Tricks revision

One key word or phrase is swapped for a synonym, while the rest of the sentence keeps almost the same shape. This is the lightest kind of paraphrase — only the vocabulary changes, the grammar barely moves.

⚠️ But note: this clean, same-structure swap is actually the rarer case. Far more often, IELTS rewrites the whole sentence — that is Trick 2, and that is the frequent style. Learn this one first because it is the easiest to see, then expect Trick 2's fuller paraphrase far more often.

Line the two sentences up side by side. If the order of ideas matches and only the words differ, it is a synonym swap — trust the meaning and mark it. If the whole structure has moved, switch to Trick 2.

Sometimes the whole sentence is rebuilt — active becomes passive, or the clause order flips — but the underlying facts stay identical. Watch out: sometimes a "restructured" option secretly changes the meaning too — that's a trap.

Do not compare word-for-word. Ask one question: are the underlying facts the same, and does the meaning point the same way? If yes, it is a valid restructure. If the direction has flipped (as in the Try it below), it is the trap.

Correct options are usually paraphrased, so an option that repeats the passage's exact words is not automatically right. Test-writers often reuse a striking phrase from the passage in a distractor and then change what is said about it — the wrong person, the wrong time, the opposite direction. The same applies to numbers: a figure like 75% or 1995 that appears in both the passage and an option proves only that you have found the right area — you still have to check that the option says the same thing about it.

When an option matches the passage word-for-word or number-for-number, treat it as a warning light, not a green light: check the other options first, and confirm the exact-match option still says the same thing the passage does before you trust it.

A single adverb — hardly, fully, often, always, sometimes, rarely, previously, nearly — can completely reverse a sentence's meaning, even when every other word is identical. Two sentences can read almost the same and mean opposite things, so an option that looks like a match is not automatically the answer.

Before you accept a look-alike option, put its adverbs against the passage's. If a degree, frequency, or time adverb has changed — partly↔fully, rarely↔frequently, sometimes↔always, previously↔now — the meaning has changed with it. Same shape, opposite sense = trap.

Words like however, in contrast, therefore, despite, and as a result tell you what KIND of information is about to follow — a contrast, a cause, a result, a concession. They help you find the part of a paragraph that answers the question, which you then read carefully.

Reading has 60 minutes for three passages and no extra transfer time, so you cannot read everything with the same slow care. Signal words show where the writer turns, and many questions — especially on the writer's view — are decided right after a however, yet or even so.

In Multiple Choice, more than one option can be true according to the passage — but only ONE actually answers the exact question asked. Before choosing, don't ask "is this true?"; ask "does this answer this question?"

Underline the exact focus of the question stem (why / when / who / what effect). Then test each option against that focus, not against the passage in general. Discard any option that is true but does not address the stem.

TRUE/YES = the information in the passage confirms the statement. FALSE/NO = the information in the passage contradicts it — the passage and the statement cannot both be true. NOT GIVEN = the passage gives no information that lets you confirm or contradict this specific claim. Be careful: Not Given does not mean "never mentioned". The topic — even the same words — may well appear in the passage; what is missing is information about the exact claim the statement makes. It might be true in real life, but you cannot know that from this text.

Candidates lose many marks in this task family, and most mistakes come from skipping this three-way test — especially from marking Not Given as soon as a word seems "missing", or False as soon as something "sounds unlikely".

Every statement makes one core claim (usually carried by its main verb + noun, plus any qualifier such as first, only, more than, because). Use the topic words to locate the right part of the passage — then read that part carefully for meaning, because it will be paraphrased. Three outcomes decide the answer:

1) the passage confirms the claim (in any wording) → True
2) the passage says something that cannot be true at the same time as the claim → False
3) the passage discusses the topic but neither confirms nor rules out this claim → Not Given

"False" needs evidence: a sentence you can point to that clashes with the statement. If all you have is "it seems unlikely" or "it is probably true in real life", you have no evidence — that is the signal for Not Given.

Underline the statement's exact claim. Locate the matching part of the passage and read the whole sentence (and the one around it) for meaning. Then ask two questions in order: "Which words confirm it?" → True. "Which words make it impossible?" → False. If you can answer neither with words from the text → Not Given.

