Unlike Listening, Reading gives you no transfer time. Sixty minutes is everything you get, and that single fact shapes every tactic worth having.
The fact that governs Reading is the mirror image of the one that governs Listening:
There is no transfer time.
In Listening, the paper test gives you ten extra minutes to copy answers onto the answer sheet. Reading gives you none. Those sixty minutes have to cover reading, answering and filling in the sheet. A great many marks are lost not to wrong answers but to answers never written down.
So time allocation is a skill to train, on a level with any question type.
Three passages, forty questions, sixty minutes. Difficulty rises across the three, so the working frame is 15 – 20 – 25 minutes.
More important than the frame: every question is worth exactly one mark. A brutal Matching Headings item is not worth more than an easy gap-fill. When you fall behind, abandon whatever is eating the clock, collect the easy marks, and return if time allows.
A flat list of fourteen is hard to hold and useless in the room. Grouping them by how they behave is far more usable.
The answers appear in the text in the same order as the questions. Tactically this is the easiest family, because you never have to read backwards.
1. Sentence completion. 2. Summary completion. 3. Table completion. 4. Flow-chart completion. 5. Diagram completion. 6. Short answer.
How to practise: this family is won on two things — recognising paraphrase, and respecting the word limit. Before you hunt, decide what kind of word the gap needs and where in the argument the gap sits.
Shared trap: exceeding the word limit, or miscopying a word that is printed right there in the passage. Both are marked wrong however well you understood.
Answers are scattered. This is the family that eats time, and where candidates get bogged down.
7. Matching headings. 8. Matching information — find the paragraph containing X. 9. Matching features — attach characteristics to people or things. 10. Matching sentence endings.
How to practise: do this family last within each passage, even when it is printed first. By then you have been through the text once and roughly know what lives where.
The particular trap in Matching Headings: people match the heading to the most striking detail instead of the paragraph's main idea. A paragraph may spend three sentences on a memorable example, but that example only serves an argument sitting in the first or last sentence. The paper always includes a bait heading matching the detail. The question to ask is: what does this paragraph exist to do within the whole passage?
11. True / False / Not Given — about facts. 12. Yes / No / Not Given — about the writer's views. 13. Multiple choice. 14. List of options, and choosing a title for the passage.
The single most important line in the paper lives here: False means the passage says the opposite; Not Given means the passage does not carry enough to decide — even when the statement sounds entirely reasonable given what you know.
The test: if you have to use knowledge from outside the passage to decide, the answer is Not Given. The exam asks about the text, not about the world.
The difference between the two sets: True/False/Not Given asks about facts; Yes/No/Not Given asks about the writer's opinion. Using the wrong set of words loses the mark even when the reasoning was right.
Skimming is fast reading for the main idea — for Matching Headings, author's purpose, choosing a title.
Scanning is sweeping for specific information: proper nouns, figures, technical terms — for completion, True/False/Not Given, Matching Information.
The commonest mistake is reading every question type the same way, usually reading everything carefully, and running out of time in passage three.
Only stop for words that decide the answer — words inside the sentence carrying the information asked about. For those, infer from context: what part of speech, is the tone positive or negative, what do the neighbours suggest.
Every other unfamiliar word: skip it without mercy. High-band readers do not know as many more words as you think — they are better at knowing which words they do not need to know.
For the long-run foundation, Hu and Nation (2000) put comfortable reading at about 98% known words, and Nation (2006) estimates eight to nine thousand word families for reading without lookups. There is no shortcut to those numbers except reading widely, for a long time.
Reading is the section where the computer format helps most, and where the fewest candidates take advantage of it.
The timing is identical: three passages, forty questions, sixty minutes, and no transfer time in either format. What changes is the interface, and it comes with three tools worth learning before test day rather than discovering during it.
The split screen. The passage sits on the left, the questions on the right, and both scroll independently. You never lose your place turning a page, and you never hunt for the paragraph you were reading. For Matching Information in particular — the type that sends you back through the whole passage — this is a real advantage.
