Madrid PAU English - Additional Foreign Language cheat sheet

Global questions ask for the main idea, the author's purpose, the tone, or the best title; the answer must cover the whole text, not one paragraph.

Locate the thesis in the first or last paragraph of an expository text; in argumentative texts it usually follows a contrast marker such as 'however' or 'yet'.

Reject options that are true but too narrow (one detail), too broad (beyond the text), or evaluative in a way the author never states.

Text-type recognition drives purpose: a review evaluates, a news report informs, an opinion column argues, a leaflet persuades or instructs.

Tone words to discriminate: neutral, ironic, critical, enthusiastic, nostalgic, cautious; irony is signalled by quotation marks, exaggeration, or an unexpected adjective.

Paragraph-function mapping: identify whether each paragraph narrates, exemplifies, contrasts, or concludes; the main idea usually spans the introduction and conclusion.

If two options seem right, prefer the one whose scope matches the number of paragraphs it summarizes; a single-paragraph option rarely answers a global question.

Structural signposts (however, therefore, in fact, on the whole) mark the author's stance more reliably than individual adjectives.

Specific-information questions are answered by locating a paraphrase, not a verbatim match; the correct option rewrites the source with synonyms or a changed word class.

Scan for proper nouns, numbers, dates, and capitalised terms first; they anchor the relevant line faster than reading sequentially.

Beware of distractors that copy the text's wording but invert it with 'not', 'only', 'always', or a swapped subject.

Inference answers must be logically forced by the text plus general knowledge; if the option needs an extra assumption, it is wrong.

Pronoun and reference questions (it, this, they, the former) require tracing back to the nearest compatible noun phrase, respecting number and gender.

Vocabulary-in-context items: substitute each option into the sentence and keep the one that preserves both meaning and register.

Negative factual questions ('which is NOT mentioned') are solved by verifying three options against the text and eliminating them.

Numbers and quantities: check whether the text gives an absolute figure, a proportion, or an approximation; options often change the unit or the comparison base.

Verb-form choice depends on time markers: 'since/for' plus present perfect, 'by the time' plus past perfect, 'this time next week' plus future continuous.

Conditionals: zero (present/present), first (present/will), second (past/would), third (had + participle/would have + participle); mixed conditionals pair a past condition with a present result.

Reported speech shifts tense one step back and adjusts time and place words (now→then, today→that day, here→there), except when the reporting verb is present tense.

Passive transformation keeps the original tense: 'They built it' → 'It was built'; progressive passive is 'is being + participle', perfect passive is 'has been + participle'.

Word formation: suffixes change class (-ness, -ment, -tion for nouns; -ful, -less, -ive, -able for adjectives; -ise/-ize, -en for verbs), prefixes usually negate (un-, in-, im-, dis-, mis-).

Collocation and preposition rules: depend on, insist on, be interested in, be good at, look forward to + -ing; 'used to + infinitive' (past habit) versus 'be used to + -ing' (accustomed).

Relative clauses: defining clauses take no comma and allow 'that'; non-defining clauses take commas and reject 'that'; 'whose' marks possession.

Modals of deduction: must (certain), may/might/could (possible), can't (impossible); 'must have + participle' expresses past certainty, 'needn't have + participle' a past action done unnecessarily.

Countability and agreement traps: 'news', 'information', 'advice', 'furniture' are uncountable; 'the number of' takes a singular verb, 'a number of' a plural one.

Standard task formats: an email or letter (formal or informal), an opinion essay, a for-and-against essay, a narrative, or a description based on a prompt; word limits are usually 120–150 or 150–180 words.

Paragraph architecture: introduction stating the topic and position, two body paragraphs each with one controlling idea plus support, conclusion restating without new arguments.

Cohesion devices to deploy deliberately: addition (furthermore, in addition), contrast (however, whereas, despite + noun/-ing), cause (therefore, as a result), exemplification (for instance, such as), sequencing (firstly, subsequently, finally).

Range is scored through varied structures: at least one conditional, one passive, one relative clause, and one impersonal construction ('It is widely believed that…') in a 150-word text.

Accuracy priorities in order of damage: verb tense and agreement, subject–verb inversion after 'not only', article use with countable singulars, and preposition collocations.

Register control: formal writing avoids contractions, phrasal verbs where a Latinate verb exists, exclamation marks, and direct address; informal emails use contractions and fixed openings (Hi X, / Best wishes,).

Punctuation rules that examiners penalise: no comma splice between independent clauses (use a full stop, semicolon, or conjunction); a comma before 'which' in non-defining clauses but not before 'that'.

Self-editing checklist: verify every verb has a subject and correct tense, every sentence has a main clause, and the word count stays within the stated limit.

Global questions ask for the main idea, the author's purpose, the tone, or the best title; the answer must cover the whole text, not one paragraph.

Locate the thesis in the first or last paragraph of an expository text; in argumentative texts it usually follows a contrast marker such as 'however' or 'yet'.

Reject options that are true but too narrow (one detail), too broad (beyond the text), or evaluative in a way the author never states.

Text-type recognition drives purpose: a review evaluates, a news report informs, an opinion column argues, a leaflet persuades or instructs.

Tone words to discriminate: neutral, ironic, critical, enthusiastic, nostalgic, cautious; irony is signalled by quotation marks, exaggeration, or an unexpected adjective.

Paragraph-function mapping: identify whether each paragraph narrates, exemplifies, contrasts, or concludes; the main idea usually spans the introduction and conclusion.

If two options seem right, prefer the one whose scope matches the number of paragraphs it summarizes; a single-paragraph option rarely answers a global question.

Structural signposts (however, therefore, in fact, on the whole) mark the author's stance more reliably than individual adjectives.

Specific-information questions are answered by locating a paraphrase, not a verbatim match; the correct option rewrites the source with synonyms or a changed word class.

Scan for proper nouns, numbers, dates, and capitalised terms first; they anchor the relevant line faster than reading sequentially.

Beware of distractors that copy the text's wording but invert it with 'not', 'only', 'always', or a swapped subject.

Inference answers must be logically forced by the text plus general knowledge; if the option needs an extra assumption, it is wrong.

Pronoun and reference questions (it, this, they, the former) require tracing back to the nearest compatible noun phrase, respecting number and gender.

Vocabulary-in-context items: substitute each option into the sentence and keep the one that preserves both meaning and register.

Negative factual questions ('which is NOT mentioned') are solved by verifying three options against the text and eliminating them.

Numbers and quantities: check whether the text gives an absolute figure, a proportion, or an approximation; options often change the unit or the comparison base.

Verb-form choice depends on time markers: 'since/for' plus present perfect, 'by the time' plus past perfect, 'this time next week' plus future continuous.

Conditionals: zero (present/present), first (present/will), second (past/would), third (had + participle/would have + participle); mixed conditionals pair a past condition with a present result.

Reported speech shifts tense one step back and adjusts time and place words (now→then, today→that day, here→there), except when the reporting verb is present tense.

Passive transformation keeps the original tense: 'They built it' → 'It was built'; progressive passive is 'is being + participle', perfect passive is 'has been + participle'.

Word formation: suffixes change class (-ness, -ment, -tion for nouns; -ful, -less, -ive, -able for adjectives; -ise/-ize, -en for verbs), prefixes usually negate (un-, in-, im-, dis-, mis-).

Collocation and preposition rules: depend on, insist on, be interested in, be good at, look forward to + -ing; 'used to + infinitive' (past habit) versus 'be used to + -ing' (accustomed).

