Miyazaki Entrance English common mistakes
Miyazaki Entrance English · Understand the traps and how to avoid them.
On main-idea items, the candidate picks the answer choice that repeats the most words from the first sentence of the passage, without checking whether it covers the whole passage.
Under time pressure the opening sentence is the most recently and vividly read, and lexical overlap feels like evidence. Passage openings often introduce a topic narrowly or with an example, so the highest-overlap option is frequently a detail, not the main idea.
What to do instead
Cover the choices and write a one-line summary of the passage in your own words first, then choose the option closest to your summary. Reject any option that matches only one paragraph, and check that the chosen option is broad enough to cover the passage's conclusion as well as its opening.
The candidate treats a plausible real-world inference as the answer on inference questions, choosing what 'must be true in general' rather than what the passage supports.
Inference items are designed with options that are true in everyday life but not stated or implied in the text. Candidates answer from background knowledge instead of from textual evidence, especially on familiar topics like school life or technology.
What to do instead
For every inference option, locate the specific sentence or pair of sentences that licenses it and say the link aloud: 'X plus Y implies Z.' If you cannot point to the text, eliminate the option even if it is factually true. Practice by marking one supporting line per inference question.
The candidate answers vocabulary-in-context items by recalling the most common dictionary meaning of the word rather than the meaning it carries in the sentence.
High-frequency words (run, hold, figure, matter, mean) have many senses, and the first-learned sense is retrieved automatically. The test deliberately selects a secondary or idiomatic sense, so the familiar gloss is often a trap.
What to do instead
Substitute each option back into the original sentence and read the whole sentence for coherence. Also check the word's grammatical role and the prepositions or nouns it collocates with; a change of part of speech or collocation usually signals a different sense.
The candidate builds collocations by translating a phrase word-for-word from their first language, producing forms like 'do a mistake', 'open the light', or 'very interested in' with the wrong preposition.
Collocation is arbitrary and stored as chunks, not derived from rules. Direct translation from Japanese produces grammatically legal but unnatural English, and candidates often cannot hear the difference because both versions seem logical.
What to do instead
Memorize collocations as whole chunks with their prepositions and articles (make a mistake, take an exam, be interested in, depend on). Keep a personal list of verb+noun and adjective+preposition pairs, and test yourself by filling gaps in full sentences rather than isolated words.
The candidate chooses a tense based on a time word alone, for example selecting past simple whenever 'yesterday' or 'last week' appears, even when the sentence requires past perfect or past continuous.
Textbook keyword-to-tense mapping is over-learned and applied mechanically. It fails when the sentence contains two past events with a sequence, an interrupted action, or a duration, where the time adverb does not determine the aspect.
What to do instead
Before choosing, identify how many events or states the sentence contains and their relationship (before/after, interrupted, ongoing, completed with present relevance). Then select the aspect that encodes that relationship, and confirm the auxiliary and participle forms are correct.
The candidate misses subject-verb agreement when the subject and verb are separated by a long phrase or when the subject is an indefinite pronoun, collective noun, or a 'there is/are' construction.
Agreement is computed locally: the noun nearest the verb feels like the subject. Inserted prepositional phrases, relative clauses, and expressions like 'a number of' or 'each of' override the true head noun in the reader's mind.
What to do instead
Bracket or cross out every phrase between the subject and the verb before checking agreement, then match the verb to the head noun only. Memorize the fixed cases: each/every/either/neither + singular, 'the number of' + singular, 'a number of' + plural.
On word-order and clause-construction items, the candidate accepts an option because every word is correct, without checking the order of adverb, object, and relative-clause constituents.
Candidates proofread for vocabulary and grammar errors but not for sequencing, and Japanese word order (verb-final, noun-modifying clauses) interferes with English placement of adverbs and relative clauses.
What to do instead
Check order separately from word choice: adverb placement relative to auxiliaries and objects, indirect object before direct object, and relative pronoun immediately after its antecedent. Rearrange the sentence into a statement first, then convert to the tested form (question, passive, cleft).
In free writing, the candidate produces one long block of sentences with no paragraph divisions, or writes paragraphs that each introduce a new topic without development.
Exam prompts reward speed, and paragraphing feels like decoration rather than structure. Many candidates have never been taught a paragraph template, so ideas are listed in the order they occur.
What to do instead
Use a fixed skeleton: topic sentence stating one claim, two or three supporting sentences with reasons or examples, and a closing sentence that links to the prompt. Before writing, jot a three-point outline and allocate one paragraph per point; check that each paragraph has exactly one controlling idea.
The candidate writes a free-composition answer that ignores part of the prompt, for example answering 'Do you agree?' with a description of the issue, or omitting the required reason and example.
Prompt analysis is skipped in the rush to start writing, and familiar topic vocabulary pulls the writer toward description rather than argument. The task type (opinion, compare, problem-solution) is not identified before drafting.
What to do instead
Underline the task verb and every required element in the prompt, then write a one-sentence thesis that directly answers the question. After drafting, re-read the prompt and tick each requirement against your text; add any missing element in the conclusion or a new paragraph.
In the listening section, the candidate answers while the audio is still playing, using the first number, name, or keyword heard rather than waiting for the speaker's correction or conclusion.
Working memory is overloaded by continuous speech, and candidates grab the first plausible detail for security. Dialogues frequently include self-correction ('actually, make it...'), negation, and distractors that reverse the first-mentioned fact.
What to do instead
Use the pause before each item to read the question and predict the answer type. Keep your pen still during the first mention, listen for contrast markers (but, actually, instead, however), and write the answer only after the speaker finishes the relevant turn. Practice with self-correction dialogues specifically.
The candidate responds to situational items with grammatically correct but pragmatically inappropriate English, for example 'Give me the salt' to a host or 'No, you are wrong' to a teacher, or uses casual forms in a formal request.
Politeness and register are rarely tested explicitly in grammar study, and directness norms differ between Japanese and English. Candidates optimize for clarity and grammatical accuracy, not for the social relationship.
What to do instead
Classify each situation by relationship (peer, superior, stranger) and by function (request, apology, refusal, disagreement) before choosing. Learn one softened pattern per function: 'Could you possibly...?', 'I'm afraid I...,' 'Would you mind if...?', and avoid bare imperatives with superiors.