Spelling test that reads the word aloud

English spelling is not random, and the places it goes wrong are not random either. Almost every misspelling an adult produces falls into one of five situations where the sound of the word does not determine its letters. A consonant may or may not double before a suffix, and in American spelling that depends on where the stress falls: refer is stressed on the last syllable so it becomes referring, target is stressed on the first so it becomes targeted, and the second one is what people write wrongly once they have learned the rule. Two spellings may share a pronunciation exactly, in which case nothing audible can decide between them and only the sentence can. An ending may be -ible or -able, -ence or -ance, -ary or -ery, all of them reduced to the same sound in speech. An unstressed vowel collapses into a schwa, so separate hides an a that no pronunciation reveals. And a letter may simply not be pronounced, as in subtle or receipt, because English kept the spelling when it dropped the sound.

That taxonomy is why this page dictates rather than offering four spellings to choose from. A multiple-choice item shows you the correct form and asks you to recognize it, which is a genuinely easier task and a different one: the doubling rules and the -ence endings only have to be known when nothing on screen contains them. Dictation also makes the carrier sentence load-bearing rather than decorative. Read complement on its own and there is no information in the sound at all; read it in a sentence about two studies completing each other and the spelling follows. The one thing dictation cannot fix is a word whose pronunciation actively misleads — most speakers say February without the first r, so that word has to be spelled from knowledge and the voice is a hindrance. It is in the bank precisely for that reason.

What this page will not do is borrow somebody else’s norms, and the reason is worth stating plainly. No population figure exists for spelling a word list, because a word list is chosen rather than sampled: the average would describe whoever picked these thirty words and the voice that read them, not the people who took it. So there is no spelling age here, no grade equivalent and no rank — a count, a median time and five class rates, and the only comparison they support is with your own next run. A published spelling test carries norms because its word list was administered to a standardization sample under controlled conditions, and the norms belong to that list. Twenty words chosen by us, read by whichever voice your operating system ships, cannot borrow them. So the output is a count, a median time and five rates, and the comparison it supports is with your own next run. If what you actually wanted was the size of your vocabulary rather than your control of its spelling, the vocabulary test estimates that with a guessing correction, and the synonym test finds the frequency band where your recognition starts to fail. If you are being placed rather than diagnosed, the English level test walks a CEFR ladder and reports a band, and the sentence completion test does the same trick this page does — errors sorted by the rule behind them — for grammar rather than for spelling. If reading rather than writing is where the effort goes, the reading speed test measures words per minute with a comprehension check attached.

A reaction time here is the interval between the frame that painted the stimulus and the timestamp the browser attached to your key, both read from the same monotonic clock. What neither can see is the display pipeline behind it, so on a 60 Hz screen roughly 16 ms of every figure below is the machine rather than you. That is the timing floor: two numbers closer together than that are the same number, and this page reports no precision it cannot support.

Nothing here is scored on speed, and the median below is descriptive rather than graded: a word typed in four seconds and one typed in twenty both count once. The interval it measures starts at the frame that painted the prompt, so all three spoken readings sit inside it — which is why it is reported in seconds and never finer, and why comparing it with a typing speed measured on visible text would compare two different things. Words are not discarded for arriving fast; a run you may sit on has no such thing as too fast.

This is a measurement exercise, not a clinical assessment. It reports what you did on this page against a stated reference and nothing more — it cannot establish dyslexia, dysgraphia or a specific learning difficulty. Only a qualified professional, working with more than a browser, can make that judgment.

Spelling difficulty has many causes that a word list cannot separate — a hearing difference, an unfamiliar keyboard layout, a language learned later, or simply not having read the word before. A run of twenty words distinguishes none of them.