Compare the strength of the statement with the strength of the passage. A statement using "all," "every," "always," "only" or "never" is False when the passage shows an exception or uses a limiting word that excludes the rest ("most … though a few did not", "usually", "in some cases"). But check the other two outcomes too: if the passage itself is absolute ("without exception"), the statement can be True; and if the passage simply says "many" or "some" without telling you about the rest, an "all" statement may be Not Given.

Circle the absolute word in the statement, then find the passage's matching word. If the passage limits the claim (most … but not all / usually / with a few exceptions) where the statement says all / always, that clash makes it False. If the passage is equally absolute, it is True. If the passage only tells you about some and is silent about the rest, ask whether anything rules "all" out — if nothing does, it is Not Given.

A statement that gets most details right but changes just ONE — a number, a name, a date, a cause, who did what — is False, not True, provided the passage gives that detail differently. Read every detail, not just the general shape of the sentence. If the passage does not give that detail at all (it gives the duration but never the number of companies, say), the extra detail is unconfirmed → Not Given.

This is a deliberately sneaky trap — the statement reads so similarly to the passage that skimming students mark it True without checking the one changed detail.

In this task you are given the first half of several sentences (the stems) and a longer list of possible endings. The correct ending must pass two tests at once:
1) Meaning — it must say what the passage actually says.
2) Grammar — it must join onto the stem to make one correct sentence.
An ending can pass one test and fail the other — and there are always more endings than stems, so some endings are pure distractors that fit nothing.

Read the stem, find the matching sentence in the passage, and decide what idea the ending must carry. Only then look at the endings list — first cross out any whose meaning is wrong, then, among what's left, pick the one that also joins grammatically onto the stem.

Before matching a heading, ask what the paragraph is doing — introducing, explaining, comparing, criticising, concluding — not just what topic it's about.

A paragraph usually announces its job in the first sentence or two. Read those, underline the signal words, and the job names itself. Three short samples — the underlined words are what give it away:

1. "For example, many primary schools in Japan teach every pupil to play the recorder; similarly, a number of Hungarian schools hold a singing lesson every day." → Illustrating

2. "However, critics argue that the studies linking music lessons to better maths results drew on too few schools to trust." → Critiquing

3. "In short, the schools that added daily music lessons saw measurably better attendance across the decade." → Concluding

Match the heading to that job — "Examples of the trend", "Doubts about the evidence", "The overall verdict" — not just to the shared topic (music lessons).

A heading that matches only a small detail inside a paragraph is usually a distractor — the correct heading summarises the main idea of the whole paragraph, which every sentence supports, not one sentence in it.

This is one of the most common Matching Headings traps — the list often includes a heading that is true of one sentence (an example, a date, a single statistic) to catch candidates who match the first familiar idea they see.

Don't summarise every paragraph. First decide what type of information each statement describes — a figure, a reason, an example, a comparison, a criticism, a recommendation. Then scan the whole passage for that type of content (numbers, names and dates are quick anchors), and read the matching sentence carefully to confirm it really says what the statement says. The statement will be paraphrased, so the exact words are rarely there.

Matching Information answers are not in passage order, some paragraphs may contain no answer, and if the instructions say NB You may use any letter more than once, one paragraph can hold two answers. Candidates lose time by re-reading every paragraph for gist — and lose marks by stopping at the first paragraph that shares a word or a number with the statement.

Quick method first: when you're short on time or unsure, reading the first one or two sentences plus the last sentence of a paragraph often reveals its main idea — a fast, reliable first pass for Matching Headings.
But the caution: don't rely on the first sentence alone — sometimes it's only a transition or a single example, and the real topic appears later. If the first and last sentences point in different directions, read the middle before you commit.

Students who only skim first sentences for Matching Headings get tricked whenever a paragraph doesn't follow simple topic-sentence-first structure — which is common in real academic writing.

When multiple named things share similar traits, look for language that ties a specific claim to a specific name — don't match by idea alone, match by exactly whose idea it is.

