How to Solve Inference Questions on the GMAT

On the GMAT, there is often a fine line between a statement possibly being true and a statement always being true. Inference questions ask about which statement must be true, and often provide many statements that each seem to be correct. However, must be true is a high standard to achieve, and many statements fall short of this benchmark despite being perfectly reasonable assumptions on their own.

The fact that the standard is so high does allow us an easy way to verify any statement provided. If you can find one single counter-example that does not have to be true, then the entire statement does not have to be true. Even if 99.44% of the time the statement works, it does not have to be true because we can find at least once instance where it is not true.

The point of inference questions is that they must be true based on some piece of evidence in the text. The exam will rarely ask you about a universal truth (say: The Earth turns around the Sun) but rather will ask you to infer something that must because of what is written in the passage. This is why the simplest examples are just verbatim transcriptions of what’s in the text, or simple counterfactuals that must logically be true.

Sometimes, it can be easier to think of this relationship in the inverse direction. If the passage guarantees the veracity of one of the statements, does that statement also guarantee the veracity of the passage? Not necessarily, but it could easily be.

Let’s demonstrate with two simple examples. Firstly: John slept badly, so John is tired this morning. If John is tired this morning, does that mean he slept badly last night? Not necessarily. Perhaps he slept fine but forgot his coffee. Now think of the example: Ron is taller than Tom, therefore Tom is shorter than Ron. If Tom is shorter than Ron, does that mean that Ron is taller? In this case, yes it does. There is no possible other explanation that will be consistent with the statement.

It’s often beneficial to think of the answer choices in this commutative way. After determining whether the answer choice has to follow the passage, we can verify whether the passage follows from the passage as well. Let’s look at an example:

All of John’s friends say they know someone who has smoked 40 cigarettes a day for the past 40 years and yet who is really fit and well. John does not know anyone like that and it is quite certain that he is not unique among his friends in this respect.

If the statements in the passage are true, then which one of the following must also be true?

(A) Smokers often lie about how much they smoke.
(B) People often knowingly exaggerate without intending to lie.
(C) All John’s friends know the same lifelong heavy smoker.
(D) Most of John’s friends are not telling the truth.
(E) Some of John’s friends are not telling the truth.

Now instead of using our usual approach of looking at each answer choice and seeing whether it fits with the stimulus, we can reverse things and look at the answer choice and determine whether the stimulus logically follows! (It’s backwards-land!)

A)  Smokers often lie about how much the smoke: While this is often true, would this fact guarantee that the stimulus necessarily follows? Not really. For starters, smokers may lie about how much they smoke by understating or overstating the number of cigarettes they smoke. If they’re revising the number downwards, then maybe everyone’s been smoking 40+ cigarettes for the past 4 decades. Or maybe no one has. If this statement were true, the stimulus would not necessarily follow.

B)  People often knowingly exaggerate without intending to lie: Another statement that seems to be true, but doesn’t really help in determining whether the lies are overstating or understating reality. This is like answer choice A, but even more general, so therefore it cannot be the correct selection.

C)  All of John’s friends know the same lifelong heavy smoker: This one starts getting a little more conceptual. If this were true, and all of John’s friends know the same chain-smoker (who John has miraculously avoided for about a half century), then John would not be able to say that he is unique among his friends. In fact, he would likely be well aware that all of his friends are referring to the same individual. This contradicts John’s belief that he is not unique among his friends, and therefore cannot be the right choice either.

D)  Most of John’s friends are not telling the truth: This is an interesting one. If most of John’s friends are not telling the truth, then John would be correct in thinking he wasn’t alone among his friends. This answer seems to support the stimulus, however it goes too far. If even a single one of John’s friends were lying, then the stimulus would still follow. There’s no need for a 50%+1 majority of his friends. This answer is close, but the stimulus can be true without this choice being true.

E)  Some of John’s friends are not telling the truth: This is the best answer, as “some” means non-zero, which implies that a single one of John’s friends can be fibbing and the stimulus would still follow. “Most” is too strong, “some” will guarantee that the stimulus follows, and vice versa.

On questions that ask you what must be true, there can be no margin for error. As a single counterexample can sink your entire argument, the statement must be true in all situations. The bidirectional nature of the data can help clue you in on which answer choice to select as well, as the veracity of one must imply the veracity of the other. On the GMAT (as in music), it’s important to not always go in One Direction.

Plan on taking the GMAT soon? We have GMAT prep courses starting all the time. And, be sure to find us on Facebook and Google+, and follow us on Twitter!

Ron Awad is a GMAT instructor for Veritas Prep based in Montreal, bringing you weekly advice for success on your exam.  After graduating from McGill and receiving his MBA from Concordia, Ron started teaching GMAT prep and his Veritas Prep students have given him rave reviews ever since.