The Pitfalls of Confusing Correlation and Causation on GMAT Critical Reasoning Questions

causation goslingIn Stephen Pinker’s book, The Blank Slate, there’s an entertaining discussion illustrating the pitfalls of confusing correlation and causation. Pinker cites an old Russian folktale in which a Tsar discovers that, of his many provinces, the one that has the highest disease rate also has the most doctors. So he orders all the doctors killed. I’ll often make reference to this passage when I’m teaching Critical Reasoning because the absurdity of the argument is immediately apparent. Just because two variables are correlated, it doesn’t mean that one is necessarily causing the other.

Causality arguments show up frequently on the GMAT and they can be quickly encapsulated with a simple arrow diagram. So the above discussion involving the Tsar could be depicted on scratch paper like so:

Doctors –> Disease

x –> y

Typically, if we need to weaken one of these arguments, we’ll do so in one of two ways. First, it’s possible that cause and effect are reversed. Here it would mean that the disease was causing the doctors to come to the province. In arrow diagram form, it would look like this:

Disease –> Doctors

y –> x

Secondly, there may be a different underlying cause. In the case of our folktale, maybe it’s the case that poor sanitation is causing the disease.

Poor sanitation –> Disease

z –> y

To summarize: whenever we see a causality argument that needs to be weakened, we can distill it into an arrow diagram and then search for one of the two above scenarios.

Here’s an example from the Official Guide:

In the last decade there has been a significant decrease in coffee consumption.  During this same time, there has been increasing publicity about the adverse long-term effects on health from the caffeine in coffee.  Therefore, the decrease in coffee consumption must have been caused by consumers’ awareness of the harmful effects of caffeine.

Which of the following, if true, most seriously calls into question the explanation above?

A. On average, people consume 30% less coffee than they did 10 years ago.
B. Heavy coffee drinkers may have mild withdrawal symptoms, such as headaches, for a day or so after, significantly decreasing their coffee consumption.
C. Sales of specialty types of coffee have held steady, as sales of regular brands have declined.
D. The consumption of fruit juices and caffeine-free herbal teas has increased over the past decade.
E. Coffee prices increased steadily in the past decade because of unusually severe frosts in coffee-growing nations.

This one is straightforward enough to diagnose – we actually get the phrase “caused by” in the argument! As an arrow diagram, it looks like this:

Awareness of harmful effects of caffeine –> decrease in coffee consumption

According to our earlier analysis, this can be weakened in one of two ways. If cause and effect were reversed, the diagram be:

Decrease in coffee consumption –> Awareness of harmful effects of caffeine

Well, that doesn’t make sense. How could a decrease in coffee consumption cause a heightened awareness of the ill effects of caffeine? So we must be looking for an alternative cause:

Something else –> decrease in coffee consumption.

So that’s what we’re after: that alternative underlying cause.

A. On average, people consume 30% less coffee than they did 10 years ago.

There’s no different underlying cause here. In fact, this is reiterating the notion that coffee consumption has decreased. We already knew this. Eliminate A.

B. Heavy coffee drinkers may have mild withdrawal symptoms, such as headaches, for a day or so after, significantly decreasing their coffee consumption.

This isn’t an alternative reason for why people are drinking less coffee. In fact, the unpleasant withdrawal symptoms would be a pretty compelling reason to continue drinking plenty of coffee! Eliminate B.

C. Sales of specialty types of coffee have held steady as sales of regular brands have declined.

Again, no real alternative cause presented here. And, logically, this doesn’t weaken the argument at all. It’s certainly possible that while many coffee drinkers have cut back on their coffee consumption, the kind of aficionados who drink specialty coffee will continue to drink their double latte espressos without reservation. Eliminate C.

D. The consumption of fruit juices and caffeine-free herbal teas has increased over the past decade.

This one is often tempting. Students sometimes argue that it’s the appeal of fruit juices that is the alternative underlying cause we’re looking for. The problem is that we’re trying to weaken the argument, and this answer choice really isn’t incompatible with the conclusion. To see why, imagine that the argument is true: people find out that caffeine is bad for them, and so drink less coffee. It would be perfectly reasonable for them to then replace that morning coffee with alternatives like fruit juice and herbal tea. In other words, the increase in the consumption of other beverages wouldn’t be a cause of the decrease in coffee, but rather, a consequence of that decrease. D is out.

E. Coffee prices increased steadily in the past decade because of unusually severe frosts in coffee-growing nations.

Now we have our alternative cause. Perhaps it’s not the awareness of the ill effects of caffeine that’s caused this drop in coffee consumption, it’s an increase in price. The new arrow diagram looks like this:

Increase in price –> Decrease in consumption

And this makes perfect sense. E is our answer.

The takeaway: A simple arrow diagram can powerfully simplify the logic of any causality argument.

* Official Guide® question courtesy of the Graduate Management Admissions Council.

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!

By David Goldstein, a Veritas Prep GMAT instructor based in Boston. You can find more articles by him here