Test-writers often create features that could plausibly apply to more than one source — precision about WHO said or did something is what actually separates the options. The statements are paraphrased and not in passage order, and an instruction such as NB You may use any letter more than once means one source can match several statements while another matches none.

A heading using the passage's exact vocabulary can still be the wrong answer if it doesn't match the paragraph's overall topic — familiar words create false confidence.

This is the Matching Headings version of the Exact-Word Trap from Chapter 2 (Multiple Choice & Paraphrase), Trick 3 — recognising it here stops you from picking a heading just because it "sounds right."

Take the answer word(s) exactly as they appear in the passage, and respect the word limit strictly — "NO MORE THAN TWO WORDS" means one or two; three words is wrong even with perfect spelling and correct meaning. Read the instruction every time: "ONE WORD ONLY", "NO MORE THAN TWO WORDS AND/OR A NUMBER" and so on. A number can be written in figures, and a hyphenated word (e.g. "high-pressure") counts as one word. (When a summary gives you a box of words to choose from, you pick from the box instead — read the instruction.)

You still need paraphrase to find the answer — the sentence around the gap is reworded — but the answer itself is marked strictly: the right word(s), correctly spelled, within the limit. Your own synonym, a misspelling or an extra word all score zero.

The blank's surrounding grammar tells you what KIND of word fits before you even find it — after "a/an" expect a noun, after "to" expect a verb, a plural verb needs a plural noun.

Knowing what you're looking for narrows the scan dramatically — you can rule out verbs and adjectives instantly if the blank needs a noun.

Many completion answers are nouns, and numbers, dates and named things are common. If the gap clearly needs a number or a name ("___ participants", "in the year ___"), scan for figures or capital letters to find the right area quickly — then read the sentence to choose the right one. Passages often give several numbers close together (a target and an actual figure, an old and a new value), and only one matches the meaning of the question.

Scanning for the right type of item narrows your search, but it only locates. The distractor is usually a second number or name in the same sentence, so the final choice always depends on understanding.

The question sentence is written so that a word exactly as printed in the passage fits the gap. If you feel you need to change the form — "purify" into "purification", "rapid" into "rapidly" — stop: either the passage uses that form somewhere else, or you have misread the gap's grammar, or you are in the wrong place. Don't invent a new word form yourself.

Answers must be words from the passage — a self-made "corrected" form is a different word, and it is marked wrong even if it is grammatically reasonable.

For flow-chart completion, the blanks usually follow the order in which the passage describes the process — use that sequence to narrow down WHERE in the passage to search, instead of scanning randomly. The boxes are paraphrased summaries of each stage, so match the meaning of each step, and copy the answer word in the form the passage prints it. (A true labelled-picture diagram uses a different trick — Chapter 6 (Table Completion & Diagram Labelling) covers that separately.)

Process-description passages are usually written in the same order as the flow-chart — this can save minutes of searching. But a passage sometimes mentions a stage early (for example in an overview sentence) and explains it later, so confirm the meaning before you copy.

Short-answer questions often ask for a specific count or list ("what TWO things...") — your answer must match that exact number and type, not just be "an" answer.

A common, entirely avoidable error is giving only one item when two are asked, or giving a description instead of the specific named things requested.

A table answer must fit the row you're told AND the column header's category — check both before writing. A completely true fact from the passage is still wrong if it belongs in a different cell.

This is the single most common table-completion error — copying a real, correctly-spelled fact from the passage into the wrong cell, because only one of the two coordinates (row or column) was actually checked.

Just like grammar signals the answer type in sentence completion (Chapter 5, Trick 2), a table's own column heading tells you what KIND of information belongs in every cell beneath it — a "Cost" column expects a price or cost-word, a "Duration" column expects a time expression, before you've even found the passage's exact wording.

Reading the column heading first narrows your search — you can rule out any passage detail that isn't the right KIND of fact. You still have to find and read the exact sentence: the heading tells you what to look for, not what the answer is.

