Oxford ELLT cheat sheet

The Listening section is delivered once, audio plays a single time for each task, and answers are entered as you go rather than after a replay.

Task types typically include multiple-choice, matching, form/note completion, and short-answer items; question stems appear before the audio starts, so use the pre-audio window to predict answer type (number, name, place, reason).

Word limits on completion items are enforced literally: 'NO MORE THAN TWO WORDS' rejects three words and rejects hyphenated compounds counted as two.

Spelling and grammar are marked in Listening completion tasks; a correct-sounding answer with wrong spelling or wrong singular/plural form scores zero.

Distractors follow a correction pattern: a speaker states a plan, then revises it ('Actually, make that...'), so the second stated value is usually the answer.

Numbers, dates and times are frequent targets; note that 'half past', 'quarter to', and 24-hour clock forms may be read aloud differently from how they must be written.

Signposting language ('firstly', 'moving on', 'to sum up') marks transitions between question groups; losing the marker usually means losing the next answer.

In matching tasks with more options than items, options are not reused unless stated; track used letters to avoid duplicate answers.

Reading is timed as a single block covering all passages; there is no separate transfer time, so answers must be recorded in the test interface as you read.

Question order within a passage normally follows the order of information in the text, except for matching-headings and matching-features tasks, which are not sequential.

Completion tasks enforce word limits strictly; articles and numbers count as words, and contracted forms are usually disallowed.

True/False/Not Given hinges on the distinction: False means the text contradicts the statement; Not Given means the text is silent on it, even if the statement is plausible.

Matching-headings answers describe the whole paragraph's main idea, not a detail; a heading that matches one sentence is usually a trap.

Skimming for gist and scanning for names, dates and capitalized terms are separate passes: locate the answer zone first, then read that zone closely.

Answers are marked on exact wording for completion items, but paraphrased meaning for multiple-choice; do not copy text into multiple-choice options.

Time allocation rule of thumb: roughly 20 minutes per passage, with the final minutes reserved for transferring and checking spelling.

Two tasks: Task 1 requires a minimum of 150 words and Task 2 a minimum of 250 words; under-length responses lose marks for task achievement, and suggested timing is about 20 and 40 minutes respectively.

Task 2 carries roughly twice the weight of Task 1 in the overall Writing score, so protect time for it.

Assessment uses four criteria: Task Achievement/Response, Coherence and Cohesion, Lexical Resource, and Grammatical Range and Accuracy, each contributing equally.

Task 1 (Academic) reports data from a chart, graph, table or process; it must include an overview of the main trends, with no personal opinion or causes.

Task 1 (General) is a letter with a defined purpose, recipient and register (formal, semi-formal, informal); register consistency is marked.

Task 2 requires a clear position sustained throughout, not just in the introduction and conclusion; each body paragraph should develop one controlling idea with an example.

Coherence and Cohesion penalizes mechanical linkers; overuse of 'Moreover' and 'Furthermore' at every sentence start reads as memorized language.

Memorized or copied phrases are discounted by examiners and can lower Lexical Resource; original phrasing tied to the specific prompt scores higher.

Three parts: Part 1 is short personal questions on familiar topics (roughly 4-5 minutes), Part 2 is a 1-minute preparation plus a 2-minute individual long turn, Part 3 is abstract discussion linked to the Part 2 topic.

Assessment uses four criteria: Fluency and Coherence, Lexical Resource, Grammatical Range and Accuracy, and Pronunciation, weighted equally.

Part 2 cue cards include bullet prompts plus a final 'explain' prompt; covering all bullets is not required, but sustained speech for the full 2 minutes is expected.

The examiner stops the long turn at 2 minutes; running short of time signals insufficient fluency development and limits the score.

Fluency is judged on flow and repair, not accent; self-correction is acceptable, but long pauses and repeated restarting lower the band.

Part 3 rewards extended answers with reasons, comparisons and speculation; one-line answers cap the score regardless of grammar accuracy.

Pronunciation is rated on intelligibility and feature use (stress, intonation, connected speech), not on native-like accent.

Answers are recorded; examiners assess from the recording, so clarity of volume and pace matters as much as content.

Map and plan-labelling tasks require tracking directional language ('just past', 'opposite', 'at the end of'); orientation is fixed to the map, not the speaker's movement.

Multiple-choice options are often paraphrased in the audio; the option that repeats the speaker's exact words is frequently the distractor.

Answers arrive in the same order as the questions within a section, so a missed answer means moving forward rather than waiting.

Section difficulty generally increases: earlier sections are transactional and concrete, later sections are academic monologue or discussion with abstract vocabulary.

Numbers with similar sounds (13/30, 14/40) are common traps; stress and '-teen' versus '-ty' endings distinguish them.

If a speaker spells a name aloud, write exactly the letters given, including doubled letters, rather than guessing a conventional spelling.

Answer sheets accept one answer per item; writing two alternatives scores zero even if one is correct.

Matching-features tasks (e.g., matching researchers to statements) allow options to be used more than once unless the instructions say otherwise; check the instruction wording before answering.

