Reading Comprehension: How to Find the Main Idea and Evidence

Read for the Author's Purpose

A passage is not a collection of equally important sentences. Before looking at answer choices, ask what the author is trying to do: explain a process, defend a position, compare approaches, describe a problem, or evaluate a proposed solution. Write that purpose in one phrase — 'defends X', 'compares X and Y', 'explains how X works' — before you touch the questions. Learners who skip this step tend to answer each question as an isolated puzzle instead of checking it against what the passage is actually trying to accomplish.

The main idea is usually broad enough to cover the whole passage but specific enough to distinguish it from a topic label. 'Urban transport' is a topic. 'Reliable public transport can improve access to work when cities combine frequency with affordable fares' is a possible main idea — it names the topic, the author's position, and the condition attached to it.

A quick test for whether you've actually found the main idea: could every paragraph in the passage be read as either support for it, a complication of it, or a response to an objection against it? If a paragraph doesn't fit any of those three roles, you've probably misread the purpose, or that paragraph is doing something narrower — background, a transition, or a single example — that shouldn't be mistaken for the passage's core claim.

Separate Claims, Evidence, and Examples

A claim is an idea the author wants you to accept. Evidence supports it through data, observations, research, or reasoning. An example illustrates the point but may not prove it on its own. Marking these roles while reading makes questions about tone and structure easier, and it directly prevents one of the most common wrong-answer traps: an option that restates an example from the passage as if it were the author's main claim. Examples are almost always narrower than the point they're illustrating, so an answer choice that only covers the example is usually too specific to be correct for a main-idea question.

Pay attention to transitions. Words such as however, therefore, although, for example, and in contrast reveal relationships between ideas. The sentence after a contrast often contains the author's qualification or most important shift — exam passages frequently build toward exactly that sentence, then test whether you caught it.

A concrete way to practise this: take a passage you've already read, and without looking back at it, list every transition word in order from memory. If you can reconstruct roughly where the argument turned, you were tracking structure, not just reading words. If you can't, that's useful diagnostic information — it means transitions are sliding past you, which is worth deliberately slowing down for on the next passage.

Answer Inference Questions Carefully

An inference must be supported by the passage even when it is not stated word for word. Eliminate options that introduce outside knowledge, extreme language, or a claim broader than the evidence. A plausible statement is not necessarily a valid inference — the test isn't 'could this be true in general,' it's 'does this passage, specifically, support it.'

Extreme-language options ('always', 'never', 'proves', 'completely') fail more often than they pass, simply because passages are usually careful to qualify their claims, and an inference that goes further than the source text almost always oversteps. When two options both sound reasonable, check which one requires you to add information the passage never gave you — that's the one to eliminate.

For vocabulary-in-context questions, replace the word with each option and read the sentence again. The correct meaning is the one that fits the author's logic, not necessarily the most common dictionary meaning. A word like 'critical' can mean 'essential' or 'disapproving' depending entirely on the sentence around it — the passage decides which one you need, not a general definition.

A Timed Practice Routine

For one week, practise one passage without a timer and write a one-sentence summary before answering questions. On the next week, add a moderate time limit and review why each wrong option was tempting. Record whether the problem was comprehension, vocabulary, rushing, or evidence tracking — these four categories call for different fixes, so lumping every wrong answer together as 'I need to read faster' usually treats the wrong problem.

Do not reread the full passage for every question. Return to the relevant paragraph, locate the evidence, and compare the wording of the option with the wording and scope of the text. If you find yourself rereading the whole passage more than once per question set, that's a sign your first pass isn't building a strong enough map of where each idea lives — try annotating paragraph purposes in the margin on your next attempt (one phrase per paragraph) so you can navigate back directly instead of re-scanning from the top.

A caveat: reading speed develops through regular reading of varied, well-edited material. Speed without accurate evidence tracking usually lowers performance — a learner who finishes fast but can't point to the sentence that supports their answer is optimising for the wrong thing.