Relative clauses: defining clauses take no comma and allow 'that'; non-defining clauses take commas and reject 'that'; 'whose' marks possession.

Modals of deduction: must (certain), may/might/could (possible), can't (impossible); 'must have + participle' expresses past certainty, 'needn't have + participle' a past action done unnecessarily.

Countability and agreement traps: 'news', 'information', 'advice', 'furniture' are uncountable; 'the number of' takes a singular verb, 'a number of' a plural one.

Standard task formats: an email or letter (formal or informal), an opinion essay, a for-and-against essay, a narrative, or a description based on a prompt; word limits are usually 120–150 or 150–180 words.

Paragraph architecture: introduction stating the topic and position, two body paragraphs each with one controlling idea plus support, conclusion restating without new arguments.

Cohesion devices to deploy deliberately: addition (furthermore, in addition), contrast (however, whereas, despite + noun/-ing), cause (therefore, as a result), exemplification (for instance, such as), sequencing (firstly, subsequently, finally).

Range is scored through varied structures: at least one conditional, one passive, one relative clause, and one impersonal construction ('It is widely believed that…') in a 150-word text.

Accuracy priorities in order of damage: verb tense and agreement, subject–verb inversion after 'not only', article use with countable singulars, and preposition collocations.

Register control: formal writing avoids contractions, phrasal verbs where a Latinate verb exists, exclamation marks, and direct address; informal emails use contractions and fixed openings (Hi X, / Best wishes,).

Punctuation rules that examiners penalise: no comma splice between independent clauses (use a full stop, semicolon, or conjunction); a comma before 'which' in non-defining clauses but not before 'that'.

Self-editing checklist: verify every verb has a subject and correct tense, every sentence has a main clause, and the word count stays within the stated limit.

Global questions ask for the main idea, the author's purpose, the tone, or the best title; the answer must cover the whole text, not one paragraph.

Locate the thesis in the first or last paragraph of an expository text; in argumentative texts it usually follows a contrast marker such as 'however' or 'yet'.

Reject options that are true but too narrow (one detail), too broad (beyond the text), or evaluative in a way the author never states.

Text-type recognition drives purpose: a review evaluates, a news report informs, an opinion column argues, a leaflet persuades or instructs.

Tone words to discriminate: neutral, ironic, critical, enthusiastic, nostalgic, cautious; irony is signalled by quotation marks, exaggeration, or an unexpected adjective.

Paragraph-function mapping: identify whether each paragraph narrates, exemplifies, contrasts, or concludes; the main idea usually spans the introduction and conclusion.

If two options seem right, prefer the one whose scope matches the number of paragraphs it summarizes; a single-paragraph option rarely answers a global question.

Structural signposts (however, therefore, in fact, on the whole) mark the author's stance more reliably than individual adjectives.

Specific-information questions are answered by locating a paraphrase, not a verbatim match; the correct option rewrites the source with synonyms or a changed word class.

Scan for proper nouns, numbers, dates, and capitalised terms first; they anchor the relevant line faster than reading sequentially.

Beware of distractors that copy the text's wording but invert it with 'not', 'only', 'always', or a swapped subject.

Inference answers must be logically forced by the text plus general knowledge; if the option needs an extra assumption, it is wrong.

Pronoun and reference questions (it, this, they, the former) require tracing back to the nearest compatible noun phrase, respecting number and gender.

Vocabulary-in-context items: substitute each option into the sentence and keep the one that preserves both meaning and register.

Negative factual questions ('which is NOT mentioned') are solved by verifying three options against the text and eliminating them.

Numbers and quantities: check whether the text gives an absolute figure, a proportion, or an approximation; options often change the unit or the comparison base.

Verb-form choice depends on time markers: 'since/for' plus present perfect, 'by the time' plus past perfect, 'this time next week' plus future continuous.

Conditionals: zero (present/present), first (present/will), second (past/would), third (had + participle/would have + participle); mixed conditionals pair a past condition with a present result.

Reported speech shifts tense one step back and adjusts time and place words (now→then, today→that day, here→there), except when the reporting verb is present tense.

Passive transformation keeps the original tense: 'They built it' → 'It was built'; progressive passive is 'is being + participle', perfect passive is 'has been + participle'.

Word formation: suffixes change class (-ness, -ment, -tion for nouns; -ful, -less, -ive, -able for adjectives; -ise/-ize, -en for verbs), prefixes usually negate (un-, in-, im-, dis-, mis-).

Collocation and preposition rules: depend on, insist on, be interested in, be good at, look forward to + -ing; 'used to + infinitive' (past habit) versus 'be used to + -ing' (accustomed).

Relative clauses: defining clauses take no comma and allow 'that'; non-defining clauses take commas and reject 'that'; 'whose' marks possession.

Modals of deduction: must (certain), may/might/could (possible), can't (impossible); 'must have + participle' expresses past certainty, 'needn't have + participle' a past action done unnecessarily.

Countability and agreement traps: 'news', 'information', 'advice', 'furniture' are uncountable; 'the number of' takes a singular verb, 'a number of' a plural one.

Standard task formats: an email or letter (formal or informal), an opinion essay, a for-and-against essay, a narrative, or a description based on a prompt; word limits are usually 120–150 or 150–180 words.

Paragraph architecture: introduction stating the topic and position, two body paragraphs each with one controlling idea plus support, conclusion restating without new arguments.

Cohesion devices to deploy deliberately: addition (furthermore, in addition), contrast (however, whereas, despite + noun/-ing), cause (therefore, as a result), exemplification (for instance, such as), sequencing (firstly, subsequently, finally).

Range is scored through varied structures: at least one conditional, one passive, one relative clause, and one impersonal construction ('It is widely believed that…') in a 150-word text.

Accuracy priorities in order of damage: verb tense and agreement, subject–verb inversion after 'not only', article use with countable singulars, and preposition collocations.

Register control: formal writing avoids contractions, phrasal verbs where a Latinate verb exists, exclamation marks, and direct address; informal emails use contractions and fixed openings (Hi X, / Best wishes,).

Punctuation rules that examiners penalise: no comma splice between independent clauses (use a full stop, semicolon, or conjunction); a comma before 'which' in non-defining clauses but not before 'that'.

Self-editing checklist: verify every verb has a subject and correct tense, every sentence has a main clause, and the word count stays within the stated limit.

Global questions ask for the main idea, the author's purpose, the tone, or the best title; the answer must cover the whole text, not one paragraph.

Locate the thesis in the first or last paragraph of an expository text; in argumentative texts it usually follows a contrast marker such as 'however' or 'yet'.

Reject options that are true but too narrow (one detail), too broad (beyond the text), or evaluative in a way the author never states.

Text-type recognition drives purpose: a review evaluates, a news report informs, an opinion column argues, a leaflet persuades or instructs.

Tone words to discriminate: neutral, ironic, critical, enthusiastic, nostalgic, cautious; irony is signalled by quotation marks, exaggeration, or an unexpected adjective.

Paragraph-function mapping: identify whether each paragraph narrates, exemplifies, contrasts, or concludes; the main idea usually spans the introduction and conclusion.

If two options seem right, prefer the one whose scope matches the number of paragraphs it summarizes; a single-paragraph option rarely answers a global question.

Structural signposts (however, therefore, in fact, on the whole) mark the author's stance more reliably than individual adjectives.

Specific-information questions are answered by locating a paraphrase, not a verbatim match; the correct option rewrites the source with synonyms or a changed word class.

