KZ Foreign Language Exam common mistakes
KZ Foreign Language Exam · Understand the traps and how to avoid them.
In the Listening section, candidates read the answer options in detail while the recording is still playing, then miss the next two or three utterances entirely.
The recording plays once, and candidates treat the pause before it as reading time rather than using the pre-listening window; anxiety about the options pulls attention away from the audio stream, and working memory cannot hold both the text of the options and the incoming speech.
What to do instead
Use the pre-listening seconds to scan only the stems and mark one keyword per option (a number, a place, a verb). During playback keep your eyes on the stem, not the options, and answer from memory immediately after each item. Practise with single-play recordings of 30–45 seconds and force yourself to write nothing while the audio runs.
Candidates hear a word from an option in the recording and select that option, even when the speaker says 'I didn't say X' or 'not X, but Y'.
Phonological matching is the cheapest strategy available under time pressure, so the brain latches onto the first familiar string. Negation, contrastive stress ('I wanted the BLUE one'), and distractors built from the same lexical field are exactly what the item writer exploits.
What to do instead
Train on minimal-pair and negation sets: record yourself saying pairs like 'I did want it' vs 'I didn't want it' and transcribe. In practice, before choosing, state aloud what the speaker actually asserted, and physically cross out any option whose only support is a repeated word rather than a repeated meaning.
In Reading, candidates read the passage word by word, then read the questions, then reread the passage, and run out of time before the final passage.
School habit rewards full comprehension of every sentence. Exam passages contain 30–40% material that no item tests, and the candidate has no time budget, so the first passage absorbs the time meant for three.
What to do instead
Adopt a fixed three-pass routine: (1) read the questions and underline the locating keyword; (2) scan for that keyword and read only the sentence plus one sentence either side; (3) answer and move on. Set a hard per-passage timer in practice (e.g. 8 minutes for a 4-question passage) and leave the last 5 minutes for transfer of answers.
Candidates answer global questions ('What is the main idea?', 'What is the author's attitude?') from a single striking sentence, usually the first or last, without checking the whole text.
Topic sentences are salient and feel authoritative, but exam passages often open with a counter-argument the author later rejects. The candidate also confuses a true detail with the main idea because both are true statements about the text.
What to do instead
For global items, cover the options, write a six-word summary of the passage yourself, then match. Eliminate any option that is true but narrow (a detail) or true but too broad (beyond the text). Practise by labelling each paragraph's function (claim, evidence, concession, conclusion) before answering.
In Lexical and Grammatical Competence, candidates choose the option that 'sounds right' in their own language's pattern, e.g. picking a wrong preposition, aspect, or article because the direct translation of their first language would use it.
Collocation and government (verb + preposition, case, aspect) are stored as chunks and are the last layer of a language to become automatic. Under time pressure the first-language pattern wins, and the candidate has no conscious rule to check against.
What to do instead
Keep a personal error log of only the patterns you personally get wrong (not a general grammar list), with the full chunk written out: 'depend on', 'be interested in', 'despite + noun'. Review it weekly and test yourself by producing the chunk from the noun, not by rereading the list.
Candidates treat multiple-choice grammar items as a search for the one correct form and stop at the first plausible option, without checking whether a second option is also grammatical but ruled out by context.
Many items are designed with two or three grammatically possible forms where only one fits the surrounding tense, register, or logic. The candidate reads the gap in isolation rather than the whole sentence.
What to do instead
For every gap item, read the full sentence before looking at the options, decide what meaning and time reference are needed, then check each option against that requirement. In practice, write down why each rejected option fails (wrong tense, wrong register, wrong agreement), not just which one you chose.
In Writing, candidates produce an essay that answers the prompt in content but ignores the required genre features - no clear position statement, no paragraph topic sentences, or an informal register in a formal task.
Candidates prepare topics rather than task types, and under exam pressure they write what they know about the subject. They also transfer spoken fluency markers ('well', 'you know', contractions everywhere) into written text.
What to do instead
Build a one-page template per task type (opinion, argument, formal letter, report) with the required moves in order. Before writing, spend 3 minutes on a skeleton: position, two supporting points with one example each, concession. Then write to the skeleton, and reserve 4 minutes to check register, articles, and verb endings.
Candidates spend 25 of 40 writing minutes on planning and drafting the first paragraph, then finish with a rushed, underdeveloped conclusion and no time to proofread.
Perfectionism attaches to the opening sentences, which are the most visible, and there is no timed practice history to calibrate effort. The result is a strong start and a weak, error-dense finish that the marker sees last.
What to do instead
Rehearse the clock: 3–5 minutes planning, 25 minutes writing, 5–7 minutes checking, and never rewrite the introduction. Practise full timed essays weekly and mark where the time went; if the plan exceeds 5 minutes, cap it with a timer that you cannot ignore.
In Speaking, candidates memorise answers to predicted questions and deliver them verbatim, then freeze when the examiner asks a follow-up that does not fit the script.
Memorised chunks feel like fluency insurance, but examiners are trained to probe and to notice recited intonation. The candidate has practised production, not interaction, so turn-taking and repair strategies are untrained.
What to do instead
Prepare topic vocabulary and two personal examples per topic, never full sentences. Practise with a partner who must ask one unscripted follow-up per answer, and drill three repair phrases ('Let me put that another way…', 'What I mean is…', 'Could you rephrase that?') until they are automatic.
In Speaking, candidates give one- or two-sentence answers to every question, treating the exam as a short interview rather than an opportunity to demonstrate range.
Short answers are safe: fewer errors, less risk of losing the thread. But raters assess range of structures and discourse, and a string of minimal answers cannot show it, so the score is capped regardless of accuracy.
What to do instead
Use a fixed extension frame for every answer: direct answer, one reason, one concrete example, one contrast or result. Time yourself producing 30–45 seconds per answer on random prompts, and record and replay to confirm you actually used a complex sentence and a linking phrase each time.