Discover more from The Wisdom Project
Correlation Causation Fallacy
#123 | Mental Models to Live By🧠
Welcome to Wisdom Letter #123.
Today we are talking about the Correlation Causation Fallacy.
This mental model derived from statistics is perhaps the most important tool we need to make sense of all the information thrown at us every minute of every day these days.
But first, if you’re new here, consider subscribing to The Wisdom Project. We send 1 article every week that helps you think better, live better :)
Now, let’s dive in👇
What is the Correlation Causation Fallacy?
It states that just because two events are correlated does not imply that one causes the other.
Causation may or may not exist.
Their correlation may be explained by a third external factor that caused both of them.
How it works
Our brain is a very high energy-consuming organ. It has evolved to function in ways that can improve its energy efficiency.
Making connections(even if none exist) makes it easy for us to understand and remember a fact.
We humans, find it more difficult to resist connecting two facts when presented together.
This bias served us well when we were in the African savannah.
As homo sapiens, if someone died after eating a berry, it helped to conclude that those berries are dangerous.
It did not matter if that was true. The fear kept us alive.
We didn’t need data and randomized controlled trials to prove that the berries were poisonous.
We trusted correlation.
We evolved to derive causality in correlation.
Why it matters
In the information age, we generate millions of data points every second. Data-backed studies follow us in every walk of life.
Our inability to differentiate correlation and causation can lead us to draw wrong conclusions from these studies.
The scale of if this decision can vary from something personal - eg. which latest fad diet to follow to something global like what leaders we elect to power.
Where it works
Does eating red chilies lead to longevity? Does drinking green tea lead to weight loss?
These claims are often correlations presented as causations.
When reading through such claims, it helps to know what other factors could have led to the same results, who were the participants and how was the study concluded.
In fact, who is to gain if these claims were indeed true.
Would you attribute your business problems to COVID or would you care to find out deeper issues that were accelerated with COVID?
Was the increase in sales of your recently launched product indicative of product-market fit?
Or you are a beneficiary of a booming market?
Where it does not work
Correlation is indeed the basis of a lot of scientific theories.
Studies about smoking leading to lung cancer, high consumption of sugar leading to diabetes were evolved from correlation studies.
The scientific way to go about it is to list down all the possible factors and work on elimination.
Here is a good explanation -
Causation vs. Correlation | Farnam Street Blog
Last week we wrote a thread about a popular question on the internet -
Would you rather take a guaranteed $100k or a 50/50 chance at $10Mn?
What is your answer?
We broke down the process to think about it. Do check the thread and share your thoughts -
📂 From the archives
This Week Last Year - 2021
This Week - 2020
We would appreciate it if you can move us to your primary inbox.
This year we’re doubling down on curating the best content online to help you think better and live better.
And in the process help you transform your life and become the best version of yourself.
We don’t want you missing any emails just because the email went in your promotions tab.
Also, if you’ve signed up recently, do tell us what is the 1 thing you want us to focus on in 2022, what are your expectations from us, we would do our best to live up to them.
Ayush & Aditi