EIKEN Grade Pre-2 Plus common mistakes
EIKEN Grade Pre-2 Plus · Understand the traps and how to avoid them.
In the email/message response task, the candidate answers the prompt's question but never acknowledges the sender's stated situation or request, producing a reply that reads like a standalone paragraph rather than a message in a thread.
Under time pressure the candidate treats the task as a mini-essay and writes only the required content points, skipping the natural opening that refers back to what the sender wrote. Pre-2 candidates often have never written a real reply in English and have no model of message register.
What to do instead
Before writing, underline the sender's situation and every question in the prompt. Open with one sentence that mirrors the situation (e.g. thanking them for the invitation or apologising for the delay), then answer each question in its own sentence, and close with one short forward-looking line. Practise by rewriting three sample prompts where the only change is the opening sentence.
In the opinion essay, the candidate states a position and then lists two reasons with no support, so each body sentence is a bare claim like 'It is good because it is convenient.'
The candidate confuses 'giving a reason' with 'explaining a reason' and lacks the connective vocabulary (for example, for instance, this means that) to extend a point. Japanese-style paragraph logic also encourages stating the conclusion once and assuming the reader will infer the support.
What to do instead
Use a fixed two-sentence pattern per reason: claim plus one concrete example or consequence drawn from school or daily life. Drill by taking a bare claim and writing two follow-on sentences with 'For example' and 'This means that'. Aim for roughly 60-80 words total so there is room for support without running out of time.
On information-retrieval questions, the candidate reads the whole notice or timetable top to bottom and then picks the option that repeats the most words from the text, instead of locating the specific conditions the question asks about.
Pre-2 retrieval tasks embed two or three conditions (a date, a fee, a category such as 'members only') and the correct answer combines them, while the distractor quotes the text verbatim. Candidates match surface words because they have not been taught to scan for keywords first.
What to do instead
Read the question stem first and circle its conditions, then scan the passage only for those keywords, checking each condition against the candidate answer. Train with a two-column table: condition on the left, where in the text it is satisfied on the right. Deliberately practise rejecting the verbatim-match option.
On short and medium reading passages, the candidate answers detail questions from general knowledge or personal opinion rather than from what the passage actually says, and answers main-idea questions with a detail that appears early in the text.
Topic familiarity makes the candidate feel confident and skip the confirming sentence. For main idea, the first sentence is the most memorable, so it gets chosen even when the passage's point is stated or implied later.
What to do instead
For every answer, point to the exact line in the passage that supports it and say the line silently before choosing. For main-idea items, check that the chosen option covers the whole passage, not one paragraph, and reject options that are true but narrow. Do five passages where you must underline evidence before marking the answer.
In listening, the candidate picks the answer choice that contains a word or number heard in the audio, even when the speaker later corrects it or says it is not available.
Listening is processed word by word rather than for meaning, and Pre-2 dialogues routinely include a self-correction ('Actually, make that...') or a rejected option. The heard word feels safe, so the candidate stops listening after the first mention.
What to do instead
Train for correction signals: actually, instead, but, sorry, not until. Take notes as a small table (time, place, price) and update the entry whenever a correction marker appears. Practise with dialogues where the first number given is always wrong, so the habit of waiting for the final value is reinforced.
In longer talks and announcements, the candidate loses the thread after missing one sentence and stops listening, so the answer to a later question is missed entirely.
The candidate tries to hold every word in memory rather than tracking the structure of the talk, and a single unknown word triggers panic. Announcements also move fast with no chance to replay.
What to do instead
Preview the questions before the audio starts and note what each one asks for (a time, a reason, a person). During the talk, listen for signpost phrases (first, also, however, so, finally) and write one or two keywords per question rather than full sentences. If a word is missed, drop it and rejoin at the next signpost.
On the speaking read-aloud, the candidate reads word by word with flat intonation, pausing mid-phrase at line breaks or unfamiliar words, and pronouncing grammatical endings such as -ed and -s weakly or not at all.
The candidate has practised reading silently far more than aloud, so chunking and stress patterns are unfamiliar. Nervousness speeds up delivery and makes unfamiliar vocabulary cause a stall rather than a smooth skip-and-continue.
What to do instead
Before reading aloud, use the preparation seconds to mark phrase boundaries with slashes and circle the stressed word in each phrase. Read in thought groups, not word by word, and keep going past unknown words. Record yourself, then compare against a model reading for rhythm, paying specific attention to final -ed and plural -s.
On the illustration-description task, the candidate produces a list of isolated nouns or present-progressive sentences with no location or relation, e.g. 'A boy. He is running. A dog.' and runs out of things to say before the time limit.
There is no practised framework for scanning a picture, so the candidate describes whatever catches the eye first and then stalls. Pre-2 candidates also lack the prepositional phrases (in the middle, on the left, next to) needed to place people and objects.
What to do instead
Use a fixed scan order: overall scene, then left to right or foreground to background, giving each person an action plus a location. Build a bank of frames: 'In the foreground, a woman is ... while behind her ...'. Practise 30-second descriptions of three different pictures with a timer, aiming for five to six linked sentences rather than a list.
On the speaking opinion question, the candidate gives a one-sentence answer with a single reason and stops, or answers 'I don't know' or repeats the question, leaving the examiner with nothing to score for fluency and coherence.
The candidate has no rehearsed structure for spontaneous opinion and fears making a grammatical error, so brevity feels safer. Limited topic vocabulary makes the first reason hard to extend.
What to do instead
Rehearse a three-part frame: position, reason with a personal example, and a closing restatement ('So I think ...'). Keep a mental list of ten common school and lifestyle topics with one prepared example each. Practise answering in exactly three sentences under a 20-second limit so the length becomes automatic.
In the vocabulary, grammar, and usage section, the candidate chooses options by how familiar the word looks or by translating the whole sentence into Japanese, rather than by checking the grammatical slot the blank requires.
Pre-2 items test collocation, tense agreement, and word class, and translation-based choice fails when a Japanese equivalent fits but the English part of speech or preposition does not. Familiar-looking distractors exploit recognition memory.
What to do instead
Before looking at the options, decide what the blank needs: a noun, a verb form, or a preposition, and check the words on either side for clues (a following preposition, a time expression, a subject). Eliminate options that are the wrong word class first, then choose among the rest. Log every wrong item with the rule it tested and review the log weekly.