An arrow vs letter flanker task is a side-by-side format of the classic Eriksen flanker paradigm in which one variant uses directional arrows as stimuli and another uses letters such as H and K; only the central item in a row determines the correct response, and the surrounding items are distractions whose direction either matches or conflicts with the centre. The browser-based Flanker Test keeps the arrow side of that comparison and drops the letter side entirely, so every one of its twelve rows is a five-arrow display with the middle arrow outlined. The page labels its rows matching when the four flanking arrows point the same way as the centre arrow and conflicting when they point the opposite way, and the route is fixed at six matching and six conflicting rows so the final panel can compare two like-sized groups from your own session. None of this is a hidden calculation about you; the summary is a transparent record of what happened on the device in front of you, kept inside the browser.
What an Arrow vs Letter Flanker Task Actually Tests
The Eriksen flanker task was introduced in 1974 as a way to measure how well someone can ignore nearby information that pulls attention away from a chosen target. In its most-cited form, a row of five items is shown and the participant responds to the centre item alone while the two items on either side act as either supportive or distracting context. When the original task is split into a letter version (typically strings built from H and K, with a centre letter such as S or K that points left or right) and an arrow version (strings such as >>>>> versus <>><<), the comparison between the two is sometimes used to argue that arrows produce stronger interference effects because their directional meaning is more automatic than letter reading.
That background is the entire reason a search for "arrow vs letter flanker task" exists at all: people are curious which stimulus type produces a bigger interference cost, and many tutorials present both versions side by side. The browser tool linked in the next section is the arrow-only form; there is no letter condition to compare against, and the page makes that clear by removing one variable at a time. If the goal is to see how the arrow half of that comparison feels on its own, that is exactly what the fixed twelve-row route is built to show.
The Centre-Arrow Version on This Page
The arrow side of an arrow vs letter flanker task lives on a page whose product contract is short and transparent. A fixed twelve-row route is set up with six matching rows and six conflicting rows, every row shows five arrows, and only the outlined middle arrow counts as the answer. The four surrounding arrows are pure visual context and are never a second rule you have to infer. The table below shows how the two row types are defined on the page.
| Aspect | Matching row | Conflicting row |
|---|---|---|
| Centre arrow direction | Same as the four surrounding arrows | Opposite to the four surrounding arrows |
| Number of rows in the fixed route | 6 | 6 |
| Role of the four surrounding arrows | Visual context only; not a second rule | Visual context only; not a second rule |
| What the page records from your response | One accuracy value and one row response time | One accuracy value and one row response time |
| Used in the interference index | Yes — contributes the matching average | Yes — contributes the conflicting average |
The number of each row type is fixed so the two response-time averages come from equal-sized samples within a single round. The route order is also fixed, so two runs on the same device can be compared on equal terms; only the input, not the stimulus sequence, changes between plays.
How to Play the Fixed Twelve-Row Route
- Open the Flanker Test page in a browser tab and keep the tab visible from the moment you press Start. Hidden-tab timing is paused by design, so starting in another window would just delay your first row.
- Wait for the first five-arrow row to render with the centre arrow clearly outlined. The route is fixed and identical on every play, so what changes between sessions is your input, not the row order.
- Choose the direction of the outlined centre arrow. Use the two large on-screen buttons labelled Left and Right, or use the physical Left and Right Arrow keys as an optional keyboard shortcut.
- Let the page accept the response and advance to the next row. The game takes exactly one response for the visible row; a repeated key event or a second click on the same row cannot add another data point.
- If you switch apps, answer a call, or get a notification that hides the tab, the current row will pause instead of counting background time. When you come back, press Resume this row to start a new visible response interval for that same row.
- After the twelfth response, the result panel replaces the play area. Read your accuracy, the matching-row average response time, the conflicting-row average response time, and the interference index, then use Restart to play the same fixed route again.
The whole route is twelve responses, so a single uninterrupted play fits in well under a minute. Restart does not change the row order; it returns to the beginning of the same fixed twelve rows.
Reading the Matching, Conflicting, and Interference Results
The result panel reports four figures from the route you just finished, all computed locally inside the browser. Accuracy is the percentage of correct centre-arrow choices out of the twelve responses. A faster route with more wrong answers is not automatically ranked higher than a slower route with more right answers; the page keeps accuracy visible because faster is not the same as better.
The two average response times come from the same one round of play. The matching-row average is the mean response time across the six matching rows, and the conflicting-row average is the mean across the six conflicting rows. Both numbers describe what happened on this device for this session; they are not compared with a population, an age range, a classroom, or a published reference.
The interference index is the simplest piece of arithmetic on the page. The page defines it as the conflicting-row average response time minus the matching-row average response time, drawn from this one fixed route. As a worked example, if your conflicting-row average is 480 ms and your matching-row average is 410 ms, the index is 480 − 410 = 70 ms, with the difference purely directional: a higher conflicting average produces a positive number, an equal one produces zero, and a lower conflicting average produces a negative number. The sign is not a verdict about your attention span; it is a transparent subtraction of two numbers the page just showed you. For a related look at how a similar browser game reports accuracy and speed side by side, see Digit Symbol Accuracy vs Speed: How This Game Reports Both.
A positive or negative index is not ranked against another player, age range, classroom, workplace, clinical sample, or published reference. The numbers describe what happened in this session on this device, and that is the only context the summary is meant to carry.
What the Numbers Do Not Measure
The page is an entertainment exercise, not a clinical, IQ, or cognitive assessment. It does not diagnose attention, measure intelligence, or show whether your brain has improved over time. It has no clinical interpretation, no health threshold, no norm group, no educational score, and no professional use. Screen size, input method, browser scheduling, interruptions, familiarity with the route, and ordinary day-to-day variation can all affect a single session, so two runs on the same device can land at noticeably different averages even when nothing meaningful has changed.
The game runs entirely in your browser and does not ask for an account, profile, name, camera, microphone, or upload. The shared game shell may keep a local best score in browser storage, but it is not an online leaderboard and is not sent to Lizely. If you need support for health, learning, work, or accessibility, the page is not a substitute for a qualified assessment; treat the figures as transparent feedback from one round of play, not as a claim of personal ability.
Related reading: Commission vs Omission Errors: A Twelve-Cue Game.