Scan for proper nouns, numbers, dates, and capitalised terms first; they anchor the relevant line faster than reading sequentially.

Beware of distractors that copy the text's wording but invert it with 'not', 'only', 'always', or a swapped subject.

Inference answers must be logically forced by the text plus general knowledge; if the option needs an extra assumption, it is wrong.

Pronoun and reference questions (it, this, they, the former) require tracing back to the nearest compatible noun phrase, respecting number and gender.

Vocabulary-in-context items: substitute each option into the sentence and keep the one that preserves both meaning and register.

Negative factual questions ('which is NOT mentioned') are solved by verifying three options against the text and eliminating them.

Numbers and quantities: check whether the text gives an absolute figure, a proportion, or an approximation; options often change the unit or the comparison base.

Verb-form choice depends on time markers: 'since/for' plus present perfect, 'by the time' plus past perfect, 'this time next week' plus future continuous.

Conditionals: zero (present/present), first (present/will), second (past/would), third (had + participle/would have + participle); mixed conditionals pair a past condition with a present result.

Reported speech shifts tense one step back and adjusts time and place words (now→then, today→that day, here→there), except when the reporting verb is present tense.

Passive transformation keeps the original tense: 'They built it' → 'It was built'; progressive passive is 'is being + participle', perfect passive is 'has been + participle'.

Word formation: suffixes change class (-ness, -ment, -tion for nouns; -ful, -less, -ive, -able for adjectives; -ise/-ize, -en for verbs), prefixes usually negate (un-, in-, im-, dis-, mis-).

Collocation and preposition rules: depend on, insist on, be interested in, be good at, look forward to + -ing; 'used to + infinitive' (past habit) versus 'be used to + -ing' (accustomed).

Relative clauses: defining clauses take no comma and allow 'that'; non-defining clauses take commas and reject 'that'; 'whose' marks possession.

Modals of deduction: must (certain), may/might/could (possible), can't (impossible); 'must have + participle' expresses past certainty, 'needn't have + participle' a past action done unnecessarily.

Countability and agreement traps: 'news', 'information', 'advice', 'furniture' are uncountable; 'the number of' takes a singular verb, 'a number of' a plural one.

Standard task formats: an email or letter (formal or informal), an opinion essay, a for-and-against essay, a narrative, or a description based on a prompt; word limits are usually 120–150 or 150–180 words.

Paragraph architecture: introduction stating the topic and position, two body paragraphs each with one controlling idea plus support, conclusion restating without new arguments.

Cohesion devices to deploy deliberately: addition (furthermore, in addition), contrast (however, whereas, despite + noun/-ing), cause (therefore, as a result), exemplification (for instance, such as), sequencing (firstly, subsequently, finally).

Range is scored through varied structures: at least one conditional, one passive, one relative clause, and one impersonal construction ('It is widely believed that…') in a 150-word text.

Accuracy priorities in order of damage: verb tense and agreement, subject–verb inversion after 'not only', article use with countable singulars, and preposition collocations.

Register control: formal writing avoids contractions, phrasal verbs where a Latinate verb exists, exclamation marks, and direct address; informal emails use contractions and fixed openings (Hi X, / Best wishes,).

Punctuation rules that examiners penalise: no comma splice between independent clauses (use a full stop, semicolon, or conjunction); a comma before 'which' in non-defining clauses but not before 'that'.

Self-editing checklist: verify every verb has a subject and correct tense, every sentence has a main clause, and the word count stays within the stated limit.

Global questions ask for the main idea, the author's purpose, the tone, or the best title; the answer must cover the whole text, not one paragraph.

Locate the thesis in the first or last paragraph of an expository text; in argumentative texts it usually follows a contrast marker such as 'however' or 'yet'.

Reject options that are true but too narrow (one detail), too broad (beyond the text), or evaluative in a way the author never states.

Text-type recognition drives purpose: a review evaluates, a news report informs, an opinion column argues, a leaflet persuades or instructs.

Tone words to discriminate: neutral, ironic, critical, enthusiastic, nostalgic, cautious; irony is signalled by quotation marks, exaggeration, or an unexpected adjective.

Paragraph-function mapping: identify whether each paragraph narrates, exemplifies, contrasts, or concludes; the main idea usually spans the introduction and conclusion.

If two options seem right, prefer the one whose scope matches the number of paragraphs it summarizes; a single-paragraph option rarely answers a global question.

Structural signposts (however, therefore, in fact, on the whole) mark the author's stance more reliably than individual adjectives.

Specific-information questions are answered by locating a paraphrase, not a verbatim match; the correct option rewrites the source with synonyms or a changed word class.

Scan for proper nouns, numbers, dates, and capitalised terms first; they anchor the relevant line faster than reading sequentially.

Beware of distractors that copy the text's wording but invert it with 'not', 'only', 'always', or a swapped subject.

Inference answers must be logically forced by the text plus general knowledge; if the option needs an extra assumption, it is wrong.

Pronoun and reference questions (it, this, they, the former) require tracing back to the nearest compatible noun phrase, respecting number and gender.

Vocabulary-in-context items: substitute each option into the sentence and keep the one that preserves both meaning and register.

Negative factual questions ('which is NOT mentioned') are solved by verifying three options against the text and eliminating them.

Numbers and quantities: check whether the text gives an absolute figure, a proportion, or an approximation; options often change the unit or the comparison base.

Verb-form choice depends on time markers: 'since/for' plus present perfect, 'by the time' plus past perfect, 'this time next week' plus future continuous.

Conditionals: zero (present/present), first (present/will), second (past/would), third (had + participle/would have + participle); mixed conditionals pair a past condition with a present result.

Reported speech shifts tense one step back and adjusts time and place words (now→then, today→that day, here→there), except when the reporting verb is present tense.

Passive transformation keeps the original tense: 'They built it' → 'It was built'; progressive passive is 'is being + participle', perfect passive is 'has been + participle'.

Word formation: suffixes change class (-ness, -ment, -tion for nouns; -ful, -less, -ive, -able for adjectives; -ise/-ize, -en for verbs), prefixes usually negate (un-, in-, im-, dis-, mis-).

Collocation and preposition rules: depend on, insist on, be interested in, be good at, look forward to + -ing; 'used to + infinitive' (past habit) versus 'be used to + -ing' (accustomed).

Relative clauses: defining clauses take no comma and allow 'that'; non-defining clauses take commas and reject 'that'; 'whose' marks possession.

Modals of deduction: must (certain), may/might/could (possible), can't (impossible); 'must have + participle' expresses past certainty, 'needn't have + participle' a past action done unnecessarily.

Countability and agreement traps: 'news', 'information', 'advice', 'furniture' are uncountable; 'the number of' takes a singular verb, 'a number of' a plural one.

Standard task formats: an email or letter (formal or informal), an opinion essay, a for-and-against essay, a narrative, or a description based on a prompt; word limits are usually 120–150 or 150–180 words.

Paragraph architecture: introduction stating the topic and position, two body paragraphs each with one controlling idea plus support, conclusion restating without new arguments.

Cohesion devices to deploy deliberately: addition (furthermore, in addition), contrast (however, whereas, despite + noun/-ing), cause (therefore, as a result), exemplification (for instance, such as), sequencing (firstly, subsequently, finally).

Range is scored through varied structures: at least one conditional, one passive, one relative clause, and one impersonal construction ('It is widely believed that…') in a 150-word text.

