Stroop Test: Tap the Ink, Not the Word

How to play the Stroop test

This Stroop test takes 30 seconds. A color word appears on a card in some ink color. Answer the ink, not the word. Your reading habit will fight you.

1. Pick a mode. Test mode has fixed difficulty and counts toward your records. Training mode adapts your level after each run.
2. Look at the ink. Coach Mochi’s speech bubble reminds you: Tap the ink color! In Test mode, half the words match their ink (GREEN printed in green) and half don’t (GREEN printed in red). For that second one, the right answer is red.
3. Tap one of four buttons. They sit in a 2×2 grid, each with a color swatch and the color’s name in plain dark text, so nothing depends on color alone. In Test mode they never move.
4. Shake off mistakes. A wrong answer shakes the card, puts an ✕ on your button and a check on the right one, and locks the buttons for half a second (500 ms). Keep moving.
5. Read your result. Your score is correct answers in 30 seconds. A run with fewer than 5 answers, or fewer than 40% right, doesn’t count.

Training mode has eight levels. Mismatching words grow from 50% to 85% of trials. From level 4, the buttons swap places on every word, and the bubble says Ink color! Buttons shuffle. From level 6, a thick frame with a Word tag sometimes appears around the word (about one trial in four), and the bubble says Frame? Tap the word!: then answer the word, not the ink. You level up with 90% accuracy and at least 17 correct (13 with Large text & slower pace), and drop a level below 70% accuracy.

What it measures

The Stroop test works two skills: inhibitory control, stopping an automatic response, and selective attention, staying on one feature (the ink) while another (the word) shouts at you.

For a fluent reader, reading is automatic. When the word says RED and the ink is blue, two answers show up at once and you have to hold one back. That slowdown is the Stroop effect, named after the psychologist J. Ridley Stroop, who described it in 1935. The task has been a standard tool in attention research ever since.

It still runs in online studies. In the public data of one (Barzykowski et al., 2022; 466 adults answering on a keyboard), people were on average 121 ms slower on mismatching words.

Some honest limits:

• It is a game, not a diagnosis. A 30-second phone session can’t tell you anything medical about attention, ADHD or anything else.
• Interference is noisy for one person. In that dataset the standard deviation (116 ms) was almost as large as the average, and one person in ten showed 14 ms or less. That’s why Brainsquat scores correct answers instead of timing the gap.
• Practice improves the practiced game. A large review found that people reliably improve on the tasks they train, with little evidence of gains carrying over to everyday performance (Simons et al., 2016). A higher score means you got better at this game. That’s the promise.

Averages and ranges

Reference values from the open data of Barzykowski et al. (2022): an online study of 466 adults answering on a keyboard (the 199 who gave their age averaged 24.7 years).

How to compare yourself fairly:

• Phone taps differ from key presses. Touchscreen browsers add about 58–70 ms to each response (Pronk et al., 2020), and tapping on-screen buttons is a different movement from pressing keys. Those reference times are background, not a target.
• Age. A meta-analysis found that the larger Stroop effect in older adults is largely explained by general slowing (Verhaeghen & De Meersman, 1998). We don’t show age-group averages.
• No peer band for your score. Peer comparison may come later. For now, beat your own record.

How your score works: