same over the maps of 15 points.
- Lower is closer to the shortest routes
- Shortest Routes Foundcount
- Maps whose route was as short as the shortest route.
- Higher means more maps solved exactly
- Share of Shortest Routes%
- Shortest routes found as a share of the maps scored.
- Higher means more maps solved exactly
- Routes with Crossingscount
- Routes with at least one pair of legs that cross; a shortest route never has one.
- Lower is better
- Edge Order Kept%
- Share of routes that visit the points on the outer edge of the map in their order around it, as every shortest route does.
- Informational
- Median Time to First Tapms
- Median time from a map appearing to its first tap.
- Informational
- Median Route Timems
- Median time from a map appearing to Done.
- Informational
- Undoscount
- Points taken back with Undo over the test maps.
- Informational
- Restartscount
- Routes cleared with Start over over the test maps.
- Informational
- Maps Scoredcount
- Test maps with a complete route; 12 in the standard session.
- Informational
Use and evidence
1 papers from 2021 to 2025 name this task in their title or abstract. That ranks it 492 of 495 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 0 |
| 2024 | 0 |
| 2023 | 1 |
| 2022 | 0 |
| 2021 | 0 |
| 2020 | 0 |
| 2019 | 0 |
| 2018 | 0 |
| 2017 | 0 |
| 2016 | 0 |
| 2015 | 0 |
| 2014 | 1 |
| 2013 | 0 |
| 2012 | 0 |
| 2011 | 0 |
| 2010 | 1 |
| 2009 | 0 |
| 2008 | 1 |
| 2007 | 0 |
| 2006 | 0 |
| 2005 | 0 |
| 2004 | 0 |
| 2003 | 0 |
| 2002 | 0 |
| 2001 | 0 |
| 2000 | 0 |
Use grade E
Fewer than 30 papers from 2021 to 2025.
Evidence grade D
Little published measurement evidence found.
- Papers on test-retest reliability
- 0
- Papers on reliability or validity
- 0
- 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 0.22.0, counted 3 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Maps
- Points drawn at random on a square board from the session's seed, at least 7% of the side from its edges and 14% from each other, so each keeps a 44 px touch target on a phone. One point, shown as a square, is the start.
- Session
- 12 maps (4 of each size as standard; 2 to 6 as a setting) of 9, 12 and 15 points, each size once in every three maps in an order drawn from the seed. One practice map of 6 points comes first.
- Route
- Points are tapped, clicked or chosen with the keyboard in the order of the route, and the program draws a line from the point before. After the last point the route closes back to the start by itself. Undo takes back the last point, Start over clears the route, and Done ends the map. No time limit.
- Shortest route
- Worked out exactly for each map by the dynamic program of Held and Karp (1962) from the stored points, when the map is made and again by the scorer.
- Feedback
- After the practice map, the participant's route and the shortest route are drawn together, with how much longer the route was. The test maps have no feedback.
- Scoring
- How much longer each route was than the shortest, in percent, averaged over the maps, with the same for each map size, shortest routes found, routes with crossing legs and routes that kept the order of the map's outer points.
- Data
- Each map's points, shortest route, the participant's route and the time of each tap. No pointer path is kept.
The points, maps and scoring are this task's own; no published problem set is used. Careful routes come within a few percent of the shortest, so differences between people are small and best read across groups. 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: shortest_route_puzzle from the scored run, one per map, each a complete route worked out again from the stored points. A route has no chance level, so accuracy is not judged against chance, and a map ends only with a complete route, so there are no missed responses. Routes finished in under 200 ms a point on at least half the maps invalidate. Fewer than half the planned maps scored invalidates, and any other incomplete run is flagged. Routes 50% or more longer than the shortest on average are flagged for review.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. 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. Here the time is the route's time divided by its points, since each point is one choice. Careful routes come within a few percent of the shortest (MacGregor and Ormerod 1996; Graham, Joshi and Pizlo 2000), so routes as long as random ones suggest the points were tapped without planning.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Maps of each size | 4 | 2 to 6 |
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.
- 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 Shortest Route Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; MacGregor and Ormerod, 1996; Graham, Joshi and Pizlo, 2000), with the standard settings (maps of each size: 4). The main score was percent above shortest (how much longer each route was than the shortest route through the same points, in percent, averaged over the maps).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- MacGregor, J. N., & Ormerod, T. (1996). Human performance on the traveling salesman problem. Perception & Psychophysics, 58(4), 527-539. https://doi.org/10.3758/BF03213088The method and its main score: routes drawn by eye through 10 points were on average 0.8% to 3.8% longer than the shortest, an order of magnitude better than simple rules of thumb, and nearly always kept the order of the outer points.
- Graham, S. M., Joshi, A., & Pizlo, Z. (2000). The traveling salesman problem: A hierarchical model. Memory & Cognition, 28(7), 1191-1204. https://doi.org/10.3758/BF03211820A computer version: points clicked in order with a line drawn to each, the last step undone or the route cleared, and routes of 6 to 30 points compared with the shortest found by an exact method.
- Mueller, S. T., Perelman, B. S., Tan, Y. Y., & Thanasuan, K. (2015). Development of the PEBL Traveling Salesman Problem Computerized Testbed. The Journal of Problem Solving, 8(1), 64-90. https://doi.org/10.7771/1932-6246.1169A free computer version with a given start, the time to the first click as planning time, and a composite score with a test-retest reliability of .75 for a set of about five minutes.
- MacGregor, J. N., & Chu, Y. (2011). Human performance on the traveling salesman and related problems: A review. The Journal of Problem Solving, 3(2), 1-29. https://doi.org/10.7771/1932-6246.1090A review of human performance on the problem, its variants and its uses in neuropsychology.
- Held, M., & Karp, R. M. (1962). A dynamic programming approach to sequencing problems. Journal of the Society for Industrial and Applied Mathematics, 10(1), 196-210. https://doi.org/10.1137/0110015The exact method this task uses to work out each shortest route.
Source papers
- MacGregor JN, Ormerod T (1996). Human performance on the traveling salesman problem. Perception & Psychophysics, 58(4):527-539.Paradigm source, Adults
- Mueller ST, Perelman BS, Tan YY, Thanasuan K (2015). Development of the PEBL Traveling Salesman Problem Computerized Testbed. The Journal of Problem Solving, 8(1):64-90.Paradigm source, Adults
- No normative data for this task.Paradigm source, N/A
Questions
How long does the Shortest Route 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?
Percent Above Shortest, Percent Above Shortest, 9 Points, Percent Above Shortest, 12 Points, Percent Above Shortest, 15 Points, Shortest Routes Found, Share of Shortest Routes, Routes with Crossings, Edge Order Kept, Median Time to First Tap, Median Route Time, Undos, Restarts and Maps Scored. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Shortest Route Task used in research?
1 papers from 2021 to 2025 name it in their title or abstract. That ranks it 492 of 495 test families in the ConductCognition test census (use grade E). Source: ConductCognition test census 0.22.0, counted 3 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Shortest Route Task have?
In the published research, 0 papers report its test-retest reliability and 0 registered clinical studies use it as an outcome (evidence grade D in the ConductCognition test census).
Does it report a percentile?
No. Scores and quality flags only.
Which settings can a researcher change?
Maps of each size (2 to 6). 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.
Related tests