A labelled diagram (a picture of a device or structure) is described mainly through position and connection language: above, beneath, attached to, fitted near, connects to, at the base of. Use those words as your map to each blank. When the diagram shows something moving through a system (water, air, heat), the description may also follow that flow — but each label is still fixed by where the part is, so check its position, not just the order.

Treating a picture-based diagram like a flow-chart (hunting only for step-order words) leads to labels in the wrong place, because the passage is mostly describing a fixed structure. Diagram labels also do not always follow passage order, so match each numbered pointer by position.

Unlike sentence completion, a diagram label has no room for a full clause — it's almost always just the name of a component (a noun or short noun phrase), copied from the passage with the same spelling and the same form (singular or plural). Scan for concrete object-names, not actions or descriptions, and keep within the word limit.

Knowing the answer will be a short noun phrase stops you from writing an explanation or a full sentence where only a 1–3 word label is expected — a very common, entirely avoidable mark loss.

Rewrite the question's own sentence in your introduction using synonyms and a different sentence structure — change the key noun and verb, but keep the meaning and add no data.

Words copied from the question are not counted towards your 150 words, and they give the examiner no evidence of your own vocabulary for Lexical Resource. A paraphrased opening shows control from the first line. Keep it to one sentence — the introduction earns little on its own, so the overview and the selected details matter far more.

Before writing anything else, look at the data as a whole shape, not individual numbers, and ask: what are the one or two most obvious things anyone would notice at a glance? Those become the Overview — no specific numbers yet.

The public Task Achievement descriptors are explicit: Band 7 needs "a clear overview of main trends, differences or stages", while an answer with no clear overview fits Band 5 — however accurate the numbers are. The overview may come straight after the introduction or at the end, but it must be easy to find, so give it its own short paragraph.

Never describe data in the order it's given — row by row, year by year. Instead, group data points that share a pattern (two similar countries, or a before-vs-after split) and describe the group together in one sentence.

The Task Achievement descriptors describe Band 5 as recounting detail "mechanically". Listing number after number shows you have not selected or compared anything, and it also weakens Coherence and Cohesion, because the paragraphs have no logical organisation — even if every number is accurate.

Comparison language isn't interchangeable. "More than double," "nearly triple," and "roughly half" each describe a specific mathematical relationship — pick the phrase that matches the actual ratio, not just a general sense of "a lot more" or "a lot less."

Examiners check for accuracy, not just vocabulary variety — using "double" for something that's actually triple is a factual error that directly costs Task Achievement marks, no matter how fluent the sentence sounds.

Every body-paragraph sentence should make a comparative point first, then use the number as evidence for it — not state a bare number with no framing.

Band 7 Task Achievement asks you to "clearly present and highlight key features". A point-first sentence does exactly that, and it naturally produces the comparison language and grouping that Tricks 3 and 4 depend on. The numbers still matter: key features must be supported with accurate data.

Modifying a movement verb or noun with an intensity word — an adverb before the verb ("rose sharply") or an adjective before the noun ("a sharp rise") — must match how steep the actual change is on the chart. Use a three-tier scale: dramatic, moderate, or gentle.

This earns marks in two criteria at once: Task Achievement (a mismatched intensity word is a factual misrepresentation of the data) and Lexical Resource (varied, precise vocabulary scores higher than repeating "increased" five times).

Before describing any change, identify the 2–3 fixed features that appear unchanged in the SAME position on both maps — a river, a main road, a landmark. These become your reference points for locating everything else ("to the north of the river," "next to the school").

Without fixed reference points, location language becomes vague ("something changed near the middle") — Task Achievement rewards key features that are reported accurately, and on a map a change is only accurate if the reader can tell where it happened. Fixed reference points are what make precise location possible.

Describe the earlier map's features in ordinary past simple ("A park was situated in the north"), then switch to passive past for what happened to them ("...it was converted into a shopping mall"). Passive voice puts the FEATURE first, not an unnamed "someone" who changed it.

One more case — a map that changes up to NOW. If the second map shows the present (labelled now / present day / today, not a past year), report the change in the present perfect passive: "The school that was located near the theatre has now been replaced with a garage." ⛔ Use the present perfect only when the newer map is the present situation. If both maps are in the past (e.g. 1990 → 2010), stay in past simple + passive past — do not use the present perfect.

