The flanker task, and how it differs from the arrow Stroop

September 26, 2026 · 6 min read · Tyamka team

What the Eriksen flanker task is, how the arrow version works, how it differs from the Stroop test, and what our Arrow Rule game borrows from both.

The flanker task asks you to respond to one item in the middle and ignore the items on either side of it, the flankers. When the flankers point to a different answer than the middle item, people answer more slowly and make more mistakes. It was introduced by Barbara and Charles Eriksen in 1974, and it is one of the standard ways to study how attention filters out nearby distractions. Our game Arrow Stroop test, called Arrow Rule in the app, uses arrows too, but works differently, and this article explains both.

The flanker task in one table

A common arrow version uses a row of five arrows. You press left or right for the direction of the middle arrow only.

Most people are slower and less accurate on the incongruent rows. The difference in time between incongruent and congruent rows is called the flanker effect, or flanker interference.

Where it comes from

Barbara A. Eriksen and Charles W. Eriksen published the task in Perception & Psychophysics in 1974 under the title "Effects of noise letters upon the identification of a target letter in a nonsearch task". Their version did not use arrows. The target was a letter in a known place. The letters H and K meant one lever, S and C the other. The target was surrounded by noise letters that could belong to the same response, the other response, or neither.

The key result was that noise letters tied to the other response slowed people down even though they knew exactly where the target would be. Moving the noise letters further from the target made the effect smaller. Attention, it seemed, could not be narrowed to a single spot on command. Things close to what you look at get processed anyway.

The arrow version became popular later. A well-known use is the Attention Network Test by Jin Fan, Michael Posner and colleagues in 2002, which combines a row of arrows with warning and location cues to look at three aspects of attention in one short session. Arrows are easy to explain, need no letter mapping, and work across languages.

How the flanker task differs from the Stroop test

Both tasks measure interference, the extra time a distractor costs. The difference is where the distractor is.

In the Stroop test the conflict sits inside one object. You have to ignore one feature of the thing you are looking at, the word, and respond to another, the ink. In the flanker task the target is clean, and the conflict comes from its neighbours. So the flanker task is more about filtering in space, and the Stroop test more about holding back a strong habit, reading. Both are covered with more examples in the Stroop effect explained and what the Stroop test measures.

And what is an arrow Stroop

The name is used loosely. In research it usually means a task where the word and the arrow disagree, such as the word LEFT inside an arrow that points right, or the conflict between where an arrow points and where it appears. Online, "arrow Stroop test" often just means any task where arrows try to trick you. That covers flanker rows, arrows that appear on the wrong side of the screen, and tasks where you answer against the arrow.

What our Arrow Rule game does

Our game has no flankers. There is one arrow at a time, and a rule above it tells you what to do. The game is built on the idea that the pull toward the wrong answer comes from habit, the reflex to follow where an arrow points. We have not measured this on players yet. That idea is why we file it under the Stroop family and not under flanker tasks.

• An arrow appears with a rule. Same means tap the button that points where the arrow points. Opposite means tap the button that points the other way.
• There are four directions and four buttons. On a keyboard you use the arrow keys or W, A, S and D.
• Each arrow has a time limit. A wrong answer or a miss costs one of three lives. Faster right answers earn more points.

The 15 levels come in four phases. The numbers are game settings, not measurements of players.

The Opposite phase is close to what researchers call an anti-response. You see a direction and have to answer with its opposite, which means suppressing the obvious move first. In the mixed phase either rule can come next, and the same rule can repeat, so you have to read it every time rather than settle into one mode.

On the last three levels the arrow sits off center, near the top, bottom, left or right edge of the panel. Where it sits does not matter, only where it points. But position pulls too. This is the Simon effect, named after J. Richard Simon. In 1967 he and Alan Rudell showed that people answer faster when the correct button is on the same side as the signal, even when the side is irrelevant to the task. Their signals were sounds in the left or right ear, and later work found the same with visual positions. The arrow's position is chosen at random, so on average in three of every four off-center items the side it sits on is not the side you should answer.

Try the difference yourself

If you want to feel the flanker effect, read the table at the top out loud, saying only the direction of the middle arrow, and notice the pause on the incongruent rows. Then play a few levels of the Arrow Stroop test and compare. In the flanker rows the problem is what surrounds the arrow. In Arrow Rule it is the rule and your own habit.

None of these tasks gives a score that means something outside the task. Your points show how you played a level on that day, and they vary with rest, device and practice. Compare them with your own best. For the wider picture of choosing what to attend to, read selective attention with examples, or find more games on the all games page.