EJU common mistakes
EJU · Understand the traps and how to avoid them.
In the Japanese as a Foreign Language reading section, reading each passage word by word from the first line and trying to understand every unknown kanji before moving on, so the last questions are answered with seconds left.
Candidates trained on sentence-translation tests treat reading as full decoding. EJU reading is timed at roughly one question per minute and rewards locating the paragraph that answers the question, not comprehending the whole text. Unknown vocabulary triggers panic and stalling.
What to do instead
Practise a three-step routine on every passage: read the question stem first, scan for the keyword or its paraphrase, then read only the surrounding two to three sentences closely. Deliberately skip unknown kanji when the surrounding grammar carries the meaning. Do full 40-minute timed sets weekly and log where the clock was lost.
In Japanese listening and listening-reading, taking notes on every word and therefore missing the next utterance, or answering from a single keyword heard early in the track instead of from the conclusion.
School note-taking habits carry over, but EJU listening tasks hinge on the final decision or the speaker's correction of an earlier statement. Candidates also latch onto a word that appears in an answer choice without checking whether the speaker affirmed or rejected it.
What to do instead
Train note-taking with symbols only: arrows for change of plan, X for rejected options, a circle around the final decision. Before each track, predict from the question what kind of ending to listen for (a number, a reason, a refusal). Drill tasks where the first-mentioned option is explicitly overturned so you learn to wait for the conclusion.
In Japanese writing, producing a memorised essay on a familiar topic regardless of the prompt, or writing long complex sentences with inconsistent desu/masu and unclear paragraph structure.
The writing task is scored on relevance to the prompt, organisation, vocabulary and grammar range, and correct register. Candidates who have memorised templates force the topic to fit the template, and candidates who chase complexity lose points on sentence breaks and consistency.
What to do instead
Spend the first three minutes underlining the prompt's key instruction and writing a two-line plan: position plus two supporting reasons. Use one register (desu/masu) throughout and keep sentences to a length you can finish correctly. Practise 20 prompts in 20-minute blocks, then self-check against four criteria: task relevance, structure, vocabulary variety, grammar accuracy.
In Mathematics, attempting every problem in order and spending ten minutes on one difficult integration or probability question, leaving easy early questions unfinished or unchecked.
EJU Mathematics is scored on accuracy within a fixed time, and the paper mixes routine and demanding items. Candidates treat it as a school test where completion matters, and they fear leaving blanks.
What to do instead
Do two passes: first pass solves everything you recognise within about one minute each, second pass returns to the rest. Before the exam, rank the syllabus topics by your own accuracy and decide a cut-off time per question (for example 4 minutes). In practice, mark any question exceeding the cut-off and move on, then review it at the end.
In Mathematics, losing marks on the final numerical answer because of a sign error, a misread condition such as 'integer' or 'positive', or forgetting to convert units and round as instructed.
EJU Mathematics uses a fill-in format that often asks for a specific form, so a correct method with a single arithmetic slip produces a wrong entry. Candidates also skim conditions under time pressure and carry intermediate rounding errors.
What to do instead
Adopt a written checklist applied at the end of each question: reread the condition line, verify the sign of the final substitution, confirm the requested form (fraction, decimal, integer), and re-derive the last two lines. Practise with the answer sheet format so you get used to entering values in the required shape.
In Science (Physics, Chemistry, Biology), memorising formulas and terms without being able to apply them to an unfamiliar experimental setup or graph, and skipping the graph-reading questions.
EJU Science items frequently present data, experimental descriptions, or combined situations rather than direct recall. Candidates who study by rereading textbooks recognise the vocabulary but cannot translate a described setup into equations or interpret a slope.
What to do instead
For each topic, practise converting a described situation into a diagram and equations before looking at the answer. Spend a fixed share of study time on graph and data interpretation: identify axes, units, slope meaning, and intercept meaning. Work past-paper style experimental questions and write one sentence explaining what the experiment measures.
In Science, answering calculation questions without carrying units through the working, or mixing up similar quantities and constants (for example energy versus power, molarity versus moles, or the direction of a force).
Candidates learn formulas as symbol strings and substitute numbers without checking dimensional consistency. Under time pressure, similar-looking definitions collapse into each other, and answers that are numerically plausible but dimensionally wrong go unnoticed.
What to do instead
Write units next to every substituted value and cancel them explicitly. Keep a one-page sheet of confusable pairs with their units and a physical example of each, and review it before practice sets. When an answer looks odd, check the order of magnitude against a known reference value.
In Japan and the World, treating the subject as pure fact recall and memorising dates and names in isolation, without linking events to causes, consequences, or the source-based reasoning the questions require.
The subject combines history, geography, economics and contemporary society and often asks candidates to interpret data, maps, or short texts. Candidates who study only lists cannot answer questions that ask why a policy was adopted or what a graph implies.
What to do instead
Build cause-and-effect chains for each major period and theme: condition, policy response, outcome, later consequence. For data and map questions, practise stating what the figure shows, what it does not show, and which option the evidence supports. Test yourself with 'why' and 'what followed' questions rather than date recitation.
In Japan and the World, relying on general knowledge or current news from the home country to answer questions about Japanese and international institutions, and ignoring the specific periodisation and terminology used in the syllabus.
Candidates assume the subject rewards broad awareness, but questions use precise framing (for example, a named era, a specific treaty, or a defined economic indicator) and expect the syllabus's own categories. Home-country examples can actively mislead.
What to do instead
Study from a syllabus-mapped outline and, for every term, write a one-line definition with its time period and a concrete example. When a question names a period or institution, answer strictly within that frame. Review past-paper style questions to see which categories recur, and build flashcards keyed to those categories.
Across all domains, sitting full past papers only at the end and never simulating the real sequence and timing, so stamina, section transitions, and answer-sheet discipline are untested.
EJU runs several subjects in one day with strict timing and separate answer sheets. Candidates who study topic by topic never experience fatigue, the switch between Japanese and content subjects, or the risk of misaligning answers on the sheet.
What to do instead
Schedule at least four full simulations in the final two months, in the real order and duration, including breaks. After each, record time per section, errors caused by fatigue, and any answer-sheet mistakes. Adjust your per-question cut-offs and your strategy for the final ten minutes of each section based on that log.