Answers written in the active voice often invent an agent the map never shows ("the council built a mall"). The passive keeps you accurate to what the diagram actually shows — WHAT changed, not who changed it — and a correct mix of tenses and passives also counts towards Grammatical Range and Accuracy.

Group your description by geographic zone (north side / south side, or "near the river" vs. "further inland") rather than jumping between unrelated changes in random order — describe everything in one area fully before moving to the next.

Coherence and Cohesion rewards information that is logically organised with clear progression. An answer that hops between corners of the map loses marks under this criterion, even if every individual change is accurately described.

Before writing the overview, step back from the individual buildings and ask two questions: (1) what is the overall character of the change — did the place become more built-up, more residential, more commercial, greener, better connected? (2) what major features were left untouched? Answer both in one or two sentences, with no single building, position or date-by-date detail. The individual changes belong in the body paragraphs.

Task Achievement at Band 7 needs "a clear overview of main trends, differences or stages". On a map, the most common failures are an overview that simply lists every change (which is detail, not a trend) or no overview at all. Saying what stayed the same is part of the big picture too — it shows the examiner you have compared the whole map, not just spotted the new buildings.

Describe positions as if you were standing in the real place, not looking at a sheet of paper. Use compass zones (in the north-east of the town, in the south-west) and prepositions anchored to features (just north of the school, on the eastern bank, between the river and Mill Road, beside Mill Road, across the river from). Avoid "at the top of the map", "on the left side" or "in the corner of the picture" — they describe the page, not the place. Watch one small but important difference: in the north of the town = inside it; to the north of the town = outside it.

On a map, a change is only reported accurately if the reader knows where it happened — that is Task Achievement. Precise location vocabulary is also exactly what Lexical Resource rewards in this task type, while "top/bottom/left" reads as imprecise, picture-describing language, and a wrong in/to can move a building out of the town altogether.

A man-made process diagram shows WHAT happens to a material, not WHO does it — so the passive ("the paper is collected," "the pulp is then heated") is the natural, accurate way to describe most stages, not just a style option. Natural processes and life cycles are different: verbs such as evaporate, hatch, grow, flow take no object and stay active ("the eggs hatch", not "the eggs are hatched"). Many diagrams need both.

Inventing an agent the diagram never shows ("workers collect the paper") adds information that isn't there, which weakens Task Achievement. The passive keeps the description accurate to what is shown, and using active and passive forms correctly side by side is also evidence of range under Grammatical Range and Accuracy.

Make the order of the stages clear, and vary how you mark it: connectors (first/initially, once this is complete, following this, subsequently, finally), time clauses ("after being sorted, …") and relative clauses ("…a vat, where it is mixed with water"). Don't just list stages with commas or repeat "then" for every stage — but don't force a connector into every sentence either.

Coherence and Cohesion rewards cohesive devices that are used appropriately; missing, repetitive or mechanical linking ("Firstly… Secondly… Thirdly…") is typical of Band 5–6. Sequencing language is what turns a list of stages into a described PROCESS, not just a set of facts.

Some processes are a one-way flow (raw material → finished product, ends) — others are a cycle that returns to the start (the water cycle, a recycling loop). State which it is in your overview, and make your final sentence match: a linear process ends with the finished product ("…emerging as new paper"), while a cyclical one returns to the start ("…at which point the cycle begins again"). Task 1 does not need a separate conclusion paragraph.

Calling a cyclical process linear, or vice versa, misreports a key feature of the diagram, which affects Task Achievement — it is not just a style issue. Check the diagram's own arrows before you write the overview and the final sentence.

Before writing the overview, count the stages (use the diagram's own numbers if it has them) and find the first and the last. Your overview then says: how many stages, where the process starts and where it ends, whether it is linear or cyclical (Trick 3), and — in a second sentence if you like — the two or three broad phases it falls into. Keep every temperature, material and by-product out of the overview; they belong in the body.

