The Causal Revolution and Why You Should Study It

In summary, Judea Pearl revolutionized our understanding of causality in the mid-1990's and was awarded the Turing prize for his contributions. Studying causality allows us to ask and answer better questions, and there are three recommended books to learn this: The Book of Why, Causal Inference in Statistics: A Primer, and Causality: Models, Reasoning, and Inference. These books require a background in probability and statistics, with the latter two being more advanced and requiring additional knowledge in calculus and Bayesian statistics. However, studying causality can lead to important insights, such as determining causality without a randomized controlled trial and understanding when to control for confounding variables. Another approach to causality is the Potential Out
  • #1
Ackbach
Gold Member
MHB
4,155
92
In the mid-1990's, an electrical engineer/computer scientist by the name of Judea Pearl started to change the world by greatly improving our understanding of causality. He brought together many strands of thought that had gone before him, then synthesized them into an integrated whole, with many original contributions as well. For this he was awarded the Turing prize, which is the equivalent of the Nobel prize in computer science.

Here's why you should study causality: because once you've done so, you can begin to ask and answer better questions. For example, instead of merely noting that a hospital's appointments are down at the same time some virus is spreading around, you can ask the better question: is the virus causing appointment counts to go down? The new causality tools give you what you need to answer that question! It is still an inductive procedure, so it's not as though you go from induction to deduction. However, you're asking and answering the questions people really want to know: the "why" questions.

Here's how to learn the new causality. Prerequisites: probability and statistics, the more the better. If you've had a typical calculus-based version, you'd certainly be well-prepared. However, the first book on the list only requires basic probability and statistics. If you want to be able to do all the computations yourself, you would need more background to get through Books 2 and especially 3.

Study these three books, in this order.

  1. The Book of Why, by Judea Pearl and Dana Mackenzie.
  2. Causal Inference in Statistics: A Primer, by Judea Pearl, Madelyn Glymour, and Nicholas P. Jewell.
  3. Causality: Models, Reasoning, and Inference, by Judea Pearl.

Teaser: contrary to the standard doctrine of traditional statistics, which I had learned, you do not always need to have a randomized controlled trial (RCT) in order to establish causality! With the right data, even an observational study can give you causality (this is how we know that smoking causes lung cancer, e.g., when the right RCT would be unethical).

Another teaser: Have you ever wondered how you can tell when to control for a possibly confounding variable or not? The new causality not only makes the whole concept of confounding much clearer, but tells you when you need to condition on a variable, and when NOT to condition on a variable! (Hint: sometimes conditioning on a variable gives you the WRONG answer!)

Highly recommended!
 
Last edited:
  • Like
Likes Filip Larsen, scottdave, benorin and 3 others
Physics news on Phys.org
  • #3
scottdave said:
And on a lighter note, check out these Spurious Correlations - https://tylervigen.com/spurious-correlations
Right! Although, as the books above point out, the correct statement is not "Correlation does not imply causation." A better statement is, "Correlation often implies causation." Equally important: "Correlation does not imply confounding."
 
  • #4
I'm just reading "The Book of Why". It explains, among many other things, why it is actually true that beauty and intelligence can be negatively correlated, provided that we consider a population with a common feature, such as success in show business.
 
  • Like
Likes scottdave
  • #5
What sort of prerequisite in typical statistics is assumed? Is the list in increasing order of difficulty/advanced-ness?
 
  • #6
Muu9: Great question! Here are the prereqs as I see them:

The Book of Why: high school statistics.

Causal Inference in Statistics: A Primer: the usual calculus sequence (including multivariable) followed by mathematical statistics.

Causality: Models, Reasoning, and Inference: This is extremely difficult, and I have not read it. I recommend the usual calculus sequence, mathematical statistics, Bayesian statistics (to the level of Gelman's BDA), and Bayesian networks before attempting this book.

These three books constitute the Directed Acyclic Graph approach. The other main approach, the Potential Outcomes Framework, is headed up by Donald Rubin. The main book here is

Causal Inference for Statistics, Social, and Biomedical Sciences: An Introduction: Prereqs appear to be (I have not read this book) calculus, mathematical statistics, linear models, and design and analysis of experiments. Bayesian statistics wouldn't hurt.
 
Last edited:
  • Like
  • Informative
Likes Muu9 and Greg Bernhardt

FAQ: The Causal Revolution and Why You Should Study It

What is the Causal Revolution?

The Causal Revolution refers to a shift in the way scientists approach causality in research. It involves moving away from traditional methods of establishing causation, such as randomized controlled trials, and towards more complex and nuanced methods that take into account multiple factors and potential confounders.

Why is it important to study the Causal Revolution?

Studying the Causal Revolution is important because it allows scientists to better understand the complexities of causality and make more accurate and reliable conclusions about cause and effect relationships. This can lead to more effective interventions and policies in various fields, such as healthcare, economics, and social sciences.

How does the Causal Revolution impact current research practices?

The Causal Revolution has led to a shift in research practices, with more emphasis on methods such as causal inference and counterfactual analysis. It has also highlighted the importance of considering potential confounders and using more sophisticated statistical techniques to control for them.

What are some challenges of the Causal Revolution?

One of the main challenges of the Causal Revolution is the complexity of the methods involved, which can be difficult for researchers to understand and implement. There is also a need for more data and resources to support these methods, as well as a need for interdisciplinary collaboration to fully understand and utilize them.

How can studying the Causal Revolution benefit society?

Studying the Causal Revolution can benefit society by improving our understanding of causality and leading to more effective and evidence-based interventions and policies. This can ultimately lead to better outcomes in areas such as healthcare, education, and social justice.

Similar threads

Replies
1
Views
2K
Replies
40
Views
4K
Replies
1
Views
2K
Replies
3
Views
950
Replies
26
Views
7K
Replies
5
Views
5K
Replies
5
Views
4K
Back
Top