Sentence-completion items must be grammatically compatible with the stem; an answer that breaks the surrounding grammar is wrong even if factually present in the text.

Summary-completion with a word bank may include more words than gaps; unused options are distractors, and each option is normally used once.

Passage topics are academic (science, history, environment, society) and passages increase in difficulty; the third passage is usually the longest and most abstract.

For 'which paragraph contains the information' tasks, answers are not sequential and a paragraph may be used more than once.

Skimming should target the first and last sentences of paragraphs, where topic sentences and conclusions usually sit.

Do not rely on general knowledge: answers must be supported by the passage, and plausible outside facts score zero.

Flag-and-return is the correct handling for a stalled item; spending more than about 90 seconds on one question costs later easy marks.

Task 1 data description should select and report key features (highest, lowest, main change) rather than list every figure; listing all numbers without selection limits Task Achievement.

A Task 1 overview is normally placed after the introduction and states two or three general trends without numbers.

Task 2 essay structures that score well share: an introduction that paraphrases the prompt and states a position, two or three body paragraphs, and a conclusion that restates the position without new ideas.

Word counts exclude the title and anything copied from the prompt; copied prompt text is not counted toward the 150/250 minimum.

Grammar range is shown by complex sentences (relative clauses, conditionals, concession) used accurately, not by length alone.

Spelling and punctuation errors count under Lexical Resource and Grammatical Range; consistent errors in articles and subject-verb agreement are high-frequency deductions.

Off-topic responses score low on Task Response regardless of language quality; every paragraph must serve the exact question asked.

Legible handwriting matters for paper-based delivery; illegible words are marked as errors.

Part 1 answers of two to three sentences with a reason are the target length; single-word answers and memorized speeches both depress the score.

In Part 2, notes made in the 1-minute preparation should be keywords, not full sentences, so delivery stays natural.

If the examiner asks a follow-up in Part 2 ('Does this place still exist?'), answer briefly and let the examiner lead.

Lexical Resource rewards less common vocabulary and paraphrase; repeating the question's words verbatim signals limited range.

Grammatical Range is shown by mixing simple, compound and complex structures, including accurate tense use when narrating past events.

Fillers such as 'well', 'I suppose' and 'let me think' are acceptable fluency devices; long silent pauses and L1 words are not.

Part 3 questions often ask for comparison, prediction or evaluation; structuring answers as 'opinion + reason + example' sustains length.

The interview is a fixed script; examiners cannot rephrase questions, so asking for repetition is allowed but repeated requests reduce fluency impressions.

Note-completion tables and flowcharts depend on layout logic: answers fill the gap that fits the column heading or arrow direction, so read the table structure before audio begins.

Distractors often appear as a speaker rejecting an option ('not the library, it's closed') before giving the real answer.

Currency, fractions and decimals may be spoken in words; write the numeral form the instructions permit.

In discussion tasks with two speakers, the answer is usually the one confirmed by the other speaker's agreement, not the first suggestion.

Section breaks give time to read ahead; use them to underline the key word in each question stem that signals the answer type.

Answers that are grammatically plural in the stem require plural forms in the gap; singular/plural mismatch is a common avoidable error.

Vocabulary-in-context items require the meaning as used in the passage, not the most common dictionary meaning.

Inference questions ask what is implied but not stated; the answer must follow logically from the text, not from the reader's opinion.

Reference questions ('what does "this" refer to') are resolved by reading the preceding sentence for the nearest compatible noun phrase.

Scanning for numbers and proper nouns is faster than reading sequentially; use it to locate the paragraph before detailed reading.

Multiple-choice Reading items often have two partially true options; the correct option answers the whole question, not just part of it.

Paragraph-matching for main idea requires distinguishing topic (what it is about) from main idea (what it claims about the topic).

Time pressure is the main failure mode; unanswered items should be guessed rather than left blank, since there is no penalty for wrong answers.

Task 1 process diagrams require sequencing language (first, subsequently, once..., finally) and passive voice where the agent is unspecified.

Comparison language for Task 1 includes 'twice as many as', 'marginally higher than', 'overtook', and 'remained stable at'; accuracy of the comparison matters more than variety.

Task 2 discussion essays must cover both views plus the writer's own position; omitting one view caps Task Response.

Problem-solution essays need both problems and solutions in balanced proportion; a solutions-only essay is partially off-task.

Cohesion is achieved through referencing (this trend, such measures, the former) as well as linkers; pronoun and referencing errors reduce the score.

Paragraphing counts: one idea per paragraph with a clear topic sentence is the expected pattern.

Checking time should target the highest-frequency errors: articles, plurals, verb tense consistency, and sentence fragments.

Part 2 cue cards usually ask about a past experience, a person, a place or an object; tense choice should match the prompt (past for experiences, present for descriptions).

Extending a long turn is done by adding detail layers: what happened, who was involved, how you felt, and why it mattered.