Accuracy priorities in order of damage: verb tense and agreement, subject–verb inversion after 'not only', article use with countable singulars, and preposition collocations.

Register control: formal writing avoids contractions, phrasal verbs where a Latinate verb exists, exclamation marks, and direct address; informal emails use contractions and fixed openings (Hi X, / Best wishes,).

Punctuation rules that examiners penalise: no comma splice between independent clauses (use a full stop, semicolon, or conjunction); a comma before 'which' in non-defining clauses but not before 'that'.

Self-editing checklist: verify every verb has a subject and correct tense, every sentence has a main clause, and the word count stays within the stated limit.

Global questions ask for the main idea, the author's purpose, the tone, or the best title; the answer must cover the whole text, not one paragraph.

Locate the thesis in the first or last paragraph of an expository text; in argumentative texts it usually follows a contrast marker such as 'however' or 'yet'.

Reject options that are true but too narrow (one detail), too broad (beyond the text), or evaluative in a way the author never states.

Text-type recognition drives purpose: a review evaluates, a news report informs, an opinion column argues, a leaflet persuades or instructs.

Tone words to discriminate: neutral, ironic, critical, enthusiastic, nostalgic, cautious; irony is signalled by quotation marks, exaggeration, or an unexpected adjective.

Paragraph-function mapping: identify whether each paragraph narrates, exemplifies, contrasts, or concludes; the main idea usually spans the introduction and conclusion.

If two options seem right, prefer the one whose scope matches the number of paragraphs it summarizes; a single-paragraph option rarely answers a global question.

Structural signposts (however, therefore, in fact, on the whole) mark the author's stance more reliably than individual adjectives.

Specific-information questions are answered by locating a paraphrase, not a verbatim match; the correct option rewrites the source with synonyms or a changed word class.

Scan for proper nouns, numbers, dates, and capitalised terms first; they anchor the relevant line faster than reading sequentially.

Beware of distractors that copy the text's wording but invert it with 'not', 'only', 'always', or a swapped subject.

Inference answers must be logically forced by the text plus general knowledge; if the option needs an extra assumption, it is wrong.

Pronoun and reference questions (it, this, they, the former) require tracing back to the nearest compatible noun phrase, respecting number and gender.

Vocabulary-in-context items: substitute each option into the sentence and keep the one that preserves both meaning and register.

Negative factual questions ('which is NOT mentioned') are solved by verifying three options against the text and eliminating them.

Numbers and quantities: check whether the text gives an absolute figure, a proportion, or an approximation; options often change the unit or the comparison base.

Verb-form choice depends on time markers: 'since/for' plus present perfect, 'by the time' plus past perfect, 'this time next week' plus future continuous.

Conditionals: zero (present/present), first (present/will), second (past/would), third (had + participle/would have + participle); mixed conditionals pair a past condition with a present result.

Reported speech shifts tense one step back and adjusts time and place words (now→then, today→that day, here→there), except when the reporting verb is present tense.

Passive transformation keeps the original tense: 'They built it' → 'It was built'; progressive passive is 'is being + participle', perfect passive is 'has been + participle'.

Word formation: suffixes change class (-ness, -ment, -tion for nouns; -ful, -less, -ive, -able for adjectives; -ise/-ize, -en for verbs), prefixes usually negate (un-, in-, im-, dis-, mis-).

Collocation and preposition rules: depend on, insist on, be interested in, be good at, look forward to + -ing; 'used to + infinitive' (past habit) versus 'be used to + -ing' (accustomed).

Relative clauses: defining clauses take no comma and allow 'that'; non-defining clauses take commas and reject 'that'; 'whose' marks possession.

Modals of deduction: must (certain), may/might/could (possible), can't (impossible); 'must have + participle' expresses past certainty, 'needn't have + participle' a past action done unnecessarily.

Countability and agreement traps: 'news', 'information', 'advice', 'furniture' are uncountable; 'the number of' takes a singular verb, 'a number of' a plural one.

Standard task formats: an email or letter (formal or informal), an opinion essay, a for-and-against essay, a narrative, or a description based on a prompt; word limits are usually 120–150 or 150–180 words.

Paragraph architecture: introduction stating the topic and position, two body paragraphs each with one controlling idea plus support, conclusion restating without new arguments.

Cohesion devices to deploy deliberately: addition (furthermore, in addition), contrast (however, whereas, despite + noun/-ing), cause (therefore, as a result), exemplification (for instance, such as), sequencing (firstly, subsequently, finally).

Range is scored through varied structures: at least one conditional, one passive, one relative clause, and one impersonal construction ('It is widely believed that…') in a 150-word text.

Accuracy priorities in order of damage: verb tense and agreement, subject–verb inversion after 'not only', article use with countable singulars, and preposition collocations.

Register control: formal writing avoids contractions, phrasal verbs where a Latinate verb exists, exclamation marks, and direct address; informal emails use contractions and fixed openings (Hi X, / Best wishes,).

Punctuation rules that examiners penalise: no comma splice between independent clauses (use a full stop, semicolon, or conjunction); a comma before 'which' in non-defining clauses but not before 'that'.

Self-editing checklist: verify every verb has a subject and correct tense, every sentence has a main clause, and the word count stays within the stated limit.

Global questions ask for the main idea, the author's purpose, the tone, or the best title; the answer must cover the whole text, not one paragraph.

Locate the thesis in the first or last paragraph of an expository text; in argumentative texts it usually follows a contrast marker such as 'however' or 'yet'.

Reject options that are true but too narrow (one detail), too broad (beyond the text), or evaluative in a way the author never states.

Text-type recognition drives purpose: a review evaluates, a news report informs, an opinion column argues, a leaflet persuades or instructs.

Tone words to discriminate: neutral, ironic, critical, enthusiastic, nostalgic, cautious; irony is signalled by quotation marks, exaggeration, or an unexpected adjective.

Paragraph-function mapping: identify whether each paragraph narrates, exemplifies, contrasts, or concludes; the main idea usually spans the introduction and conclusion.

If two options seem right, prefer the one whose scope matches the number of paragraphs it summarizes; a single-paragraph option rarely answers a global question.

Structural signposts (however, therefore, in fact, on the whole) mark the author's stance more reliably than individual adjectives.

Specific-information questions are answered by locating a paraphrase, not a verbatim match; the correct option rewrites the source with synonyms or a changed word class.

Scan for proper nouns, numbers, dates, and capitalised terms first; they anchor the relevant line faster than reading sequentially.

Beware of distractors that copy the text's wording but invert it with 'not', 'only', 'always', or a swapped subject.

Inference answers must be logically forced by the text plus general knowledge; if the option needs an extra assumption, it is wrong.

Pronoun and reference questions (it, this, they, the former) require tracing back to the nearest compatible noun phrase, respecting number and gender.

Vocabulary-in-context items: substitute each option into the sentence and keep the one that preserves both meaning and register.

Negative factual questions ('which is NOT mentioned') are solved by verifying three options against the text and eliminating them.

Numbers and quantities: check whether the text gives an absolute figure, a proportion, or an approximation; options often change the unit or the comparison base.

Verb-form choice depends on time markers: 'since/for' plus present perfect, 'by the time' plus past perfect, 'this time next week' plus future continuous.

Conditionals: zero (present/present), first (present/will), second (past/would), third (had + participle/would have + participle); mixed conditionals pair a past condition with a present result.

Reported speech shifts tense one step back and adjusts time and place words (now→then, today→that day, here→there), except when the reporting verb is present tense.