Task Achievement at Band 7 needs "a clear overview of main trends, differences or stages". In process answers the usual mistakes are jumping straight into stage 1 with no overview at all, or squeezing every stage into the overview so that it becomes a list. Counting the stages and naming the phases gives you a real summary in two sentences.

Diagram labels are usually nouns ("Collection", "De-inking") — turn them into verbs ("the paper is collected", "the ink is removed") and vary the forms so you are not copying the labels in order. Every sub-label is information you should report: temperatures, times, materials added and anything removed. But report only what is shown — no people, sizes, reasons or results that the diagram does not give. Technical terms on the diagram (a pulper, molasses) may be used as they are; paraphrase the everyday words around them.

Task Achievement rewards key features that are accurately reported — leaving out the 90°C or the removed ink loses detail, while adding "by skilled workers" or "to make it whiter" adds information the examiner can see is not there. Changing word forms accurately (collection → collected, separating → separate) is also direct evidence for Lexical Resource and Grammatical Range and Accuracy.

Before planning anything, identify three things in the prompt: the TASK VERB PHRASE (what you're actually asked to do — this also tells you the essay type: "to what extent do you agree" = Opinion; "discuss both views" = Discuss-both; "causes...and solutions" = Problem-Solution; "advantages and disadvantages" = Adv-Dis; two separate direct questions, e.g. "why is this happening? is it a positive development?" = Two-part), the TOPIC NOUNS (the exact subject), and any ADJECTIVES/ADVERBS that narrow the scope (e.g. "increasingly," "solely," or a specific group like "young people," not everyone).

Essays that drift into a generic version of the topic, or use the wrong structure for the task verb, lose Task Response marks even with flawless English — the prompt's own words are your checklist for staying exactly on-topic.

Before writing a single sentence, jot a brief skeleton: your position (if Opinion/Discuss-Both) plus one line per body paragraph naming its main point. This takes 2 minutes and prevents running out of ideas mid-paragraph or drifting off-topic.

A rushed start without a plan is the most common reason a student's Body 2 paragraph is noticeably weaker or shorter than Body 1 — planning both paragraphs' content BEFORE writing either one prevents this imbalance.

Body 1 and Body 2 don't contain the same KIND of content for every essay type. Opinion needs two reasons (or a reason plus a rebuttal) for your view. Discuss-Both needs a View-A paragraph, then a View-B paragraph. Problem-Solution needs a causes paragraph, then a solutions paragraph. Advantages-Disadvantages needs an advantages paragraph, then a disadvantages paragraph. Cause and Effect needs a causes paragraph, then an effects/consequences paragraph — easy to confuse with Problem-Solution, but no solution is ever asked for. Two-part (direct question) needs one body paragraph answering each of the prompt's two direct questions — for example, one paragraph on why a trend is happening and a second on whether it is a positive or negative development.

Using an Opinion-essay structure (two reasons for your view) when asked for Problem-Solution (causes then solutions) answers a different question from the one set — the main parts of the prompt are left unaddressed, which is a Task Response problem, not a minor structural preference.

The introduction is usually two (sometimes three) sentences: one paraphrasing the topic in your own words (not copied — words copied from the prompt earn no credit), and one directly stating your position (Opinion / Discuss-both / "outweigh" questions) or previewing your main points (Problem-Solution / Cause-Effect / Two-part). The examiner should know your stance from the introduction alone.

An introduction with no clear position forces the examiner to guess your stance until the conclusion — this reads as unclear, poorly organised writing even when the body paragraphs are strong.

If you can't think of a reason or example, run through a mental checklist of common angles: economic, social/personal, environmental, educational, technological, health. One of these almost always produces a usable idea.

Examiners don't reward ORIGINALITY for its own sake — Task Response assesses how relevant your ideas are and how well you extend and support them. A simple, common idea explained clearly outscores a vague, ambitious one you can't fully develop.

Each body paragraph makes exactly ONE main point, then explains why it's true, then gives a specific example or consequence. Never cram two unrelated ideas into one paragraph, and never state a point without explaining it.

Two underdeveloped points usually score lower than one point fully explained — the Task Response descriptors reward main ideas that are "extended and supported", not the number of points raised, and Coherence and Cohesion rewards paragraphs with a clear central topic and logical progression.