Pronunciation features assessed include word stress, sentence stress for emphasis, and intonation for lists and questions.

Connected speech (linking, elision) aids fluency but must not reduce intelligibility; clarity outranks speed.

Repeating the examiner's question back as a paraphrase buys thinking time and demonstrates lexical range.

In Part 3, hedging language ('tends to', 'in most cases', 'it depends on') supports balanced, extended answers on abstract topics.

Interrupting or dominating the conversation is not possible in the scripted format; the examiner controls turn length.

Instructions state the word limit per task; the limit applies to each answer, not to the section as a whole.

Dates may be spoken as ordinals ('the third of May') but written as numerals or month-first depending on the format shown in the question.

Telephone numbers and postcodes are read digit by digit, with 'double' used for repeated digits.

When a speaker self-corrects a name or figure, the correction supersedes the first mention; write the corrected version.

In labelling diagrams, prepositions of place ('adjacent to', 'beneath', 'to the left of') determine the label position.

Answers are usually concrete nouns, numbers or short noun phrases; full sentences in a gap are almost always wrong.

Yes/No/Not Given items concern the writer's views or claims, whereas True/False/Not Given concern facts; mixing the two sets of labels is a scoring error.

For matching sentence endings, the grammatical subject and tense of the stem must agree with the chosen ending.

Skimming a passage's title, subheadings and first paragraph predicts content and vocabulary, improving scan speed.

Answers must be copied exactly from the text for completion items, including any plural or capital form used in the passage.

Long passages reward paragraph-by-paragraph work: read the questions for a paragraph group, then read that paragraph, rather than reading the whole passage first.

Distractor answers in Reading often reuse exact words from the passage in a context that changes the meaning.

Task 2 question types include opinion (agree/disagree), discussion (both views), advantage-disadvantage, problem-solution, and two-part questions; the type determines the essay structure.

An introduction should not include the main body's examples; it states the topic, the paraphrase, and the position in about two sentences.

Examples in Task 2 should be plausible and specific but need not be statistically true; invented statistics are discouraged, generalized examples are acceptable.

Lexical Resource rewards collocation accuracy (e.g., 'pose a risk', 'address an issue') over rare single words used incorrectly.

Grammatical Range and Accuracy penalizes errors that impede communication more heavily than slips that do not.

Task 1 must not include reasons or speculation about causes; that is an opinion and lowers Task Achievement.

The conclusion should not introduce a new argument, as it will be treated as undeveloped.

Fluency and Coherence includes the ability to develop topics logically, not just speak without pausing.

Lexical Resource is rated on paraphrase ability: when a word is unknown, describing it ('the thing you use to...') demonstrates range.

Grammatical Range and Accuracy counts error density and whether errors cause misunderstanding, not error count alone.

Pronunciation band descriptors emphasize sustained intelligibility across the whole test, so a few unclear words do not cap the score.

Part 2 notes may be consulted during the long turn; glancing at keywords is permitted and does not reduce the score.

Answering 'I don't know' ends the turn; reframing the question and speculating keeps the interaction going.

The full Speaking test lasts roughly 11-14 minutes; Part 2's 2 minutes is the only fixed-length turn.

Answer transfer is not separate in the computer-delivered format; typing answers directly means spell-check is unavailable and must be done mentally.

Synonyms are the main listening skill: the question uses one word, the audio uses another with the same meaning.

In multiple-answer items ('choose TWO letters'), both answers must be correct for credit in most formats; partial credit is not given.

Pace cues ('before we move on', 'finally') signal that the current question group is ending.

Numbers in the thousands may be spoken with 'and' or as paired digits; write the full numeral.

If an answer is missed, leave it and rejoin at the next question; audio does not pause for recovery.

The Reading section has no extra transfer time; answers typed or written during the timed block are final.

Headings tasks penalize over-specific headings that match a single detail rather than the paragraph's overall claim.

Transition words in the passage ('however', 'by contrast', 'consequently') often mark the sentence that contains the answer.

In matching-information tasks, the same paragraph can answer more than one question; do not eliminate a paragraph after one use.

Reading speed target is roughly 200-250 words per minute for gist, slower for answer zones.

Guessing strategy: eliminate options contradicted by the passage, then choose the remaining option that is supported by a specific line.

Task 1 line graphs require tense control: past for completed periods, present perfect for trends continuing to now, future forms only for projections shown on the axis.

Task 1 bar and pie charts require proportion language (a third, the majority, a negligible share) rather than raw numbers alone.

Task 2 opinion essays must state the position in the introduction and maintain it; a balanced 'both sides' essay without a stance loses Task Response marks.

Coherence and Cohesion includes paragraph organization and logical progression, not only linking words.

Sentence variety includes fronted adverbials, conditional clauses and relative clauses; overlong sentences with comma splices reduce accuracy.

Task 1 tables with many categories require grouping: report the largest group and the most notable exception rather than every row.

Time management rule: complete Task 2 first if Task 1 tends to overrun, since Task 2 carries more weight.

