Which characters actually cost you
Knowing a chart and recognising a character quickly are different skills. The shapes you meet often tend to become familiar first, while a few others stay slow. That is useful information: you can spend your next few minutes on those shapes instead of restarting every row from the beginning.
This quiz shows one character and asks you to choose its reading. It draws its wrong answers from Niwa's reviewed look-alike data and additional shape comparisons. Seeing ツ alongside the reading for シ is a more useful discrimination exercise than choosing between unrelated sounds. Options with the same romaji are removed, because two identical answers would make the question unfair. The writing differences between じ and ぢ cannot be diagnosed just by asking for their shared pronunciation.
Understanding your result
The score is the number of correct answers out of the questions you answered. Seconds per character is your average time from seeing a character to choosing an answer. Accuracy and speed belong together: choosing rapidly with many mistakes means something different from answering carefully and correctly.
The result divides the selected pool into three groups. “Correct in this test” means you answered a character correctly without a recorded miss. It does not certify lasting mastery. “Confused” shows a character you missed and the character whose reading you chose instead. If you answered the same character both correctly and incorrectly, it stays in the confused group so the useful warning is not hidden.
“Not tested” means the clock ran out before that character appeared. It does not mean you do not know it. Thirty seconds is far too short to assess every character, especially with both scripts and combined sounds selected. Repeat the test on another day to sample more of the pool. If you switch away from this tab, the quiz ends with a clearly labelled partial result.
What to practise next
If you missed a few characters, use the ten-item follow-up drill. It repeats only actual misses from this test, so a single weak character may appear several times. The drill has no timer, and its score does not change your original result. When you make a mistake, read the correction aloud and look for the stroke or orientation that distinguishes the two shapes.
If you missed nothing, the page does not invent weak characters or create an unnecessary drill. You may still have plenty of untested characters. Try a different script or a fresh test rather than treating one short perfect score as proof that you have finished learning kana.
Hiragana, katakana, or both
Test scripts separately when you want a clear diagnosis. Hiragana is used for grammatical endings and many common words. Katakana appears often in loanwords, names and emphasis; several of its shapes differ mainly in the direction of a stroke. Combined sounds use small kana, and the two voicing marks or small circle create additional distinctions to notice.
Keep romaji as a temporary aid when studying, then try short stretches of reading without it. A quiz can show which shapes need attention, but recognising them inside real words is the next step. Listen to examples, read a little, and return to the pairs that still make you hesitate.
Character, audio and stroke-data sources are listed on the tool attribution page.