Hiroshima Entrance English common mistakes
Hiroshima Entrance English · Understand the traps and how to avoid them.
When a long passage asks about the author's attitude, the candidate picks the option that merely restates a fact stated in the text (e.g., 'the author mentions that the town lost population') instead of an option that captures the stance (e.g., 'the author is critical of how the town handled the change').
Under time pressure, candidates match words they remember seeing in the passage to words in an option and stop reading; attitude/inference items are deliberately built so that a verbatim-echo option is a distractor, and the correct option usually paraphrases rather than quotes.
What to do instead
Before looking at the options, write a three-to-five-word label for the author's stance next to each paragraph (e.g., 'doubts the plan', 'sympathetic to workers'). Then eliminate any option that is only a factual restatement with no evaluative content, and require the surviving option to match one of your labels in meaning, not in wording.
The candidate answers a detail question from general knowledge or from the first mention of the topic rather than from the specific sentence the question targets, so answers about numbers, names, or sequence in the passage are wrong.
Long Hiroshima-style passages recycle the same nouns (the committee, the survey, the residents) across paragraphs, and candidates remember the gist of the first mention instead of locating the exact line; scanning skill is under-practised relative to reading for gist.
What to do instead
Train a two-step routine: identify the key word in the question (a number, a proper noun, a date), scan only for that key word, then read the whole sentence containing it plus the sentence before and after before choosing. Do ten minutes of pure scanning drills per day where you time yourself finding the target line without reading the rest.
On vocabulary-in-context items, the candidate supplies the most common dictionary meaning of the underlined word instead of the meaning it carries in that sentence (e.g., reading 'run' in 'run a campaign' as 'move quickly').
Vocabulary is studied as isolated word lists with one Japanese gloss per word, so the candidate never builds the habit of checking how the surrounding words constrain meaning; exam items deliberately choose polysemous words.
What to do instead
Study words in collocation chunks rather than singly ('run a campaign', 'run short of', 'run into trouble'), and on every practice item substitute each option back into the original sentence and read the whole sentence aloud to test whether it is coherent. Keep a personal log of words you got wrong with the sentence, not just the gloss.
The candidate confuses near-synonyms that differ in grammatical pattern or register (e.g., choosing 'provide' where the sentence needs 'provide A with B' versus 'supply A to B', or choosing an informal phrasal verb in a formal written sentence).
Word lists group synonyms by Japanese translation, which hides differences in complement structure and formality; candidates learn meaning but not the grammatical frame each word requires.
What to do instead
For each new verb, record its pattern explicitly ('provide sb with sth', 'supply sth to sb', 'prevent sb from doing'). When an option is a synonym of another option, treat the pair as a signal that the item is testing pattern or register, and decide on the frame rather than the meaning.
In sentence-completion and error-correction items, the candidate checks only whether the sentence 'sounds right' and misses a structural fault such as a missing relative pronoun, a doubled subject, or a participle with the wrong subject (dangling modifier).
Japanese learners often process English clause by clause and accept structures that Japanese would allow (e.g., a resumptive pronoun, or a modifier whose subject is unclear), because the sentence is semantically transparent even when it is ungrammatical.
What to do instead
Adopt a fixed checklist to run on every sentence: (1) one finite verb per clause, (2) every relative clause has a subject or object gap, (3) the subject of an initial participle is the main-clause subject, (4) articles and plurals agree with the noun type. Practise by correcting ten ungrammatical sentences a day and naming which checklist item each one breaks.
When rewriting or combining sentences as instructed (e.g., 'rewrite using the passive', 'combine using a relative clause'), the candidate changes the meaning or drops information, producing a grammatical sentence that no longer says the same thing.
The candidate focuses on producing the target structure and treats the original content as secondary, so adverbial information, negation, or the original emphasis is lost in the transformation.