Part 1 topics are predictable (home, work/study, hobbies, food, weather); prepared vocabulary on these domains improves range.

In Part 2, the final cue bullet ('and explain why...') is the evaluative element; addressing it shows higher-level development.

Part 3 answers benefit from concession ('although..., I still think...') which demonstrates grammatical range and balanced reasoning.

Pronunciation of word endings (-s, -ed) affects grammar intelligibility and is noticed by examiners.

Speaking at a natural pace with clear stress is rated higher than fast speech with unclear articulation.

The examiner's manner is neutral by design; lack of feedback during the test is not a signal of poor performance.

Form-completion tasks often require a name, address or reference number; the format of the field (e.g., a box per letter) dictates how to write it.

The audio may include irrelevant background conversation before the answer; the question stem's key word tells you when to listen closely.

Negative phrasing ('not included', 'except') in the question reverses the expected answer.

In matching tasks, the option letters are usually listed in a box; write the letter, not the full phrase, if instructed.

Common spelling traps include double consonants, silent letters, and British spellings (colour, centre) used consistently.

Section 1 and 2 are usually social/transactional; Section 3 and 4 are academic, with lectures using signposting and definitions.

True/False/Not Given items are answered from the passage only; a statement that is true in the real world but absent from the text is Not Given.

In matching-headings tasks, do the headings for the easiest paragraphs first, then eliminate used headings from the list.

Sentence-completion answers must fit the word limit and the grammar; check the stem's verb form before writing.

For 'choose the correct letter' items, the correct option is a paraphrase; options that copy the text verbatim are often wrong.

Passage 1 is usually easiest; completing it quickly banks time for the harder third passage.

Underlining key terms in the question before reading the passage reduces re-reading and improves scan targeting.

Task 1 maps require change language (was replaced by, was extended, was demolished, remained unchanged) and orientation language (north, adjacent, opposite).

Task 1 must have an overview; a response without one is limited to Band 5 on Task Achievement regardless of detail.

Task 2 two-part questions require both parts answered with roughly equal development; ignoring one part halves the task response.

Lexical Resource penalizes repetition of the prompt's key words; paraphrase them in the introduction and vary them in the body.

Grammatical Range and Accuracy includes punctuation; comma splices and missing apostrophes count as errors.

Word-count check should be done before the final minutes; adding a sentence to reach 250 words is better than submitting 240.

Register in Task 2 is neutral-academic; contractions and slang are inappropriate.

Part 2 preparation time is exactly 1 minute and notes are allowed; the examiner will signal when the long turn begins.

If the long turn ends before 2 minutes, the examiner may prompt with 'Can you tell me more about...?', which signals underdevelopment.

Fluency includes the ability to self-correct without losing the thread; abandoning a sentence and restarting repeatedly lowers the band.

Lexical Resource is judged on appropriacy as well as range; slang and idiom used inaccurately reduce the score.

Grammatical errors that persist across the test (e.g., missing third-person -s) are weighted more than one-off slips.

Pronunciation assessment covers individual sounds, word stress and intonation; consistent mispronunciation of common words affects intelligibility.

Part 3 rewards hypothetical language ('would', 'might', 'if...then') for speculative questions about the future or society.

Answers in note-completion tasks must fit the note's grammar; a noun is usually required after an article.

The audio often gives an example answer before the first question; do not write the example.

In multiple-choice with three options, the distractor may be mentioned first and rejected, so wait for the speaker's final decision.

Listening to the speaker's stress helps identify the key information; emphasized words often carry the answer.

Section 4 is a monologue with no pause in the middle; reading the questions in the break before it starts is essential.

Write answers in the language of the question; mixing languages or scripts is not accepted.

Matching-headings instructions may state that headings can be used once only; check before reusing a heading.

In summary-completion without a word bank, the answer must be taken from the passage, not invented.

The passage's first sentence often states the topic; the last sentence often states the conclusion, which is useful for main-idea items.

Scanning for synonyms of question words is more effective than scanning for the exact words, since passages paraphrase.

Inference answers must be the only reasonable conclusion from the text; if two options are plausible, re-read the specific lines.

Do not spend time translating unknown vocabulary; context and word parts (prefixes, roots) often suffice.

Task 1 introduction paraphrases the prompt (what the chart shows, the time period, the units) without copying its wording.

Task 2 introductions should be short; a three-sentence introduction that previews both body paragraphs wastes time and words.

Body paragraphs need a topic sentence, explanation, and example; a paragraph without a clear controlling idea lowers Coherence and Cohesion.

Linkers should match the logical relationship: 'whereas' for contrast, 'therefore' for result, 'despite' for concession followed by a noun.

Task 1 should use past tense for past data and present tense for general facts (e.g., 'the chart shows').

Task 2 conclusions restate the position and summarize the main reasons in one or two sentences.

Handwriting or typing errors that change meaning are penalized; proofread for dropped words and wrong word forms.

Part 1 questions are short and direct; answering with a reason and an example gives the examiner evidence for all four criteria.