Passive transformation keeps the original tense: 'They built it' → 'It was built'; progressive passive is 'is being + participle', perfect passive is 'has been + participle'.

Word formation: suffixes change class (-ness, -ment, -tion for nouns; -ful, -less, -ive, -able for adjectives; -ise/-ize, -en for verbs), prefixes usually negate (un-, in-, im-, dis-, mis-).

Collocation and preposition rules: depend on, insist on, be interested in, be good at, look forward to + -ing; 'used to + infinitive' (past habit) versus 'be used to + -ing' (accustomed).

Relative clauses: defining clauses take no comma and allow 'that'; non-defining clauses take commas and reject 'that'; 'whose' marks possession.

Modals of deduction: must (certain), may/might/could (possible), can't (impossible); 'must have + participle' expresses past certainty, 'needn't have + participle' a past action done unnecessarily.

Countability and agreement traps: 'news', 'information', 'advice', 'furniture' are uncountable; 'the number of' takes a singular verb, 'a number of' a plural one.

Standard task formats: an email or letter (formal or informal), an opinion essay, a for-and-against essay, a narrative, or a description based on a prompt; word limits are usually 120–150 or 150–180 words.

Paragraph architecture: introduction stating the topic and position, two body paragraphs each with one controlling idea plus support, conclusion restating without new arguments.

Cohesion devices to deploy deliberately: addition (furthermore, in addition), contrast (however, whereas, despite + noun/-ing), cause (therefore, as a result), exemplification (for instance, such as), sequencing (firstly, subsequently, finally).

Range is scored through varied structures: at least one conditional, one passive, one relative clause, and one impersonal construction ('It is widely believed that…') in a 150-word text.

Accuracy priorities in order of damage: verb tense and agreement, subject–verb inversion after 'not only', article use with countable singulars, and preposition collocations.

Register control: formal writing avoids contractions, phrasal verbs where a Latinate verb exists, exclamation marks, and direct address; informal emails use contractions and fixed openings (Hi X, / Best wishes,).

Punctuation rules that examiners penalise: no comma splice between independent clauses (use a full stop, semicolon, or conjunction); a comma before 'which' in non-defining clauses but not before 'that'.

Self-editing checklist: verify every verb has a subject and correct tense, every sentence has a main clause, and the word count stays within the stated limit.

Global questions ask for the main idea, the author's purpose, the tone, or the best title; the answer must cover the whole text, not one paragraph.

Locate the thesis in the first or last paragraph of an expository text; in argumentative texts it usually follows a contrast marker such as 'however' or 'yet'.

Reject options that are true but too narrow (one detail), too broad (beyond the text), or evaluative in a way the author never states.

Text-type recognition drives purpose: a review evaluates, a news report informs, an opinion column argues, a leaflet persuades or instructs.

Tone words to discriminate: neutral, ironic, critical, enthusiastic, nostalgic, cautious; irony is signalled by quotation marks, exaggeration, or an unexpected adjective.

Paragraph-function mapping: identify whether each paragraph narrates, exemplifies, contrasts, or concludes; the main idea usually spans the introduction and conclusion.

If two options seem right, prefer the one whose scope matches the number of paragraphs it summarizes; a single-paragraph option rarely answers a global question.

Structural signposts (however, therefore, in fact, on the whole) mark the author's stance more reliably than individual adjectives.

Specific-information questions are answered by locating a paraphrase, not a verbatim match; the correct option rewrites the source with synonyms or a changed word class.

Scan for proper nouns, numbers, dates, and capitalised terms first; they anchor the relevant line faster than reading sequentially.

Beware of distractors that copy the text's wording but invert it with 'not', 'only', 'always', or a swapped subject.

Inference answers must be logically forced by the text plus general knowledge; if the option needs an extra assumption, it is wrong.

Pronoun and reference questions (it, this, they, the former) require tracing back to the nearest compatible noun phrase, respecting number and gender.

Vocabulary-in-context items: substitute each option into the sentence and keep the one that preserves both meaning and register.

Negative factual questions ('which is NOT mentioned') are solved by verifying three options against the text and eliminating them.

Numbers and quantities: check whether the text gives an absolute figure, a proportion, or an approximation; options often change the unit or the comparison base.

Verb-form choice depends on time markers: 'since/for' plus present perfect, 'by the time' plus past perfect, 'this time next week' plus future continuous.

Conditionals: zero (present/present), first (present/will), second (past/would), third (had + participle/would have + participle); mixed conditionals pair a past condition with a present result.

Reported speech shifts tense one step back and adjusts time and place words (now→then, today→that day, here→there), except when the reporting verb is present tense.

Passive transformation keeps the original tense: 'They built it' → 'It was built'; progressive passive is 'is being + participle', perfect passive is 'has been + participle'.

Word formation: suffixes change class (-ness, -ment, -tion for nouns; -ful, -less, -ive, -able for adjectives; -ise/-ize, -en for verbs), prefixes usually negate (un-, in-, im-, dis-, mis-).

Collocation and preposition rules: depend on, insist on, be interested in, be good at, look forward to + -ing; 'used to + infinitive' (past habit) versus 'be used to + -ing' (accustomed).

Relative clauses: defining clauses take no comma and allow 'that'; non-defining clauses take commas and reject 'that'; 'whose' marks possession.

Modals of deduction: must (certain), may/might/could (possible), can't (impossible); 'must have + participle' expresses past certainty, 'needn't have + participle' a past action done unnecessarily.

Countability and agreement traps: 'news', 'information', 'advice', 'furniture' are uncountable; 'the number of' takes a singular verb, 'a number of' a plural one.

Standard task formats: an email or letter (formal or informal), an opinion essay, a for-and-against essay, a narrative, or a description based on a prompt; word limits are usually 120–150 or 150–180 words.

Paragraph architecture: introduction stating the topic and position, two body paragraphs each with one controlling idea plus support, conclusion restating without new arguments.

Cohesion devices to deploy deliberately: addition (furthermore, in addition), contrast (however, whereas, despite + noun/-ing), cause (therefore, as a result), exemplification (for instance, such as), sequencing (firstly, subsequently, finally).

Range is scored through varied structures: at least one conditional, one passive, one relative clause, and one impersonal construction ('It is widely believed that…') in a 150-word text.

Accuracy priorities in order of damage: verb tense and agreement, subject–verb inversion after 'not only', article use with countable singulars, and preposition collocations.

Register control: formal writing avoids contractions, phrasal verbs where a Latinate verb exists, exclamation marks, and direct address; informal emails use contractions and fixed openings (Hi X, / Best wishes,).

Punctuation rules that examiners penalise: no comma splice between independent clauses (use a full stop, semicolon, or conjunction); a comma before 'which' in non-defining clauses but not before 'that'.

Self-editing checklist: verify every verb has a subject and correct tense, every sentence has a main clause, and the word count stays within the stated limit.

Global questions ask for the main idea, the author's purpose, the tone, or the best title; the answer must cover the whole text, not one paragraph.

Locate the thesis in the first or last paragraph of an expository text; in argumentative texts it usually follows a contrast marker such as 'however' or 'yet'.

Reject options that are true but too narrow (one detail), too broad (beyond the text), or evaluative in a way the author never states.

Text-type recognition drives purpose: a review evaluates, a news report informs, an opinion column argues, a leaflet persuades or instructs.

