The biggest misconception about Critical Reasoning questions on the GMAT® is that they are based on common sense. Well, nothing could be farther from the truth. In fact, CR questions are based on rules of formal logic. The question from the Official Guide (OG) below is a perfect illustration.

A study of marital relationships in which one partner’s sleeping and waking cycles differ from those of the other partner reveals that such couples share fewer activities with each other and have more violent arguments than do couples in a relationship in which both partners follow the same sleeping and waking patterns. Thus, mismatched sleeping and waking cycles can seriously jeopardize a marriage.

Which of the following, if true, most seriously weakens the argument above?

(A) Married couples in which both spouses follow the same sleeping and waking patterns also occasionally have arguments that can jeopardize the couple’s marriage.
(B) The sleeping and waking cycles of individuals tend to vary from season to season.
(C) The individuals who have sleeping and waking cycles that differ significantly from those of their spouses tend to argue little with colleagues at work.
(D) People in unhappy marriages have been found to express hostility by adopting different sleeping and waking cycles from those of their spouses.
(E) According to a recent study, most people’s sleeping and waking cycles can be controlled and modified easily.

The question seems like a real-life situation that has nothing to do formal logic; test-takers typically tend to choose option A. Not if you start reading CR questions differently.

In the CR passage above, a study finds that two things are correlated: mismatched sleeping and waking cycles (X) and marital discord (Y). Based on this, the research concludes that X is causing Y.

According to rules of formal logic, correlation does not imply causation no matter how high the degree of correlation. For example, out of 100 people who have more than 3 cups of coffee a day, 99 might also report having migraines. But this cannot be used to conclude that the coffee is causing the migraines.

X and Y are correlated does not mean X is causing Y since

A. there is no evidence to prove that the direction of causation is from X to Y, it also be from Y to X?
B. there can be a different reason, Z, for the occurrence of Y.

So, for any question where the argument concludes causation based on correlation, the correct option that weakens the conclusion would show that direction of causation is reverse or Y is causing X. In this question option D does the same. It shows that it is not sleeping and waking up at different times that results in couples having violent arguments. It is because they are already having trouble in the marriage that couples choose to sleep and wake up at different times.

Solving this question based on common sense would involve going through all five options. But solving it using formal logic would mean that you always look for reverse causation whenever a weaken question-type argument is based on correlation-causation.

Option A is also incorrect based on the rules of formal logic. The argument says that mismatched sleeping and waking cycles (X) cause and marital discord/violent arguments (Y) or in formal terms All X is Y. In formal terms Option A says, there is Y without X.

This does not weaken an argument of the type All X is Y, since the argument does not claim that Only X is Y, people can have arguments even because of other reasons. If the argument claims that different sleeping and waking patterns are the ONLY reason for violent arguments then option A is the answer.

Out of the 12-14 CR questions on the Verbal Section of the GMAT® around 5 -7 will be of the weaken type, and one of them will have a correlation-causation argument. A good way of identifying such questions is to see if the passage mentions a study or research. Once you identify such arguments, all you need to do is look for a reverse causation and you can solve the question in under a minute!

We will do a few more examples of this type in the next post.