What to do instead
After each rewrite, compare it to the original clause by clause and tick off every piece of information (who, what, when, negation, emphasis). Practise transformations in pairs where you list what must be preserved before writing, and check that the topic/focus of the sentence is unchanged, not just the truth value.
In guided writing, the candidate answers the prompt with a list of short, loosely connected sentences and no clear overall claim, so the response reads as a set of notes rather than a paragraph with a position.
Candidates prepare by memorising sentence patterns and transition phrases, which encourages stringing patterns together; they rarely practise planning a claim plus two supporting reasons within a word limit.
What to do instead
Before writing, spend two minutes on a three-line plan: a one-sentence answer to the prompt, two reasons, and one example or consequence. Write the first sentence to state the claim explicitly, start each body sentence with a linking word that names the relation (because, however, for example, as a result), and finish with a sentence that restates the claim without new information.
On Japanese-to-English translation, the candidate translates word by word in Japanese order, producing sentences with the verb at the end, missing subjects, and unnatural articles (e.g., 'In Japan, is thought that many people...').
The source text omits subjects and uses topic-comment structure, and the candidate maps each Japanese chunk to one English word instead of first deciding the English clause structure.
What to do instead
Use a two-pass method: first read the Japanese and write a bare English skeleton (subject + verb + object) for each clause, supplying the subject from context; then add the modifiers. After translating, check every sentence for an explicit subject, correct tense, and articles on countable nouns, and read it aloud to catch Japanese word order.
On English-to-Japanese translation, the candidate keeps English word order and produces Japanese that is grammatically possible but unidiomatic, with literal renderings of idioms and phrasal verbs (e.g., translating 'make up for' as 上を作る rather than 埋め合わせる).
The candidate treats idioms as sums of their parts and does not have a stock of Japanese equivalents for common English multi-word expressions; fear of departing from the source wording also blocks natural reordering.
What to do instead
Build a two-way idiom notebook organised by function (apology, contrast, cause, concession) and practise translating whole sentences rather than phrases. After drafting, restructure the Japanese so the verb comes at the end and the topic is marked with は/が appropriately, then reread for naturalness as Japanese, not as a mirror of the English.
In listening, the candidate writes the answer while the audio is still playing, then misses the next piece of information; or answers from a single word heard rather than from the full statement, so items involving numbers, negation, or correction ('not on Monday, the Tuesday after') are wrong.
Note-taking is untrained and candidates try to capture everything verbatim; distractor design in listening items relies on a keyword being mentioned and then revised or negated.
What to do instead
Practise taking abbreviated notes in a fixed layout (numbers in one column, names in another) and leave the answer sheet untouched until the relevant question is reached. Train specifically for correction signals ('actually', 'sorry, I mean', 'instead') and for negation, and drill short audio clips where a number is stated, then changed, so you learn to wait for the final version.
On paragraph-ordering items, the candidate orders sentences by topic similarity rather than by discourse signals, placing a sentence that introduces a new example before the general statement it illustrates, or putting a concluding sentence in the middle.
Candidates read each sentence for meaning in isolation and do not track pronouns, demonstratives, and discourse markers (this, such, however, for instance, in conclusion) that signal what must come before and after.
What to do instead
Circle every pronoun and demonstrative and every discourse marker, and write beside each sentence what it must follow (e.g., 'this trend → needs a trend mentioned before'). Build the chain from the strongest link rather than from the first sentence, and verify the finished order by checking that each transition word makes sense at that position.
On summary and information-transfer tasks, the candidate copies sentences from the source passage or fills a table with details rather than the required category, so the summary exceeds the word limit or the transferred information does not match the column heading.
Copying feels safe and is faster than compressing, and candidates do not check each item against the column heading or the word limit before moving on.
What to do instead
For summaries, draft in note form first (one noun phrase per main idea), then write full sentences from the notes and count words against the limit. For information transfer, read the column headings before the passage, and for each cell ask 'does this answer this heading, in this grammatical form?' (e.g., a date, a reason, a person), correcting the form as well as the content.