Tone words to discriminate: neutral, ironic, critical, enthusiastic, nostalgic, cautious; irony is signalled by quotation marks, exaggeration, or an unexpected adjective.

Paragraph-function mapping: identify whether each paragraph narrates, exemplifies, contrasts, or concludes; the main idea usually spans the introduction and conclusion.

If two options seem right, prefer the one whose scope matches the number of paragraphs it summarizes; a single-paragraph option rarely answers a global question.

Structural signposts (however, therefore, in fact, on the whole) mark the author's stance more reliably than individual adjectives.

Specific-information questions are answered by locating a paraphrase, not a verbatim match; the correct option rewrites the source with synonyms or a changed word class.

Scan for proper nouns, numbers, dates, and capitalised terms first; they anchor the relevant line faster than reading sequentially.

Beware of distractors that copy the text's wording but invert it with 'not', 'only', 'always', or a swapped subject.

Inference answers must be logically forced by the text plus general knowledge; if the option needs an extra assumption, it is wrong.

Pronoun and reference questions (it, this, they, the former) require tracing back to the nearest compatible noun phrase, respecting number and gender.

Vocabulary-in-context items: substitute each option into the sentence and keep the one that preserves both meaning and register.

Negative factual questions ('which is NOT mentioned') are solved by verifying three options against the text and eliminating them.

Numbers and quantities: check whether the text gives an absolute figure, a proportion, or an approximation; options often change the unit or the comparison base.

Verb-form choice depends on time markers: 'since/for' plus present perfect, 'by the time' plus past perfect, 'this time next week' plus future continuous.

Conditionals: zero (present/present), first (present/will), second (past/would), third (had + participle/would have + participle); mixed conditionals pair a past condition with a present result.

Reported speech shifts tense one step back and adjusts time and place words (now→then, today→that day, here→there), except when the reporting verb is present tense.

Passive transformation keeps the original tense: 'They built it' → 'It was built'; progressive passive is 'is being + participle', perfect passive is 'has been + participle'.

Word formation: suffixes change class (-ness, -ment, -tion for nouns; -ful, -less, -ive, -able for adjectives; -ise/-ize, -en for verbs), prefixes usually negate (un-, in-, im-, dis-, mis-).

Collocation and preposition rules: depend on, insist on, be interested in, be good at, look forward to + -ing; 'used to + infinitive' (past habit) versus 'be used to + -ing' (accustomed).

Relative clauses: defining clauses take no comma and allow 'that'; non-defining clauses take commas and reject 'that'; 'whose' marks possession.

Modals of deduction: must (certain), may/might/could (possible), can't (impossible); 'must have + participle' expresses past certainty, 'needn't have + participle' a past action done unnecessarily.

Countability and agreement traps: 'news', 'information', 'advice', 'furniture' are uncountable; 'the number of' takes a singular verb, 'a number of' a plural one.

Standard task formats: an email or letter (formal or informal), an opinion essay, a for-and-against essay, a narrative, or a description based on a prompt; word limits are usually 120–150 or 150–180 words.

Paragraph architecture: introduction stating the topic and position, two body paragraphs each with one controlling idea plus support, conclusion restating without new arguments.

Cohesion devices to deploy deliberately: addition (furthermore, in addition), contrast (however, whereas, despite + noun/-ing), cause (therefore, as a result), exemplification (for instance, such as), sequencing (firstly, subsequently, finally).

Range is scored through varied structures: at least one conditional, one passive, one relative clause, and one impersonal construction ('It is widely believed that…') in a 150-word text.

Accuracy priorities in order of damage: verb tense and agreement, subject–verb inversion after 'not only', article use with countable singulars, and preposition collocations.

Register control: formal writing avoids contractions, phrasal verbs where a Latinate verb exists, exclamation marks, and direct address; informal emails use contractions and fixed openings (Hi X, / Best wishes,).

Punctuation rules that examiners penalise: no comma splice between independent clauses (use a full stop, semicolon, or conjunction); a comma before 'which' in non-defining clauses but not before 'that'.

Self-editing checklist: verify every verb has a subject and correct tense, every sentence has a main clause, and the word count stays within the stated limit.

Global questions ask for the main idea, the author's purpose, the tone, or the best title; the answer must cover the whole text, not one paragraph.

Locate the thesis in the first or last paragraph of an expository text; in argumentative texts it usually follows a contrast marker such as 'however' or 'yet'.

Reject options that are true but too narrow (one detail), too broad (beyond the text), or evaluative in a way the author never states.

Text-type recognition drives purpose: a review evaluates, a news report informs, an opinion column argues, a leaflet persuades or instructs.

Tone words to discriminate: neutral, ironic, critical, enthusiastic, nostalgic, cautious; irony is signalled by quotation marks, exaggeration, or an unexpected adjective.

Paragraph-function mapping: identify whether each paragraph narrates, exemplifies, contrasts, or concludes; the main idea usually spans the introduction and conclusion.

If two options seem right, prefer the one whose scope matches the number of paragraphs it summarizes; a single-paragraph option rarely answers a global question.

Structural signposts (however, therefore, in fact, on the whole) mark the author's stance more reliably than individual adjectives.

Specific-information questions are answered by locating a paraphrase, not a verbatim match; the correct option rewrites the source with synonyms or a changed word class.

Scan for proper nouns, numbers, dates, and capitalised terms first; they anchor the relevant line faster than reading sequentially.

Beware of distractors that copy the text's wording but invert it with 'not', 'only', 'always', or a swapped subject.

Inference answers must be logically forced by the text plus general knowledge; if the option needs an extra assumption, it is wrong.

Pronoun and reference questions (it, this, they, the former) require tracing back to the nearest compatible noun phrase, respecting number and gender.

Vocabulary-in-context items: substitute each option into the sentence and keep the one that preserves both meaning and register.

Negative factual questions ('which is NOT mentioned') are solved by verifying three options against the text and eliminating them.

Numbers and quantities: check whether the text gives an absolute figure, a proportion, or an approximation; options often change the unit or the comparison base.

Verb-form choice depends on time markers: 'since/for' plus present perfect, 'by the time' plus past perfect, 'this time next week' plus future continuous.

Conditionals: zero (present/present), first (present/will), second (past/would), third (had + participle/would have + participle); mixed conditionals pair a past condition with a present result.

Reported speech shifts tense one step back and adjusts time and place words (now→then, today→that day, here→there), except when the reporting verb is present tense.

Passive transformation keeps the original tense: 'They built it' → 'It was built'; progressive passive is 'is being + participle', perfect passive is 'has been + participle'.

Word formation: suffixes change class (-ness, -ment, -tion for nouns; -ful, -less, -ive, -able for adjectives; -ise/-ize, -en for verbs), prefixes usually negate (un-, in-, im-, dis-, mis-).

Collocation and preposition rules: depend on, insist on, be interested in, be good at, look forward to + -ing; 'used to + infinitive' (past habit) versus 'be used to + -ing' (accustomed).

Relative clauses: defining clauses take no comma and allow 'that'; non-defining clauses take commas and reject 'that'; 'whose' marks possession.

Modals of deduction: must (certain), may/might/could (possible), can't (impossible); 'must have + participle' expresses past certainty, 'needn't have + participle' a past action done unnecessarily.

Countability and agreement traps: 'news', 'information', 'advice', 'furniture' are uncountable; 'the number of' takes a singular verb, 'a number of' a plural one.

Standard task formats: an email or letter (formal or informal), an opinion essay, a for-and-against essay, a narrative, or a description based on a prompt; word limits are usually 120–150 or 150–180 words.

