Correlation to Causation - Part 2

Here's the link to Part 1 of this article.

Introduction

I enjoy researching the GMAT material, and I’m fascinated thinking about how life works. The concept of causation is frequently tested on the GMAT, and the same concept predominates our lives too.

In this article, I’ll use three examples to illustrate how we make jumps from correlation to causation in our lives. I’ll then talk briefly about how companies use our propensity to make this jump in their marketing. Lastly, I’ll talk about the the logic “Effect. Therefore, Cause.”

Example 1

I took CAT coaching from IMS. I scored 99.98 percentile on the CAT.

After reading these two statements, do you become more likely to recommend IMS to a CAT aspirant?

If yes, you’re forming causality in your mind even though I presented you a correlation.

Correlation means two things have happened together. I took CAT coaching from IMS, and I scored 99.98 percentile on the CAT.

However, if you have become more likely to recommend IMS for CAT coaching, you’ve formed the causality “My coaching from IMS MADE ME score high on the CAT” in your head.

You may be thinking, “Well. The coaching may not have been the sole reason, but it must have played some role.”

This is again your assumption.

What if I told you that I succeeded on the CAT, not because of coaching from IMS but in spite of coaching from IMS?

What if I argued that I had a potential to score 100 %ile but ended up scoring less because of the coaching I took?

Aren’t these situations possible?

They are.

Just because there is a correlation between X and Y DOES NOT MEAN that there is causation between X and Y.

It’s POSSIBLE that Y happened

1.NOT because of X, i.e., X didn’t help make Y happen.

2.BUT in spite of X, i.e., X tried to prevent Y from happening.

Of course, there are other possible relationships between my CAT coaching from IMS and my CAT performance:

1.My CAT coaching had ZERO IMPACT on my CAT performance. No benefit, but no harm too.

2.My CAT coaching had a very small positive impact on my CAT performance. Coaching from other institutes could have been much better.

Thus, unless you probe further, you cannot be sure whether you should recommend IMS for CAT coaching.

Example 2


Here’s the link to this article.

This article was published in April, 2020 and quoted several women political leaders whose countries or states had managed to keep Covid in control by the time the article was published.

The title of the article presents a correlation. Two things have happened together in these countries:

1.Women at the top

2.Covid cases in control

After reading the title, what comes to your mind?

“We should have more women leaders.”

If this comes to your mind, you have formed causality in your head. (Of course, I believe, that was the intent of the author of the article. She wanted you to make the causality in your head.)

You have assumed that there is a causal relationship between “Women at the top” and “Covid cases in control.”

In reality, there may not be any causal relationship between these two. Rather, it is possible that these countries had covid cases in control despite women at the top.

It is also possible that the fact that there were women at the top HAD NO IMPACT on the number of covid cases in the countries.

How can I say that women played a role in keeping covid cases in control by looking at this correlation?

Interestingly, EVERY SINGLE country cited in this article eventually had a big flare in the number of covid cases, equalling or surpassing India in terms of covid cases per day as a proportion of the population.
Perhaps, the two facts were just correlated; there was no causation between “women at the top” and “covid cases in control.”

Example 3

My wife and I meditated on God’s name, and our baby was born with pneumonia.

What does this mean?

Did our meditation lead to the disease in our baby? (Causality)

Or was there no causal relationship between the two events?

Or was is that our meditation reduced the extent of pneumonia in our kid?

The third case is the “in spite” case we have discussed earlier. Our baby had pneumonia despite our meditation, not because of our meditation.

—

Perhaps, you’re beginning to realize that while we are prone to jump from correlation to causation in our daily lives, we should not. Correlation between X and Y doesn’t necessarily mean that X led to Y. Perhaps, they are unrelated. Perhaps, Y happened despite X.

Avoid being trapped by Marketing

Our propensity to jump from correlation to causation is exploited extensively in marketing. For example, when I visit the buy page for LinkedIn Premium, here’s what it shows me:

LinkedIn cites a hard fact. The company has found that premium members get many more profile views than non-premium members.

So, there is a correlation between “being a premium member” and “having more profile views.” Of course, LinkedIn wants us to make the causality in our heads that X leads to Y, i.e., being a premium member makes one have more profile views.

However, if you’re aware of this correlation-causation jump, you wouldn’t make LinkedIn happy. You’d want to ask whether this correlation is a result of reverse causality - it is possible that people who are more popular (thus have more profile views) are more likely to take premium membership. In such a case, taking a premium membership is not going to increase my profile views.