A good example should be about the SPECIFIC topic and scope identified in Chapter 10, Trick 1 (the topic nouns and qualifiers — e.g. young people, not people in general), not just any generic supporting detail. Keep the prompt's ideas, but you don't have to repeat its exact words — rewording them with accurate synonyms also shows lexical range. This proves you're answering THIS question, not a rehearsed generic essay.

Examiners can tell when an example is a memorised template loosely bolted onto the topic — an example tied to the prompt's exact topic and qualifiers signals genuine, on-topic development, while copying the prompt's words again and again adds nothing to Lexical Resource.

For Opinion, Discuss-both-give-opinion and "Do the advantages outweigh…?" essays, your position must be clear and consistent from introduction to conclusion — not "there are good points on both sides" with no actual stance taken. A qualified position ("I agree in most cases, but…") is not fence-sitting, as long as it is stated clearly and kept the same throughout. And it is a myth that your opinion "belongs in the conclusion": the Band 7 descriptor asks for "a clear position throughout the response".

A fence-sitting essay loses marks on Task Response even when both sides are explained well — the task specifically asks for YOUR opinion, not just balanced description.

Not every phrase bank suits every essay type. Problem/Solution language belongs in Problem-Solution essays, not a pure Opinion essay — using it where it wasn't asked for signals you've misread the task.

Task Response is about addressing every part of the question that was actually asked. The phrases themselves earn nothing; what they reveal does. When "a viable solution would be" appears in an essay that never asked for solutions, the content has drifted away from the task, and the examiner will treat it as only partly addressing the question.

Using the exact same phrase for every instance of a function — "For example" every single time — reads as mechanical. Pick a different phrase from the same category each time you need it.

Linking phrases are judged mainly under Coherence and Cohesion. The public descriptors reward "a range of cohesive devices" used appropriately, and they list cohesion that is "faulty or mechanical" as a Band 6 feature. So variety helps only when each phrase fits the logic of the sentence. A linker at the start of every sentence, or a showy phrase that does not match the relationship between ideas, counts against you. Referring back with pronouns and synonyms (this policy, such measures) is part of cohesion too.

Absolute words — all, always, never, everyone, completely, proves — turn a reasonable point into a claim that one counter-example destroys. For claims about facts and effects, use hedging language: tend to, can, may, often, many, most, in many cases, research suggests. Hedge the evidence, not your position: "I believe…" stays firm, while "this can lead to…" stays accurate.

The public Task Response descriptor notes that even at Band 7 there may be "a tendency to over-generalise". A sweeping claim makes your support look weak, because the reader can immediately think of exceptions. The opposite mistake also costs you: hedging every sentence ("it might perhaps possibly…") hides your position, which Task Response needs to see clearly.

English words come with fixed partners (collocations). Many essay errors come from pairing words the way they pair in Bangla: "do a crime", "make a research", "give harm", "strong rain". Learn the partner together with the word: commit a crime, carry out research, cause harm, heavy rain, play a role in, pose a threat to, raise awareness of, meet the demand for.

Collocation is assessed under Lexical Resource. The public Band 7 descriptor mentions "some awareness of style and collocation", and frequent errors in word choice keep the score below Band 7 even when your vocabulary is wide. A plain word in the right partnership is worth more than an advanced word in the wrong one.

Before submitting, ask 3 quick questions: (1) Did I address every part of the prompt's key words (Chapter 10, Trick 1)? (2) Does my structure actually match the essay type (Chapter 10, Trick 3), not a generic template? (3) Is my position clear and consistent throughout, if this essay type requires one (Chapter 11, Trick 4)? Finally, glance at the word count — Task 2 answers under 250 words are penalised.

Task Response is one of the four equally weighted criteria, and it is limited whenever an essay addresses only part of the task or drifts from it — however good the language is. This check catches that costly mistake — a fluent essay that doesn't fully answer the question asked — in under a minute, while there is still time to fix it.