Part 2 cue cards may include a bullet about feelings or reasons; these support higher-band development.

Part 3 may include questions that ask for comparison between past and present, requiring tense contrast.

Fluency is supported by using discourse markers ('to be honest', 'as far as I'm concerned') appropriately, not excessively.

Memorized answers are detectable because they do not fit the question; examiners discount language that is not responsive.

Pronunciation of thought groups (pausing at clause boundaries) aids intelligibility more than individual sound accuracy.

Answers must be written in the space provided; extra words outside the gap are ignored but the gap content is what is marked.

In table-completion tasks, the column heading indicates the answer type (e.g., 'Cost' requires a price, 'Duration' requires a time).

The speaker may give a range ('between 9 and 11'); write the form the question expects, not both numbers unless asked.

Idiomatic expressions ('a stone's throw away') may be paraphrased in the question as 'very close'; match meaning, not words.

Section 3 discussions often include agreement and disagreement; the answer is the point both speakers accept.

Check that the answer is a real word from the language of the test, not a phonetic guess.

For matching-features items, locate each option's name in the passage first, then read the surrounding sentences.

Yes/No/Not Given answers require identifying the writer's claim; if the writer merely reports others' views, the statement may be Not Given.

In sentence-completion, the answer's part of speech must match the gap; a verb gap cannot take a noun.

The passage may contain a definition in parentheses or after a dash; these are useful for vocabulary items.

Reading questions are not equally difficult; answer the easiest types (completion) before the hardest (matching headings).

Always check the instruction word limit before writing; exceeding it invalidates an otherwise correct answer.

Task 1 comparison of two charts requires linking the data (e.g., 'while X rose, Y fell') rather than describing each separately.

Task 2 advantage-disadvantage essays should state whether advantages outweigh disadvantages if the question asks for an opinion.

Coherence and Cohesion rewards clear paragraph function: introduction, body 1, body 2, conclusion, each with a distinct role.

Lexical Resource includes spelling; commonly misspelled academic words (environment, government, necessary) should be checked.

Grammatical Range includes passive voice, which is useful in Task 1 process descriptions and in formal Task 2 statements.

Task 2 must answer the exact question; a memorized essay template that does not address the prompt scores low on Task Response.

Leave two to three minutes to proofread; errors corrected before submission do not count.

Part 2 long turn should be structured: what/who, when/where, details, and why it matters, matching the cue bullets.

Part 3 answers should avoid one-word responses; a minimum of two sentences with a reason is the practical target.

Fluency and Coherence penalizes repetition of the same linker; vary with 'that said', 'on the other hand', 'in that sense'.

Lexical Resource rewards topic-specific vocabulary; abstract Part 3 topics (education, technology, environment) benefit from prepared collocations.

Pronunciation includes stress in multi-syllable words; misplaced stress (e.g., 'deVELopment') reduces intelligibility.

The examiner may ask 'Why?' to prompt elaboration; this is a cue to extend, not a sign of failure.

Speaking scores are based on the recorded performance; speaking too quietly can reduce perceived intelligibility.

In form completion, the answer must match the field label exactly (e.g., a 'surname' field takes only the family name).

The speaker may spell out an email address; '@' is read as 'at' and '.' as 'dot'.

Words like 'approximately' and 'about' signal that the answer is an estimate; write the number given, not a rounded one unless asked.

In matching tasks, the order of the options in the box does not correspond to the order in the audio.

Section 2 often includes a map with a starting point; the speaker's route begins at the marker, not at the top of the map.

If the audio gives two possible answers and one is rejected ('we could meet Monday, but Tuesday suits better'), the accepted one is the answer.

True/False/Not Given statements that use absolute words ('all', 'never', 'only') are often False because the passage qualifies them.

Matching-headings answers should be chosen after reading the whole paragraph, not just the first sentence.

In multiple-choice, the stem's question word (why, how, what) determines which option is relevant.

Completion answers taken from the passage must preserve the passage's word form; do not change a plural to singular.

Time spent on a single hard item should be capped; mark it and return if time allows.

The final passage often has the most complex sentence structures; slower reading there is expected and should be planned for.

Task 1 overview language includes 'overall', 'in general', 'the most striking feature'; these signal the summary rather than data.

Task 2 thesis statements should be specific to the prompt; a generic statement ('this essay will discuss...') does not count as a position.

Coherence and Cohesion includes the use of referencing to avoid repetition; 'this', 'such', 'the former' must have clear antecedents.

Lexical Resource penalizes word-form errors (e.g., 'economical' for 'economic'); check derivations.

Grammatical Range includes complex noun phrases; overuse without accuracy reduces clarity.

Task 1 must not evaluate the data as good or bad; evaluation is an opinion.

Task 2 examples should support the argument, not replace it; a paragraph of examples without explanation loses Coherence.

Part 1 answers should not be rehearsed speeches; natural, responsive answers score higher on Fluency and Coherence.

Part 2 notes should include the cue bullets and one or two personal details to ensure the turn lasts 2 minutes.

