6 Comments
Jan 29·edited Jan 29Liked by Adam Kucharski

Interpretations like "Education is so important, just look at how much it lowers risk of death" remind me how disconnected I am from the "general public". I'm getting flashbacks of the covid rhetoric I've been seeing for the past four years.

Framing things in a causal way is much more helpful. I suspect that editors who shy away from causal language have an overly simplistic understanding of what it is to follow a scientific approach, and/or are trying to hedge their bets so they are less vulnerable to criticism.

Expand full comment
Jan 29Liked by Adam Kucharski

What are your thoughts on causal inference techniques, which seem to be at least aimed at this issue ie addressing causality questions in observational data?

Expand full comment
author

Yes, I think there's a lot of useful and modern methods we could be drawing on here (the Richard McElreath post I linked to has more)

Expand full comment

Adam, I am hoping that your writing on this issue of association vs. causation will be read by a wide audience! I am very sad with the state of our vaccination lately and hoping that people reading your blog on this issue of association vs. causation will get better understanding. Thank you Adam and thank you for those "references" you mentioned in this post.

Expand full comment

There are some papers (including new takes) on Bradford Hill's criteria for establishing causality. We dont need to reinvent the wheel when trying to establish causality...

Expand full comment

But where would be the profit in that?

Expand full comment