MET (Michigan English Tests) cheat sheet

MET Writing contains two tasks: a shorter transactional message (email or letter) and a longer opinion or discussion essay.

Ratings weigh task completion, organisation, grammatical range and accuracy, vocabulary, and mechanics; an off-topic response cannot score well however accurate its sentences are.

Email decision rule: state the purpose in the first sentence, address every element named in the prompt, and close with a sign-off matching the required formality.

Essay formula: paraphrase the prompt, state a thesis, write two or three body paragraphs (one controlling idea each, supported by an example), then restate the position in the conclusion.

Under-length responses lose content marks, so aim for the upper half of the stated word range rather than the minimum.

Vary sentence openings and structures; beginning every sentence with 'I' or 'The' limits the range score even when grammar is correct.

Tense rule: past for narrative examples, present for general claims; do not switch tense inside a sentence without an explicit time signal.

Formal register markers: avoid contractions, slang, and generic 'you' in the essay; use full forms and impersonal structures such as 'it is argued that'.

The section runs in three parts: short exchanges, longer conversations, then talks or monologues, with difficulty increasing across the section.

The recording is played once; question stems are printed in advance and follow the order of the audio.

Elimination rule: an option that repeats words heard but does not address the question's focus is a distractor; correct options are usually paraphrases.

Pre-read the stems during the pause and predict the answer category (person, place, time, number, reason) before the audio begins.

In talks, the opening sentence usually announces the topic and the closing sentence often restates the main point.

Correction pattern: numbers, prices, and times are frequently revised mid-sentence, so transcribe the final version rather than the first.

Tone and attitude items are decided from adjectives, stress, and intonation rather than from content words alone.

There is no penalty for incorrect answers, so every item should be marked; a four-option item gives a 25% chance when guessed.

The section moves from short practical texts to longer passages, with more items attached to the longer texts.

Within a question set, items generally follow the order of the passage, so answers can be located progressively from top to bottom.

Vocabulary items are solved by substituting each option into the original sentence and keeping the one that preserves the sentence's logic.

Elimination rule: options containing always, never, all, or none are wrong unless the text itself uses equally absolute language.

Main-idea rule: the correct option must be true of every paragraph; an option that fits one detail is a distractor.

Time budget: roughly one minute per item; flag and return rather than stalling on a single question.

Reference items: trace a pronoun to the nearest preceding noun phrase that agrees with it in number.

Sentence-insertion items: check both the sentence before and the sentence after the gap for linking words and pronoun reference.

Speaking is a short interview-style test of roughly 10 to 15 minutes, recorded and rated by trained examiners.

Tasks typically progress from personal warm-up questions to describing a picture or prompt and then giving an opinion on a topic.

Opinion formula: direct answer, then a reason, then an example; three to four sentences per turn gives raters enough language to assess.

Silence is the main fluency penalty; fillers such as 'well' or 'let me think' hold the turn while the next idea is planned.

Picture description relies on present continuous verbs for actions plus positional language such as 'in the background' or 'next to'.

Read-aloud performance is scored on chunking and intelligibility, not speed; pausing at commas and full stops is expected.

Grammatical range counts, so attempt complex structures such as conditionals and relative clauses even at some risk of error.

Self-correction rule: repair only errors that block meaning; constant restarting lowers the fluency rating.