Part 3 questions may ask for speculation about the future; use 'will likely', 'may well', 'it's conceivable that'.

Fluency is supported by paraphrasing when a word is forgotten, rather than stopping to search.

Lexical Resource includes idiomatic expressions used appropriately; forced idioms reduce the score.

Pronunciation assessment includes intonation for meaning; flat intonation can reduce the coherence impression.

The examiner's questions in Part 3 build on each other; listening carefully prevents off-topic answers.

Answers in completion tasks should not include articles unless the gap requires them; extra words can exceed the limit.

In multiple-choice, options that are too general or too absolute are often wrong; the correct option matches the speaker's qualification.

The audio may include numbers that are corrected ('£50, no, £15'); the final number is the answer.

Section 4 lectures often define terms; the definition may be the answer to a completion item.

Check that the answer fits the question's grammar before moving on; a grammatically wrong answer is usually wrong.

Listening answers are marked on meaning and form; a synonym is not accepted if the question requires the exact word from the audio.

For 'which paragraph contains' items, the answer may be a paragraph that mentions the idea briefly, not the main topic.

Yes/No/Not Given items require distinguishing the writer's opinion from reported facts; reported views may be Not Given for the writer's view.

In matching sentence endings, eliminate endings that are grammatically incompatible before checking meaning.

Summary-completion with a word bank may require changing the word form to fit the gap; check the instruction.

Reading passages often use hedging ('may', 'suggests', 'tends to'); absolute statements in questions are often False.

The instruction 'Choose the correct letter' means one answer only; writing two letters is marked wrong.

Task 1 process diagrams should state the number of stages and the start and end points in the overview.

Task 2 discussion essays should give the writer's opinion in the introduction or conclusion, and it must be consistent.

Coherence and Cohesion rewards logical paragraph order; chronological or problem-solution order should be evident.

Lexical Resource includes collocations with prepositions (e.g., 'a rise in', 'an increase of'); errors here are common.

Grammatical Range includes conditional sentences for hypothetical arguments in Task 2.

Task 1 should not include personal experience or reasons; only the data shown.

Task 2 should avoid rhetorical questions, which are considered informal in academic writing.

Part 2 cue cards may ask about a person; use present tense for description and past tense for a shared experience.

Part 3 answers should acknowledge complexity ('it depends on the context') before giving a view.

Fluency and Coherence includes the ability to link ideas across turns, not just within a sentence.

Lexical Resource includes precise verbs (e.g., 'mitigate', 'foster', 'undermine') used accurately.

Pronunciation includes pausing at meaningful boundaries; pausing mid-phrase reduces intelligibility.

The examiner does not give feedback; candidates should not interpret neutral reactions as negative.

Speaking test order is fixed: Part 1, Part 2, Part 3; the examiner manages timing throughout.

In note-completion, the answer is usually a single noun or number; writing a phrase often exceeds the limit.

The audio may include a false start ('I think it's on... no, wait'); the completed statement is the answer.

In matching tasks, the same option may be correct for more than one item if the instructions allow reuse.

Section 1 often includes a booking or enquiry; the answer is the detail the caller confirms.

Listening for the speaker's purpose (complaint, request, recommendation) helps predict answer type.

Check spelling of names against common patterns; unusual spellings are often spelled aloud.

For matching-headings, headings that use the paragraph's exact words but change the meaning are distractors.

In completion tasks, the answer must be a word or phrase from the passage; paraphrasing is not allowed unless the instruction permits it.

Multiple-choice questions may ask about the writer's purpose; the answer is the function of the sentence, not its content.

Scanning for capitalized words locates names and places quickly in dense passages.

Reading speed should slow for the answer zone and speed up between zones.

Unanswered items should be guessed; there is no negative marking.

Task 1 line graphs with multiple lines require comparing trends, not describing each line separately.

Task 2 problem-solution essays should give one or two problems and matching solutions, developed with explanation.

Coherence and Cohesion includes the use of pronouns; ambiguous pronouns reduce clarity.

Lexical Resource rewards precise quantification ('a significant minority', 'a marginal increase') over vague words ('many', 'big').

Grammatical Range includes varied sentence openings; starting every sentence with 'The' or 'This' reduces range.

Task 1 must use the units given on the axis; misreading units is a common accuracy error.

Task 2 must not exceed the time budget; an unfinished essay loses more than a slightly shorter one.

Part 1 questions may ask about preferences; answer with a preference, a reason, and a contrast ('but I also enjoy...').

Part 2 should cover the cue card's prompts in order to ensure coverage and natural progression.

Part 3 answers should use examples from society or personal observation to support abstract claims.

Fluency is supported by using fillers appropriately and avoiding long silent pauses.

Lexical Resource includes paraphrasing the examiner's question to avoid repetition.

Pronunciation includes accurate word stress in academic vocabulary (e.g., 'photography', 'economic').

Speaking performance is assessed on the whole test; a weak Part 1 can be offset by strong Parts 2 and 3.

In table completion, the answer must fit the row and column; reading across the row gives context.