Or you might wonder whether LinkedIn gives premium membership for FREE to some influencers. In that case too, premium members will have a lot more profile views on average because these influencers are part of the premium members.

—

You might already be aware that when IAS or IIT-JEE results come out (entrance tests to civil services and top engineering colleges), all the coaching institutes in the relevant industries rush to claim that the toppers took their courses.

Why do they rush to claim that the toppers took their courses?

Because when they say that the toppers took their courses (correlation between being a topper and taking their courses), people automatically make the causation in their heads - “These guys became toppers BECAUSE of these institute courses.”

However, this causation may not be true. Perhaps, the courses were useless. Perhaps, the guys succeeded despite the courses, not because of the courses.

What we know is that these guys are toppers and that these guys took these institutes’ courses. We don’t know whether there was any causation.

Follow GMAT Experts or Follow Top-scorers?

Are you likely to follow the advice of a 760-scorer over the advice of a GMAT expert?

If yes.

Let me ask you: if you are suffering from an ailment, how likely are you to follow the advice of a healthy person over a doctor?

If you’re not likely in this case, why are you likely in the case of 760-scorer?

You intuitively understand that a person may be healthy but may not be aware of what it takes to be healthy.

However, you don’t understand that a person may score 760 on the GMAT but may not be aware of what it takes to score 760 on the GMAT.

Real Story - One of my students who scored 380 on the GMAT in his first attempt asked a HBS student, “How can I get 700 on the GMAT?”

The HBS student replied, “That’s not difficult. Just do the official guides carefully and take a few mocks. That’s all it takes.”

So simple!

I’ll soon be out of business!

What the HBS student shared was what worked for him and for people around him. However, the advice was completely useless for a 380-scorer. Rather, the advice could be harmful since my student could have lost his hope after doing what he suggested and not getting the score. “I’m doing what it takes to get the score, but I’m not getting the score. Perhaps, I don’t have the capability.”

I’m not suggesting here that you discard all advice coming from 760-scorers or accept all advice coming from GMAT experts. However, I’m making you aware that a lot of junk floats around as advice, and people tend to believe it since it’s coming from the done-and-dusted-guys.

Effect. Therefore, Cause.

We look at the effect and conclude that a particular cause must be there. This reasoning is very prevalent in our daily lives; it appears quite often on the GMAT too.

Many times, we make a statement about somebody’s character or intent, we indulge in this logic.

Because we can never know somebody’s intent or character; we can know only their behaviour.

Thus, we look at the behaviour of people and make conclusions about their intents or characters.

For example,

This guy rarely comes to office on time. Clearly, he is not serious about his job.

Isn’t it possible that this guy is facing severe family health issues? Even in that case, would you say that he is not serious about his job for not coming to office on time?

Likely not.

Think about it.

I offered an alternate explanation for the fact that the person never comes on time. The conclusion was one possible cause.

There can, of course, be many possible causes for the same effect.

Let’s take another example:

Wife to Husband: I told you I had headache, but you didn’t even offer to massage my head. Clearly, you don’t care about me anymore.

What if the husband himself had severe headache and was not in a position to massage his wife’s head?

In such case, the wife’s conclusion won’t stand. The wife’s conclusion was a possible cause of the effect she observed.

When we get an alternate cause for the same effect, the conclusion is weakened.

Let me emphasize it once again: Many arguments about one’s character or intent follow the same logic. We look at their behaviours (effect) and conclude about their intents (a possible cause).

Official CR Questions following this logic

Recently political pressure groups have become far more effective

In the past most airline companies minimized aircraft weight

In malaria-infested areas, many children tend to suffer several bouts

When a female fruit fly is placed in a cage with several potential

Automobile Dealers Advertisement: The Highway Traffic Safety Institute


Help me make this article better

You can make this article better by:

1.Sharing more official or unofficial questions in which popular incorrect options do the OPPOSITE of what is asked

2.telling me which parts of the article are not clear or need more elaboration

3.pointing out spelling, grammatical, or logical mistakes in the article

If you have any questions or doubts regarding anything covered in the article, please feel free to ask. I’ll be happy to help.

I’ll be happy to hear from you in the comments. Please be aware that I also LOVE reading appreciative comments

Do you want Part-3?

I have a few more real-life interesting examples in which the logic of causation plays out. If you want me to write part 3, let me know.