CELPIP Reading Part 3: Paragraph Matching Strategy

Part 3 of the Reading test does something the other parts don't. You read one passage, then you have to remember which corner of it each idea came from. Question 4 might be in paragraph B. Question 7 in paragraph D. Question 2 might not be in the passage at all.

That last possibility is what separates Part 3 from everything else on the test. Option E. Not given.

Most candidates lose marks here, not because they didn't understand the text, but because they couldn't remember where they read what. Below: the 30-second text map, paraphrase recognition, the option E rule, and a timing budget that gets you through all nine questions in ten minutes.

The official name is Reading for Information. You get one passage of roughly 350 to 450 words, split into four labeled paragraphs (A, B, C, D). Topics: lean educational: a short biography, a natural phenomenon, the lifecycle of a Canadian songbird, the history of a public institution. Formal voice. No opinions. Just facts.

Below the passage sit nine numbered questions. Each one is a single statement of information. Your task: pick the paragraph it came from. Five options per question: A, B, C, D, or E.

Option E means the information in the question doesn't appear in any of the four paragraphs. Sounds simple. It isn't.

The trap works like this. A question mentions something that feels connected to the passage. The topic is right. The vocabulary overlaps. One or two surrounding details even line up. But the specific claim in the question? It's never actually stated. Your brain wants to assign the question to a paragraph because the vibe matches. That's how you lose the mark.

The rule that saves you

If the exact claim isn't stated, it isn't there. Not given is not the same as suggested, hinted, or probably true. The paragraph either says the thing or it doesn't.

One habit worth building. When you can't find a question's home paragraph after a 30-second search, suspect E before you start guessing. Test takers who never pick E almost always lose two to three questions per Part 3 to this trap alone.

Before you read a single question, run a fast pass over the passage and label each paragraph with two or three words. Use the noteboard. Labels don't have to be elegant. They have to be specific enough that when a question mentions a name, a year, or a stage in a process, you know which paragraph to start in.

Good labels for a passage on the history of Canadian rail look like this:

• A = early proposals, 1850s

• B = construction, Cdn Pacific

• C = workers, conditions

• D = legacy, modern impact

Bad labels are vague. A = intro, B = middle, C = more, D = end. Useless. You'll still have to re-skim every paragraph from scratch.

The first one or two sentences of each paragraph almost always announce the topic. That's where your label comes from. You're not reading the paragraph in detail. You're tagging it. Total time for the map: 60 to 90 seconds, and it pays itself back four times over once you start matching.

Questions almost never reuse the exact words from the passage. Same meaning, different vocabulary. If you're hunting for a word match, you'll miss most of your answers.

What to do instead. Read the question, restate the claim in your own words, then search your text map for the paragraph that talked about that idea.

The passage says: Production halted in 1948 when raw materials became scarce.

The question says: Manufacturing stopped due to a shortage.

Production becomes manufacturing. Halted becomes stopped. Raw materials became scarce, leading to a shortage.

The passage says: The species relies on dense forest cover for nesting.

The question says: The bird needs thick woodland to breed.

The species becomes a bird. Relies on becomes needs. Dense forest cover becomes thick woodland.

Build the habit of recognizing synonyms, verb-noun swaps, and abstract-to-concrete restatements. That's most of what Part 3 actually tests.

1. Word matching instead of meaning matching

You see the word recession in paragraph C and the word recession in question 5, so you pick C. But the question is about a different consequence of the recession. Same vocabulary, different claim.

2. Avoiding (or overusing) option E

Many candidates go through nine questions without picking E once, even though most sets contain at least one E answer. The opposite mistake is just as costly: slapping E on every question you can't immediately find. Match all four paragraphs first, then trust E only when nothing fits.

3. Spending three minutes on one question

One stubborn question can eat up your whole time budget. If a question fights you for more than 75 seconds, mark a guess, flag it on the noteboard, and come back if there's time. A wild guess is always better than a blank.

• 60-90 seconds: build the text map, label A-D

• 3-4 minutes: answer the 5-6 easy matches first

• 3-4 minutes: work the harder questions, decide on E if needed

• 30-60 seconds: review flags and fill any blanks with a best guess

Two principles drive this split. CELPIP doesn't penalize wrong answers, so a guess at 25% odds beats nothing. And the text map earns its keep on questions 4 through 9, where your eyes go straight to the right paragraph instead of scanning from scratch.