The speaker may use 'roughly', 'approximately' or 'just over'; the answer is the number stated, not a rounded one.

In multiple-choice, the correct option may be expressed with a synonym of the audio's phrase.

Section 3 often includes a tutor's advice; the answer may be the tutor's recommendation, not the student's idea.

Check that the answer is not a word already in the question; repeated words are usually not the answer.

Listening answers are usually concrete; abstract nouns are less common in completion items.

For True/False/Not Given, a statement that contradicts the passage is False, not Not Given.

Matching-headings tasks should be answered after skimming all paragraphs to understand the passage's structure.

In sentence-completion, the answer's grammar must fit the stem; check articles and prepositions.

Multiple-choice options that are true but do not answer the question are distractors.

Reading passages often contain examples that illustrate a general claim; questions may ask for the general claim.

Time allocation should prioritize completion and short-answer items, which are faster than matching headings.

Task 1 bar charts with categories require ordering by size and grouping similar categories.

Task 2 opinion essays should address counterarguments to show depth, but the writer's position must remain clear.

Coherence and Cohesion includes the use of topic sentences that preview the paragraph's content.

Lexical Resource includes avoiding repetition of the same verb; vary with synonyms that fit the context.

Grammatical Range includes accurate use of articles; this is a high-frequency error area.

Task 1 must not include data not shown in the chart; invented figures lower Task Achievement.

Task 2 should avoid absolute claims ('everyone', 'never') unless justified; hedging is more academic.

Part 1 answers should be about 2-3 sentences; longer answers may be interrupted by the examiner.

Part 2 cue cards may ask about a future plan; use future forms and reasons.

Part 3 answers should compare and contrast, using 'whereas', 'unlike', 'in contrast'.

Fluency and Coherence includes the ability to develop an argument across several sentences.

Lexical Resource includes using synonyms to avoid repeating the question's words.

Pronunciation includes clear articulation of consonant clusters, which affect intelligibility.

The examiner may ask for clarification; answering the clarified question is expected.

In form completion, the answer must match the field's expected format (e.g., a date field takes a date, not a day).

The audio may include a list of options; the answer is the one the speaker selects, not the first mentioned.

In matching tasks, the options may be paraphrased in the audio; match meaning, not words.

Section 2 often includes a talk about a place or event; the answer is the detail the speaker emphasizes.

Check that the answer is a real word; phonetic guesses are often wrong.

Listening answers should be written clearly; ambiguous handwriting is marked wrong.

For matching-features, the same feature may be used more than once if the instructions allow it.

In Yes/No/Not Given, the writer's view must be explicit; implied views may be Not Given.

Completion answers must be copied exactly; changing word forms is not allowed unless the instruction permits it.

Multiple-choice questions may ask about the main idea; the correct option summarizes the passage, not a detail.

Scanning for numbers and dates is faster than reading the whole passage.

Difficult vocabulary can often be inferred from context; do not stop to look up words.

Task 1 pie charts require comparing proportions, not raw numbers.

Task 2 advantage-disadvantage essays should give both sides and then a judgment if asked.

Coherence and Cohesion rewards clear paragraph breaks; a single block of text loses marks.

Lexical Resource includes accurate word forms; check noun/verb/adjective forms.

Grammatical Range includes varied sentence structures; simple sentences alone limit the score.

Task 1 should not include causes or reasons for the data.

Task 2 should answer all parts of the question; missing a part limits Task Response.

Part 1 questions may ask about habits; use frequency adverbs and present simple.

Part 2 should include details that make the description vivid (sights, sounds, feelings).

Part 3 answers should use conditional structures to discuss hypothetical situations.

Fluency is supported by maintaining eye contact and natural body language in face-to-face delivery.

Lexical Resource includes using precise adjectives rather than 'good' or 'bad'.

Pronunciation includes stress for emphasis; emphasizing key words aids clarity.

The examiner's questions in Part 3 may be challenging; taking a moment to think is acceptable.

In note-completion, the answer is usually a key detail, not a general category.

The audio may include a correction ('not the 15th, the 16th'); the corrected detail is the answer.

In multiple-choice, the correct option may be a summary of the speaker's point, not a direct quote.

Section 4 lectures often include examples; the answer may be the example, not the general statement.

Check that the answer fits the question's word limit before writing.

Listening answers are marked on accuracy; a near-correct spelling may be marked wrong.

For matching-headings, choose the heading that covers the whole paragraph, not one sentence.

In True/False/Not Given, a statement that is partially true but not fully supported is False or Not Given depending on the contradiction.

Completion answers must fit the gap grammatically; check the surrounding words.

Multiple-choice options that use absolute words are often wrong.

Scanning for synonyms of question words is effective because passages paraphrase.

Reading time should be monitored; the third passage often takes the longest.

Task 1 tables require grouping data by trend or size, not listing every row.

Task 2 discussion essays should give both views and the writer's opinion.

Coherence and Cohesion includes logical sequencing of ideas; each paragraph should follow from the previous.

