Find One Wrong Word in Each of 12 Original Sentences
Grammar Error Quiz asks you to find exactly one wrong word in each of 12 fixed sentences and then choose its single source-backed replacement from four visible options, so the route stays a reproducible, single-token repair task rather than a free-form proofreader. Each prompt first appears as a row of labelled word buttons. You pick the indexed wrong token, the screen swaps to four correction buttons that each cite the governing rule family beside the choice, and the corrected sentence differs from the prompt at exactly one whitespace-delimited token. That single-token contract is what makes the route auditable, keyboard-friendly, and free of the dialect or style disputes that usually derail grammar conversations.
Because the bank was written for the product rather than copied from a grammar lesson, every prompt has a narrow Standard English boundary and a published rule behind it. The same sentence can be replayed, restarted, and tested again without surprise, and the Grammar Error Quiz page always returns to that fixed 12-sentence route so behaviour stays reproducible across browsers and sessions.
How to Play Grammar Error Quiz in Three Steps
The whole route reduces to three concrete moves, all of which work with mouse, touch, or keyboard.
- Start the fixed 12-sentence route and choose the one wrong word with its button or its visible number shortcut. Number keys 1 through 9 select the visibly numbered word when that number exists.
- After finding the indexed token, choose the one-word correction that satisfies the displayed source-backed rule. The relevant rule scope and source family appear beside the four correction buttons so you can compare the candidates against the rule, not just against each other.
- Finish all sentences and treat the completed repairs, errors, and points as results from this repeated entertainment route. Restart returns to the same audited route so any later score can be compared on equal terms.
A correct repair advances to the next sentence. A miss stays on the same step, adds one error, and costs up to 25 points. Held keys, modifier shortcuts, composition events, invalid indexes, old phase tokens, duplicate clicks, input before Start, and input after completion are all treated as no-ops rather than hidden penalties, which keeps the score transparent.
The Grammar Rules Covered in the 12 Sentences
The 12 sentences each test one rule family drawn from reference sources. The table below lists the families that appear in the frozen bank alongside the published source that establishes each rule; the example sentences on those source pages are not reproduced inside the quiz, which keeps the prompts self-contained and original.
| Rule Family Tested in the Quiz | Reference Source |
|---|---|
| Subject–verb agreement | Cambridge Grammar — verb agreement |
| Does or did followed by a base verb | Cambridge Grammar — Do |
| Modal verbs followed by a base form | Cambridge Grammar — modality forms |
| A or an chosen by the following sound | Cambridge Grammar — a/an and the |
| Uncountable use of "information" | Cambridge Grammar — information |
| Object pronoun case (me, him, her, us, them) | Cambridge Grammar — personal pronouns |
| Comparative "than" with the right adjective form | Cambridge Grammar — comparison adjectives |
| Plural count noun after a numeral | Cambridge Grammar — countable and uncountable nouns |
| Irregular past "bought" from "buy" | Cambridge Dictionary — buy |
| Singular agreement with "everything" | British Council — indefinite pronouns |
Each rule family maps to one of the four correction choices in the second phase of the step, and the rule scope appears on screen so the reader is never asked to trust a silent scoring engine. The British Council reference pages back the present-simple, modal-verb, count-noun, personal-pronoun, comparative-adjective, uncount-noun, and indefinite-pronoun sides of the bank. Cambridge Grammar backs the verb-agreement, base-form, sound-based article, uncountable-noun, pronoun-case, and plural-after-numeral sides, and the Cambridge Dictionary entry for buy backs the irregular-past prompt.
What the Quiz Excludes from Its Error Count
The route deliberately narrows itself to single-answer Standard English questions so the score reflects one transparent token-level decision, not a debate about dialect or style. Several common grammar disputes are explicitly kept out of the bank:
- Collective nouns such as "team" where both singular and plural agreement are grammatical.
- Fewer versus less in measurement or idiomatic exceptions.
- Who versus whom as a register choice.
- Singular they as a pronoun preference.
- Present-perfect versus simple-past dialect preferences, including the got versus gotten split.
- Oxford-comma preferences, split infinitives, and sentence-final prepositions.
The bank also avoids article cases where pronunciation is unknown and count-or-mass nouns whose sense changes the answer. None of these items is treated as a wrong word in any sentence, and the page does not label a dialect feature, an informal construction, or a stylistic preference as universally defective. Readers therefore get a route that measures the visible token change rather than arguing about house style.
How Scoring, Errors, and Points Work
The score model is short enough to fit in your head before the first round. The full route begins at 1,200 points and never drops below zero, so even a rough first attempt still produces a final number to compare against later runs. A correct repair advances to the next sentence and does not change the running error count. A wrong word choice does not reveal the indexed answer; the screen only says that the chosen word is not the target. Once the correct token is found, a wrong correction choice adds one error and deducts up to 25 points, then leaves you on the same sentence so you can try the remaining three options. An old attempt cannot deduct twice or advance another round, because each click or keypress is validated against the current round, phase, and attempt tokens.
Tab, Enter, and Space continue to drive the native buttons alongside the number-key shortcuts, so the score reflects the same decisions whether you click or type. The spreadsheet-style boss overlay that hides the route while it is visible blocks Start, word selection, correction selection, keyboard answers, and Restart, and it does so without freezing the underlying state, so when you close it the same audited step is still waiting. Pointers and keyboard actions both feed the same guarded state transitions, which is why no input path can quietly slip past the rules.
Five Habits That Help on the 12-Sentence Route
Because the bank is fixed, a small reading routine will tighten your answers without changing what the quiz measures.
- Read the whole sentence aloud before clicking any word, so the verb form or pronoun reads cleanly in your ear before you commit to a token.
- Use the number keys when a word is visibly numbered. The route leaves the mouse path intact, but the shortcut is faster across twelve rounds.
- Compare every correction candidate against the rule scope printed beside the buttons, not just against the other three words.
- If the prompt looks ambiguous, remember that the bank excludes exactly the cases where two answers are both grammatical, so the prompt you are looking at has a single correct replacement.
- Treat any final point total as entertainment feedback rather than a proficiency estimate, and restart to repeat the same route whenever you want to compare runs.
The page does not upload answers, save a learner profile, adapt difficulty, rank users, or create a classroom record. A best point total may be retained locally by the shared game shell, but the sentence bank and scoring live entirely in the current browser, so closing the tab clears any record you did not want to keep. Use the route as a transparent source-backed puzzle and consult a current reference, editor, or teacher whenever real-world wording carries real consequences.