Paragraph architecture: introduction stating the topic and position, two body paragraphs each with one controlling idea plus support, conclusion restating without new arguments.

Cohesion devices to deploy deliberately: addition (furthermore, in addition), contrast (however, whereas, despite + noun/-ing), cause (therefore, as a result), exemplification (for instance, such as), sequencing (firstly, subsequently, finally).

Range is scored through varied structures: at least one conditional, one passive, one relative clause, and one impersonal construction ('It is widely believed that…') in a 150-word text.

Accuracy priorities in order of damage: verb tense and agreement, subject–verb inversion after 'not only', article use with countable singulars, and preposition collocations.

Register control: formal writing avoids contractions, phrasal verbs where a Latinate verb exists, exclamation marks, and direct address; informal emails use contractions and fixed openings (Hi X, / Best wishes,).

Punctuation rules that examiners penalise: no comma splice between independent clauses (use a full stop, semicolon, or conjunction); a comma before 'which' in non-defining clauses but not before 'that'.

Self-editing checklist: verify every verb has a subject and correct tense, every sentence has a main clause, and the word count stays within the stated limit.

Global questions ask for the main idea, the author's purpose, the tone, or the best title; the answer must cover the whole text, not one paragraph.

Locate the thesis in the first or last paragraph of an expository text; in argumentative texts it usually follows a contrast marker such as 'however' or 'yet'.

Reject options that are true but too narrow (one detail), too broad (beyond the text), or evaluative in a way the author never states.

Text-type recognition drives purpose: a review evaluates, a news report informs, an opinion column argues, a leaflet persuades or instructs.

Tone words to discriminate: neutral, ironic, critical, enthusiastic, nostalgic, cautious; irony is signalled by quotation marks, exaggeration, or an unexpected adjective.

Paragraph-function mapping: identify whether each paragraph narrates, exemplifies, contrasts, or concludes; the main idea usually spans the introduction and conclusion.

If two options seem right, prefer the one whose scope matches the number of paragraphs it summarizes; a single-paragraph option rarely answers a global question.

Structural signposts (however, therefore, in fact, on the whole) mark the author's stance more reliably than individual adjectives.

Specific-information questions are answered by locating a paraphrase, not a verbatim match; the correct option rewrites the source with synonyms or a changed word class.

Scan for proper nouns, numbers, dates, and capitalised terms first; they anchor the relevant line faster than reading sequentially.

Beware of distractors that copy the text's wording but invert it with 'not', 'only', 'always', or a swapped subject.

Inference answers must be logically forced by the text plus general knowledge; if the option needs an extra assumption, it is wrong.

Pronoun and reference questions (it, this, they, the former) require tracing back to the nearest compatible noun phrase, respecting number and gender.

Vocabulary-in-context items: substitute each option into the sentence and keep the one that preserves both meaning and register.

Negative factual questions ('which is NOT mentioned') are solved by verifying three options against the text and eliminating them.

Numbers and quantities: check whether the text gives an absolute figure, a proportion, or an approximation; options often change the unit or the comparison base.

Verb-form choice depends on time markers: 'since/for' plus present perfect, 'by the time' plus past perfect, 'this time next week' plus future continuous.

Conditionals: zero (present/present), first (present/will), second (past/would), third (had + participle/would have + participle); mixed conditionals pair a past condition with a present result.

Reported speech shifts tense one step back and adjusts time and place words (now→then, today→that day, here→there), except when the reporting verb is present tense.

Passive transformation keeps the original tense: 'They built it' → 'It was built'; progressive passive is 'is being + participle', perfect passive is 'has been + participle'.

Word formation: suffixes change class (-ness, -ment, -tion for nouns; -ful, -less, -ive, -able for adjectives; -ise/-ize, -en for verbs), prefixes usually negate (un-, in-, im-, dis-, mis-).

Collocation and preposition rules: depend on, insist on, be interested in, be good at, look forward to + -ing; 'used to + infinitive' (past habit) versus 'be used to + -ing' (accustomed).

Relative clauses: defining clauses take no comma and allow 'that'; non-defining clauses take commas and reject 'that'; 'whose' marks possession.

Modals of deduction: must (certain), may/might/could (possible), can't (impossible); 'must have + participle' expresses past certainty, 'needn't have + participle' a past action done unnecessarily.

Countability and agreement traps: 'news', 'information', 'advice', 'furniture' are uncountable; 'the number of' takes a singular verb, 'a number of' a plural one.

Standard task formats: an email or letter (formal or informal), an opinion essay, a for-and-against essay, a narrative, or a description based on a prompt; word limits are usually 120–150 or 150–180 words.

Paragraph architecture: introduction stating the topic and position, two body paragraphs each with one controlling idea plus support, conclusion restating without new arguments.

Cohesion devices to deploy deliberately: addition (furthermore, in addition), contrast (however, whereas, despite + noun/-ing), cause (therefore, as a result), exemplification (for instance, such as), sequencing (firstly, subsequently, finally).

Range is scored through varied structures: at least one conditional, one passive, one relative clause, and one impersonal construction ('It is widely believed that…') in a 150-word text.

Accuracy priorities in order of damage: verb tense and agreement, subject–verb inversion after 'not only', article use with countable singulars, and preposition collocations.

Register control: formal writing avoids contractions, phrasal verbs where a Latinate verb exists, exclamation marks, and direct address; informal emails use contractions and fixed openings (Hi X, / Best wishes,).

Punctuation rules that examiners penalise: no comma splice between independent clauses (use a full stop, semicolon, or conjunction); a comma before 'which' in non-defining clauses but not before 'that'.

Self-editing checklist: verify every verb has a subject and correct tense, every sentence has a main clause, and the word count stays within the stated limit.

Global questions ask for the main idea, the author's purpose, the tone, or the best title; the answer must cover the whole text, not one paragraph.

Locate the thesis in the first or last paragraph of an expository text; in argumentative texts it usually follows a contrast marker such as 'however' or 'yet'.

Reject options that are true but too narrow (one detail), too broad (beyond the text), or evaluative in a way the author never states.

Text-type recognition drives purpose: a review evaluates, a news report informs, an opinion column argues, a leaflet persuades or instructs.

Tone words to discriminate: neutral, ironic, critical, enthusiastic, nostalgic, cautious; irony is signalled by quotation marks, exaggeration, or an unexpected adjective.

Paragraph-function mapping: identify whether each paragraph narrates, exemplifies, contrasts, or concludes; the main idea usually spans the introduction and conclusion.

If two options seem right, prefer the one whose scope matches the number of paragraphs it summarizes; a single-paragraph option rarely answers a global question.

Structural signposts (however, therefore, in fact, on the whole) mark the author's stance more reliably than individual adjectives.

Specific-information questions are answered by locating a paraphrase, not a verbatim match; the correct option rewrites the source with synonyms or a changed word class.

Scan for proper nouns, numbers, dates, and capitalised terms first; they anchor the relevant line faster than reading sequentially.

Beware of distractors that copy the text's wording but invert it with 'not', 'only', 'always', or a swapped subject.

Inference answers must be logically forced by the text plus general knowledge; if the option needs an extra assumption, it is wrong.

Pronoun and reference questions (it, this, they, the former) require tracing back to the nearest compatible noun phrase, respecting number and gender.

Vocabulary-in-context items: substitute each option into the sentence and keep the one that preserves both meaning and register.