Lexical Resource includes avoiding repetition of the prompt's words; paraphrase them.

Grammatical Range includes accurate use of complex sentences; errors reduce the score.

Task 1 must include an overview; without one, Task Achievement is limited.

Task 2 should avoid informal language; use academic register.

Part 1 answers should include a reason to extend the response.

Part 2 should cover the cue card's prompts and add a concluding comment.

Part 3 answers should use examples to support abstract ideas.

Fluency is supported by using discourse markers to organize ideas.

Lexical Resource includes using collocations accurately.

Pronunciation includes clear word stress and intonation.

The examiner's role is to manage the test; candidates should not ask for feedback.

In form completion, the answer must match the field's label; a 'postcode' field takes a postcode, not a city.

The audio may include a list of options; the answer is the one the speaker chooses.

In matching tasks, the options may be paraphrased; match meaning.

Section 2 often includes a talk about a place; the answer is a detail the speaker emphasizes.

Check spelling of names and places; unusual spellings are often spelled aloud.

Listening answers should be written clearly and within the word limit.

For matching-features, locate each option's name in the passage first.

In Yes/No/Not Given, the writer's view must be explicit; implied views may be Not Given.

Completion answers must be copied exactly; changing word forms is not allowed unless permitted.

Multiple-choice questions may ask about the main idea; the correct option summarizes the passage.

Scanning for numbers and dates is faster than reading the whole passage.

Difficult vocabulary can often be inferred from context.

Task 1 pie charts require comparing proportions, not raw numbers.

Task 2 advantage-disadvantage essays should give both sides and a judgment if asked.

Coherence and Cohesion rewards clear paragraph breaks.

Lexical Resource includes accurate word forms.

Grammatical Range includes varied sentence structures.

Task 1 should not include causes or reasons for the data.

Task 2 should answer all parts of the question.

Part 1 questions may ask about habits; use frequency adverbs and present simple.

Part 2 should include details that make the description vivid.

Part 3 answers should use conditional structures for hypothetical situations.

Fluency is supported by natural body language in face-to-face delivery.

Lexical Resource includes using precise adjectives rather than 'good' or 'bad'.

Pronunciation includes stress for emphasis.

The examiner's questions in Part 3 may be challenging; taking a moment to think is acceptable.

In note-completion, the answer is usually a key detail, not a general category.

The audio may include a correction; the corrected detail is the answer.

In multiple-choice, the correct option may be a summary of the speaker's point.

Section 4 lectures often include examples; the answer may be the example.

Check that the answer fits the question's word limit.

Listening answers are marked on accuracy; a near-correct spelling may be marked wrong.

For matching-headings, choose the heading that covers the whole paragraph.

In True/False/Not Given, a statement that is partially true but not fully supported is False or Not Given depending on the contradiction.

Completion answers must fit the gap grammatically.

Multiple-choice options that use absolute words are often wrong.

Scanning for synonyms of question words is effective.

Reading time should be monitored; the third passage often takes the longest.

Task 1 tables require grouping data by trend or size.

Task 2 discussion essays should give both views and the writer's opinion.

Coherence and Cohesion includes logical sequencing of ideas.

Lexical Resource includes avoiding repetition of the prompt's words.

Grammatical Range includes accurate use of complex sentences.

Task 1 must include an overview.

Task 2 should avoid informal language.

Part 1 answers should include a reason to extend the response.

Part 2 should cover the cue card's prompts and add a concluding comment.

Part 3 answers should use examples to support abstract ideas.

Fluency is supported by using discourse markers to organize ideas.

Lexical Resource includes using collocations accurately.

Pronunciation includes clear word stress and intonation.

The examiner's role is to manage the test; candidates should not ask for feedback.

In form completion, the answer must match the field's label.

The audio may include a list of options; the answer is the one the speaker chooses.

In matching tasks, the options may be paraphrased; match meaning.

Section 2 often includes a talk about a place; the answer is a detail the speaker emphasizes.

Check spelling of names and places.

Listening answers should be written clearly and within the word limit.

For matching-features, locate each option's name in the passage first.

In Yes/No/Not Given, the writer's view must be explicit.

Completion answers must be copied exactly.

Multiple-choice questions may ask about the main idea.

Scanning for numbers and dates is faster than reading the whole passage.

Difficult vocabulary can often be inferred from context.

Task 1 pie charts require comparing proportions.

Task 2 advantage-disadvantage essays should give both sides.

Coherence and Cohesion rewards clear paragraph breaks.

Lexical Resource includes accurate word forms.

Grammatical Range includes varied sentence structures.

Task 1 should not include causes or reasons for the data.

Task 2 should answer all parts of the question.

Part 1 questions may ask about habits; use frequency adverbs.

Part 2 should include details that make the description vivid.

Part 3 answers should use conditional structures.

Fluency is supported by natural body language.

Lexical Resource includes using precise adjectives.

Pronunciation includes stress for emphasis.

The examiner's questions in Part 3 may be challenging.