IELTS Matching Features who said what, and who did not

What the task asks

You are given a box of options, usually labelled A to E, and a set of statements. You decide which option each statement belongs to. The instruction will tell you whether options may be used more than once, and it usually says they may.

The method

1. 01

   Scan for the options in the passage and mark them

   The options are usually proper nouns, so they are easy to find. Marking every occurrence first turns the task into reading around known points rather than searching the whole passage repeatedly.

2. 02

   Work from the passage, not the statements

   Take each marked name and read what is said about it, then look for the statement that matches. Going the other way means re-scanning for every statement.

3. 03

   Expect repeats and gaps

   One researcher may account for three statements and another for none. A neat one-to-one distribution is the exception, so do not force one.

4. 04

   Ignore question order

   Unlike most reading tasks, these do not follow the passage. The answer to the last question can be in the first paragraph.

Worked examples

Real questions from the practice bank with the reasoning written out. Cover the answer, decide, then read the explanation. Getting it right for the wrong reason is the thing that does not survive test day.

Where the marks go

Can a letter be used more than once in matching features?

Yes, and this is the difference between a good score on this question type and a bad one. The official IELTS Academic Reading format page states that it is possible that some options will not be used, and that others may be used more than once.

The instruction line above the set is where that is decided, and it is not boilerplate. It is the rule for that particular set, and it changes what a correct answer sheet can look like. Across the 81 matching features sets in this practice bank, 25 print an explicit note that you may use any letter only once. The other 56 print no such restriction.

Now the part that matters. In 40 of those 81 sets, just under half, at least one letter is genuinely the answer to more than one statement. Every single one of those 40 sits in the group of 56 whose instruction did not forbid it. Not one set here contradicts its own instruction line.

So the three seconds spent reading that line is the highest return three seconds available on this question type. It tells you whether crossing off is legal, and crossing off when it is not legal does not cost you one mark. It forces a wrong answer somewhere else in the set, because you have removed the correct letter from your own pool.

The reverse mistake is rarer but real: refusing to reuse a letter even when the instruction permits it, because a neat one to one grid looks more like a finished answer sheet. One researcher accounting for three statements while another accounts for none is normal here, not a sign you have gone wrong.

How to tell matching features from matching information

These two are confused more than any other pair in the Reading paper, and the reason is mechanical rather than conceptual. Both ask you to write a single capital letter in a box. On a marked answer sheet they are indistinguishable.

The test that settles it in two seconds is to ask what the letters stand for. In matching features the letters live in a printed box above the questions, with a name, a study, a country, a theory or a category next to each one. In matching information the letters are the paragraph labels of the passage itself, printed down the margin, and no box is given because the passage is the option list.

The conceptual version of the same difference is that matching features asks whose and matching information asks where. A statement about a claim regarding urban soil quality is a matching features question if you have to decide which of five researchers made it, and a matching information question if you have to decide which of nine paragraphs contains it.

The methods diverge from there. On matching features the options are usually proper nouns, so you can find and mark all of them in the passage before reading a single statement closely, which turns the whole set into reading around known points. On matching information there are no proper nouns to anchor on, so you hunt for the shape of the information instead: a reason reads as because or led to, an example reads as for instance or such as.

What they share is the thing that makes both slow. Neither follows passage order, so neither gives you the positional narrowing that carries you through True, False, Not Given. Both are worth leaving until the other sets on that passage are done, because by then you have read most of the text.

Why passage order does not help, and what to use instead

Matching features questions do not run in passage order. The answer to the last statement in the set can sit in the first paragraph. Searching question by question, top to bottom, means re-scanning the whole passage once per statement, which is how a five question set eats fifteen minutes.

The replacement signal is capitalisation. The options in this question type are overwhelmingly proper nouns, and capital letters mid sentence are visible when you are skimming far too fast to read. Finding and marking every occurrence of every option first is cheap, and it converts an unordered search into a short list of known locations.

The second signal is the pronoun chain, and it is where careful candidates lose marks rather than careless ones. A passage names a researcher once and then says she, the team, or the study for the next four sentences. Those sentences still belong to that researcher, and a statement paraphrasing the third of them is still her answer.

The practical rule that follows: when you mark a name, read forward until the next proper noun appears. That is where the chain ends. Stopping at the sentence containing the name is the single commonest way to miss an answer that was sitting two lines below the thing you highlighted.

The last signal is attribution against topic. Two researchers in the same passage work on the same subject, or they would not both be in it. The question is never what the statement is about. It is who said it, and the giveaway words are usually argued, found, disputed, proposed or concluded rather than any of the topic vocabulary.

How often does matching features appear?

More often than either of its siblings, at least in this bank. Of the 81 full papers here, 80 carry a matching features set, and the type accounts for 383 questions in total. Sets run from 3 to 7 questions and average just under 5.

The track split is the interesting part, because it runs opposite to matching headings. Matching features questions here divide 242 on General Training papers against 141 on Academic ones, and all 40 General Training papers carry a set against 40 of the 41 Academic papers.

Set against matching headings, where 244 of 280 questions sit on Academic papers and only 6 of 40 General Training papers carry the type at all, the revision advice almost inverts by track. A General Training candidate who spends an evening on matching headings and none on matching features has optimised backwards.

As always, this is the distribution of this practice bank rather than of the real exam, because the test owners do not publish theirs. Treat it as a guide to where practice time earns most, not as a prediction about your paper.

The way to test whether the method has actually stuck is to meet a set cold, inside a full paper, with the clock running and the other question types competing for the same hour. All 81 reading tests are set up that way, with a written explanation on every answer.

Matching features against the other two matching types

The fastest way to keep the three apart is to ask what the letters stand for. Counts are from this practice bank of 81 papers, not from the real exam.

Frequently asked questions