Negative factual questions ('which is NOT mentioned') are solved by verifying three options against the text and eliminating them.

Numbers and quantities: check whether the text gives an absolute figure, a proportion, or an approximation; options often change the unit or the comparison base.

Verb-form choice depends on time markers: 'since/for' plus present perfect, 'by the time' plus past perfect, 'this time next week' plus future continuous.

Conditionals: zero (present/present), first (present/will), second (past/would), third (had + participle/would have + participle); mixed conditionals pair a past condition with a present result.

Reported speech shifts tense one step back and adjusts time and place words (now→then, today→that day, here→there), except when the reporting verb is present tense.

Passive transformation keeps the original tense: 'They built it' → 'It was built'; progressive passive is 'is being + participle', perfect passive is 'has been + participle'.

Word formation: suffixes change class (-ness, -ment, -tion for nouns; -ful, -less, -ive, -able for adjectives; -ise/-ize, -en for verbs), prefixes usually negate (un-, in-, im-, dis-, mis-).

Collocation and preposition rules: depend on, insist on, be interested in, be good at, look forward to + -ing; 'used to + infinitive' (past habit) versus 'be used to + -ing' (accustomed).

Relative clauses: defining clauses take no comma and allow 'that'; non-defining clauses take commas and reject 'that'; 'whose' marks possession.

Modals of deduction: must (certain), may/might/could (possible), can't (impossible); 'must have + participle' expresses past certainty, 'needn't have + participle' a past action done unnecessarily.

Countability and agreement traps: 'news', 'information', 'advice', 'furniture' are uncountable; 'the number of' takes a singular verb, 'a number of' a plural one.

Standard task formats: an email or letter (formal or informal), an opinion essay, a for-and-against essay, a narrative, or a description based on a prompt; word limits are usually 120–150 or 150–180 words.

Paragraph architecture: introduction stating the topic and position, two body paragraphs each with one controlling idea plus support, conclusion restating without new arguments.

Cohesion devices to deploy deliberately: addition (furthermore, in addition), contrast (however, whereas, despite + noun/-ing), cause (therefore, as a result), exemplification (for instance, such as), sequencing (firstly, subsequently, finally).

Range is scored through varied structures: at least one conditional, one passive, one relative clause, and one impersonal construction ('It is widely believed that…') in a 150-word text.

Accuracy priorities in order of damage: verb tense and agreement, subject–verb inversion after 'not only', article use with countable singulars, and preposition collocations.

Register control: formal writing avoids contractions, phrasal verbs where a Latinate verb exists, exclamation marks, and direct address; informal emails use contractions and fixed openings (Hi X, / Best wishes,).

Punctuation rules that examiners penalise: no comma splice between independent clauses (use a full stop, semicolon, or conjunction); a comma before 'which' in non-defining clauses but not before 'that'.

Self-editing checklist: verify every verb has a subject and correct tense, every sentence has a main clause, and the word count stays within the stated limit.

Global questions ask for the main idea, the author's purpose, the tone, or the best title; the answer must cover the whole text, not one paragraph.

Locate the thesis in the first or last paragraph of an expository text; in argumentative texts it usually follows a contrast marker such as 'however' or 'yet'.

Reject options that are true but too narrow (one detail), too broad (beyond the text), or evaluative in a way the author never states.

Text-type recognition drives purpose: a review evaluates, a news report informs, an opinion column argues, a leaflet persuades or instructs.

Tone words to discriminate: neutral, ironic, critical, enthusiastic, nostalgic, cautious; irony is signalled by quotation marks, exaggeration, or an unexpected adjective.

Paragraph-function mapping: identify whether each paragraph narrates, exemplifies, contrasts, or concludes; the main idea usually spans the introduction and conclusion.

If two options seem right, prefer the one whose scope matches the number of paragraphs it summarizes; a single-paragraph option rarely answers a global question.

Structural signposts (however, therefore, in fact, on the whole) mark the author's stance more reliably than individual adjectives.

Specific-information questions are answered by locating a paraphrase, not a verbatim match; the correct option rewrites the source with synonyms or a changed word class.

Scan for proper nouns, numbers, dates, and capitalised terms first; they anchor the relevant line faster than reading sequentially.

Beware of distractors that copy the text's wording but invert it with 'not', 'only', 'always', or a swapped subject.

Inference answers must be logically forced by the text plus general knowledge; if the option needs an extra assumption, it is wrong.

Pronoun and reference questions (it, this, they, the former) require tracing back to the nearest compatible noun phrase, respecting number and gender.

Vocabulary-in-context items: substitute each option into the sentence and keep the one that preserves both meaning and register.

Negative factual questions ('which is NOT mentioned') are solved by verifying three options against the text and eliminating them.

Numbers and quantities: check whether the text gives an absolute figure, a proportion, or an approximation; options often change the unit or the comparison base.

Verb-form choice depends on time markers: 'since/for' plus present perfect, 'by the time' plus past perfect, 'this time next week' plus future continuous.

Conditionals: zero (present/present), first (present/will), second (past/would), third (had + participle/would have + participle); mixed conditionals pair a past condition with a present result.

Reported speech shifts tense one step back and adjusts time and place words (now→then, today→that day, here→there), except when the reporting verb is present tense.

Passive transformation keeps the original tense: 'They built it' → 'It was built'; progressive passive is 'is being + participle', perfect passive is 'has been + participle'.

Word formation: suffixes change class (-ness, -ment, -tion for nouns; -ful, -less, -ive, -able for adjectives; -ise/-ize, -en for verbs), prefixes usually negate (un-, in-, im-, dis-, mis-).

Collocation and preposition rules: depend on, insist on, be interested in, be good at, look forward to + -ing; 'used to + infinitive' (past habit) versus 'be used to + -ing' (accustomed).

Relative clauses: defining clauses take no comma and allow 'that'; non-defining clauses take commas and reject 'that'; 'whose' marks possession.

Modals of deduction: must (certain), may/might/could (possible), can't (impossible); 'must have + participle' expresses past certainty, 'needn't have + participle' a past action done unnecessarily.

Countability and agreement traps: 'news', 'information', 'advice', 'furniture' are uncountable; 'the number of' takes a singular verb, 'a number of' a plural one.

Standard task formats: an email or letter (formal or informal), an opinion essay, a for-and-against essay, a narrative, or a description based on a prompt; word limits are usually 120–150 or 150–180 words.

Paragraph architecture: introduction stating the topic and position, two body paragraphs each with one controlling idea plus support, conclusion restating without new arguments.

Cohesion devices to deploy deliberately: addition (furthermore, in addition), contrast (however, whereas, despite + noun/-ing), cause (therefore, as a result), exemplification (for instance, such as), sequencing (firstly, subsequently, finally).

Range is scored through varied structures: at least one conditional, one passive, one relative clause, and one impersonal construction ('It is widely believed that…') in a 150-word text.

Accuracy priorities in order of damage: verb tense and agreement, subject–verb inversion after 'not only', article use with countable singulars, and preposition collocations.

Register control: formal writing avoids contractions, phrasal verbs where a Latinate verb exists, exclamation marks, and direct address; informal emails use contractions and fixed openings (Hi X, / Best wishes,).

Punctuation rules that examiners penalise: no comma splice between independent clauses (use a full stop, semicolon, or conjunction); a comma before 'which' in non-defining clauses but not before 'that'.

Self-editing checklist: verify every verb has a subject and correct tense, every sentence has a main clause, and the word count stays within the stated limit.