Task 2 counts for twice as much as Task 1, so give it about 40 of your 60 minutes (you may do it first if you prefer). Inside those 40 minutes, decide the clock before you write: about 5 minutes to read and plan, about 30 to write (introduction, two body paragraphs, conclusion), and the last 3–5 to check. If you fall behind, shorten a paragraph — never drop a required part or the conclusion.

An essay that runs out of time usually loses its conclusion or the second part of the question, which limits Task Response however good the first half is, and an answer under 250 words is penalised. Running out of time is a planning failure, not a language failure — and it is the easiest one to fix.

A Task 2 conclusion is two or three sentences that restate your answer in new words and sum up your main reasons. For a Problem-Solution essay, restate the main cause and the main solution; for a two-part question, give the short answer to each part. Add no new arguments, examples or figures, and skip empty lines such as "In conclusion, this essay has discussed…".

A new point in the conclusion cannot be developed, so it adds nothing to Task Response and weakens the logical progression that Coherence and Cohesion rewards. A conclusion that simply restates the topic, or wanders, makes your final position less clear — and the descriptors look for a position that is clear throughout the response.

In an Opinion essay, examiners do not mark whether they agree with you. In your planning minutes, jot down two quick ideas for each side, then choose the position you can support with two clear, explained reasons — even if your personal view is mixed. A qualified position ("I largely agree, but…") is a good choice when you have strong support for the main view and one clear limit.

Task Response rewards a clear position whose main ideas are "extended and supported". Choosing a side because it feels more original, and then finding only one vague reason, produces thin body paragraphs — the position is clear, but the support is missing.

Words like information, advice, research, evidence, equipment, knowledge and progress are uncountable in standard academic English — never add "-s," never put "a/an" or "many" directly before them, and use a singular verb ("research shows"). Use "much," "a lot of," "little," or "a piece of" instead.

This is a very common error even in otherwise strong scripts — fluent writers add "-s" by habit. One slip will not decide your band, but repeated errors of this kind count against Grammatical Range and Accuracy (and word form under Lexical Resource), and frequent errors keep both criteria below Band 7.

Certain words always pair with one specific preposition, or none at all — "discuss" takes NO "about"; for stating your OWN opinion, use "in my opinion" or "I believe," not "according to me" (that phrase is only for citing someone else's claim).

Fixed word partnerships like these can't be reasoned out grammatically — they simply have to be learned. Errors in word choice and collocation are assessed under Lexical Resource; the Band 7 descriptor allows only occasional errors, so frequent ones hold the score down even when your vocabulary range is good.

In a long subject (subject + prepositional phrase before the verb), the verb must agree with the REAL head noun, not whichever noun happens to sit right before the verb.

This structure appears constantly in academic writing — and in Task 1 especially ("The number of…", "The proportion of…") — and agreement errors are among the most frequent Grammatical Range and Accuracy errors in otherwise strong essays, precisely because the wrong noun sits closer to the verb.

When you mean an uncountable idea or a plural noun in general, use no article — "education," "technology," "society," "young people" — not "the education," "the society." Two limits to the rule: (1) a singular countable noun always needs a/an or the ("living in a big city"); (2) some nouns take the by convention even in a general sense — the environment, the internet, the economy, the government, the media.

Article errors — an extra "the" before abstract nouns, or a missing "a/an" before singular nouns — tend to repeat throughout a script. Because it's a pattern, not a one-off mistake, fixing the pattern removes many errors at once.

Words that look almost identical but function differently — economic/economical, affect/effect — are chosen by grammatical role and meaning, not by "which one sounds right."

These pairs are easy to mishear when proofreading by ear, so checking word FORM (is it functioning as a noun, adjective, or verb here?) catches what reading-aloud won't.

Two complete sentences can't be joined with just a comma (a "comma splice"). Use a full stop, a semicolon, a conjunction (and, but, so, while, although) or a relative pronoun (which). Note that however and therefore are not conjunctions: "…, however, …" between two sentences is still a comma splice.

This is a genuine sentence-boundary error, not a style choice — examiners read it as a Grammatical Range and Accuracy fault, since it shows uncertainty about where one sentence ends and the next begins.