Highlighting and notes. You can highlight any stretch of the passage and attach a note to it. Used properly this replaces the margin annotation that paper candidates rely on: mark the paragraph topic as you skim, and Matching Headings becomes a matter of reading your own labels rather than rereading the text.
Copy and paste. For every completion task — sentence, summary, note, table, flow-chart — you can copy the exact words from the passage into the answer box. This removes an entire category of error at a stroke. Candidates lose marks on paper for miscopying a word they had correctly identified; on computer that cannot happen, provided you paste rather than retype.
And the thing you lose: the physical page. Skimming a screen is not the same motion as skimming paper, and scrolling makes it harder to hold a sense of where you are in a long passage. That is worth practising deliberately rather than assuming it will transfer.
One practical note that has nothing to do with skill: results for the computer format arrive in one to five days, against thirteen for paper. If a deadline is close, that difference may decide the format for you.
Two timed passages a week, on the real 15–20–25 frame, and with the answers actually recorded — on the answer sheet if you sit the paper test, on screen if you sit it on computer. Either way that recording happens inside the hour; if you have never practised it there, you will discover the problem on test day.
One single-type session, usually Matching Headings or True/False/Not Given — the two that bleed the most marks.
Extensive reading daily, twenty minutes, untimed, no questions. Nakanishi (2015) meta-analysed extensive reading research and found consistent positive effects on both reading speed and vocabulary. This is the foundation work that drilling papers cannot replace.
Review by cause: for each wrong answer, was it an unknown word, a missed paraphrase, a misunderstanding of what the question type demands, or simply running out of time? Those four need four different repairs.
The same article, told in plain words — for younger readers, or for anyone who wants the point quickly.
IELTS Reading has fourteen question types. They fall into three families, and knowing which family you are in changes how you should read.
Sentence, summary, table, flow-chart and diagram completion, plus short answer. The answers appear in the passage in the same order as the questions, so you never have to read backwards. Tactically this is the easiest family.
How to practise: this family is won on two things — recognising paraphrase, and respecting the word limit. Before you hunt, decide what kind of word the gap needs.
Shared trap: going over the word limit, or miscopying a word that is printed right there in the passage. Both are marked wrong however well you understood.
Matching headings, matching information, matching features, matching sentence endings. The answers are scattered. This is the family that eats time.
How to practise: do this family last within each passage, even when it is printed first. By then you have been through the text once and roughly know what lives where.
The trap in Matching Headings: people match the heading to the most striking detail instead of the paragraph's main idea. A paragraph may spend three sentences on a memorable example, but that example only serves an argument sitting in the first or last sentence — and the paper always includes a bait heading matching the detail.
The question to ask is: what does this paragraph exist to do within the whole passage?
True/False/Not Given, Yes/No/Not Given, multiple choice, list of options.
The single most important line in the whole paper lives here: False means the passage says the opposite. Not Given means the passage does not carry enough to decide — even when the statement sounds entirely reasonable.
Only stop for words that decide the answer — words inside the sentence carrying the information asked about. For those, infer from context.
Every other unfamiliar word: skip it without mercy. High-band readers do not know as many more words as you think. They are better at knowing which words they do not need to know.
For the long run: comfortable reading sits around 98% known words, and reading without lookups needs something like eight to nine thousand word families. There is no shortcut to those numbers except reading widely, for a long time.
Reading is where the computer format helps most, and where fewest candidates use the advantage. The split screen — passage left, questions right, scrolling independently — means you never lose your place, which matters most for Matching Information. You can also highlight and take notes.
What you lose is the physical page. Skimming a screen is not the same motion as skimming paper, and scrolling makes it harder to hold a sense of where you are in a long passage. Practise that deliberately rather than assuming it transfers.
These articles summarize well-established research in learning science and linguistics. Key sources and further reading:
The recording plays exactly once. That single fact governs the whole paper, and every technique worth learning is a cons...
The three parts are not three difficulty levels of the same task. They test different things — and the answer length tha...
Task 2 counts twice as much as Task 1. So the most consequential decision you make is not a good phrase — it is stopping...