Reading is not something a literate adult chooses to do. Put a word in front of someone and they have already read it, before any decision is involved. The Stroop task works by making that automatic response the wrong answer.
How to play the Stroop Test
Each trial shows one colour word printed in one colour. Your job is to identify the ink — the colour the letters are drawn in — and choose it from the buttons below. The word itself is irrelevant, and on most trials it names a different colour entirely.
When the word GREEN is printed in red, the answer is red. The pull towards “green” is the effect the task is built around, and it does not go away with practice within a round. What improves is your ability to hold the ink rule in place while it happens.
Trials advance as soon as you answer, so a round takes as long as you take. Each trial also has its own time limit; letting it expire counts as a wrong answer, because refusing to commit is not a way to avoid the conflict.
On Standard, 13 of the 24 trials are conflict trials; on Hard it is 21 of 30. You cannot wait for the hard ones to pass. Matching trials are the exception, not the baseline.
What actually helps
- Look at the letters without reading along. Softening focus on the word shape rather than tracking it left to right reduces how strongly the word competes.
- Name the ink before you look at the buttons. Deciding first and then finding the button is faster than scanning the buttons while still resolving the conflict.
- Learn the button positions. The order is fixed for the whole round. Once your hand knows where each colour sits, the remaining time is spent on the actual task.
- Accept the delay on conflict trials. Being slower on conflict trials is the normal, expected pattern. Fighting it by rushing produces errors, which cost far more than the extra fraction of a second.
- Recover immediately after a mistake. A wrong answer is one trial. Carrying it into the next two turns one error into three.
What the difficulty levels change
Three things move together, and all three are real changes in demand rather than presentation. Easy uses three colours, so there are fewer competing responses to hold. Standard and Hard add a fourth. The proportion of conflict trials rises from 40% to 54% to 70%, so there is progressively less respite. And the per-trial time limit tightens from 2.6 to 1.8 seconds.
Hard is not simply Standard sped up: with 70% conflict, the matching trials stop functioning as a rest and the ink rule has to be actively maintained for essentially the whole round.
| Level | Trials | Colours | Conflict trials | Time each | Longest round |
|---|---|---|---|---|---|
| Easy | 20 | 3 | 8 (40%) | 2.6 s | 52 s |
| Standard | 24 | 4 | 13 (54%) | 2.2 s | 53 s |
| Hard | 30 | 4 | 21 (70%) | 1.8 s | 54 s |
How the Stroop score works
Correct answers supply 82% of the score and answer pace the remaining 18%. The pace contribution is gated by accuracy and collapses at the random-choice baseline — one in three on Easy, one in four on Standard and Hard — so tapping quickly through the round earns nothing. Expired trials count as errors.
Your result also reports your interference effect: how much slower your correct conflict answers were than your correct matching ones. This is the quantity the task is famous for, and it is deliberately not part of your score. Over twenty-odd trials it is far too noisy to rank anyone, and scoring it would reward answering slowly on every trial to flatten the difference. It is shown because it is interesting, not because it is a measurement of you.
Records are kept separately for each level, since a 70%-conflict round is not comparable to a 40% one.
How each round is built and checked
Every round comes from its seed, so the same seed always produces the same trials. The generator first builds a balanced list of ink colours, so each colour is the correct answer a near-equal number of times and no button ever becomes the safe guess. It then selects which trials carry conflict, and gives each conflict trial a word drawn from the other colours in play.
The round is then re-read independently: congruency is recalculated from the word and ink alone rather than trusted from the generator, the conflict count is checked against the level, and the balance of correct answers is verified to be within one trial of even. Two structural limits apply. The same ink may not appear more than twice in a row, which would turn the round into a motor repetition rather than a naming task. And no more than four matching trials may run consecutively, because a long stretch without conflict lets the reading habit reassert itself — which is exactly what the task exists to disrupt.
Long runs of conflict trials are deliberately permitted. They are not a fairness problem; they are the difficult part of the test.
Colour, and who this game excludes
We should be direct about this. The classic Stroop effect depends on discriminating colours, and that cannot be designed away without changing the task into something else. If you cannot reliably tell these four colours apart, this game will not work for you, and no amount of interface work fixes that.
What we have done is make sure colour is never the only channel anywhere it does not have to be. Every answer button carries its colour name in text and a fixed shape — triangle, circle, square, diamond — alongside its swatch, so choosing an answer never depends on colour alone. Correct and incorrect feedback is a tick or cross plus words, never a colour change. The four colours were chosen for wide separation in both hue and lightness, and yellow was left out of the classic set because at a legible contrast against this site’s paper background it reads as gold and becomes a naming problem rather than an interference one.
That colour-free version now exists: the Shape Stroop Test puts a shape word inside an outline shape and asks you to name the outline. It is the same task with the colour dimension swapped out, and it is the version used in the Daily Brain Workout. Our accessibility page records this limitation honestly rather than claiming equivalence.
Common Stroop mistakes
Trying to not read
You cannot decide not to read a word you can see. The workable goal is not suppressing the reading but declining to act on it, which is a different and more achievable thing.
Rushing the conflict trials
Being slower on conflict trials is the expected pattern, not a failure. Accuracy is worth 82% of the score and speed 18%, so trading a correct answer for a faster one is always a bad exchange.
Hunting for the button after deciding
If you resolve the ink and then search the row, the search costs as much as the conflict did. The button order never changes within a round, so it can be learned in the first few trials.
Reading the answer buttons as words
The buttons are labelled with colour names, which is itself readable text. Using the shapes and positions rather than re-reading the labels each trial removes a second, unnecessary reading step.
Stroop Test questions
What is the Stroop effect?
The reliable finding that people are slower and less accurate at naming the colour a word is printed in when the word names a different colour. It was described by John Ridley Stroop in 1935 and is one of the most reproducible results in experimental psychology. The usual explanation is that reading is more automatic than colour naming for literate adults, so the two responses compete and the stronger one has to be overridden.
Does a low interference score mean I have better focus?
We would not read it that way, and neither should you. Over twenty to thirty trials the figure is dominated by noise, and it moves with how fast you were answering overall. It is shown as feedback because the effect is interesting to see in your own data, not as a measure of attention, focus, or anything else about you. It has no effect on your score.
Is this a clinical or diagnostic test?
No. Research and clinical versions of the Stroop task are administered under controlled conditions with normative data behind them, and none of that applies here. This is a puzzle. It does not assess attention, executive function, cognitive health, or any medical condition.
Can I play if I am colour-blind?
Probably not, and we would rather say so than pretend otherwise. The stimulus is a colour and that is the whole task. The answer buttons carry names and shapes so the response side never depends on colour, but if the four inks are not distinguishable to you the game does not work. Play the Shape Stroop Test instead: it is the same interference task built entirely without colour, and it is the one that appears in the Daily Workout.
Can I use a keyboard?
Yes. Press the number key matching an option — 1 to 3 on Easy, 1 to 4 on Standard and Hard — from anywhere in the game. You can also move between the buttons with the arrow keys and press Enter or Space.
Is the Stroop Test in the Daily Brain Workout?
No, and that is deliberate. It was in the pool briefly in workout version 5, but a daily routine that gives everyone the same five games cannot contain a station some players are unable to complete. Workout version 6 replaced it with the Shape Stroop Test, which creates the same conflict using outline shapes and no colour at all. This colour version remains a complete standalone game with its own records.