Thread by @mcmullarkey

Figuring out what causes what is SO HARD

And especially if you have a psych background, you might think we *need* an experiment to understand causes

While I love experiments, here's a thread of resources on why they're neither necessary nor sufficient to determine causes 🧵

This paper led by @MP_Grosz is a great start! It persuaded me that merely adjusting our language (eg saying "age is positively associated with happiness" instead of "happiness increases with age") isn't enough

journals.sagepub.com/doi/full/10.11…

If our underlying research question is causal, we still need causal methods! But if they're not just experiments, what are the options?

Luckily for us @dingding_peng has a must-read primer on using causal methods with non-experimental data

journals.sagepub.com/doi/10.1177/25…

Reading this paper helped me understand experiments are incredibly useful, and they aren't a "get out of jail free card" for determining causes

She also covers the basics of using directed ayclic graphs (DAGs) for understanding causes, and I found her explanations super useful!

Taking causal inference seriously can also help us more effectively answer research questions we care about!

A big question in intervention science is how intervention effects might differ if they were administered in new populations

This framework led by @dominik_deffner helps us understand how we could do just that (plus lots of other stuff too!)
psyarxiv.com/fqukp

This paper also does a great (and unsettling!) job of discussing how even if your research questions seem more descriptive you still have to think about causal models

Have I mentioned science is hard??

There are other cool frameworks for understanding intervention effects in new contexts like this from @economeager (though I'll admit a lot of it is over my head as of right now!)

lse.ac.uk/economics/Asse…

Want a textbook length treatment?

I've learned a bunch about how to think about science along doing causal inference in complex circumstances from Statistical Rethinking by @rlmcelreath

Book ($): xcelab.net/rm/statistical…
Lectures (Free): youtube.com/playlist?list=…

If a textbook is too intimidating he has a series of blog posts that get at some of the book's core concepts:

elevanth.org/blog/2021/06/1…

If wading into this lit causes you to have a bit of an existential crisis about how we do research, you're not alone!

I really vibe with this thread by @ShirleyBWang that breaks down the need for more time and space for people to learn difficult methods

https://twitter.com/ShirleyBWang/status/1443229481875873798?s=20

I'm super lucky/privileged to be in a position where I get to learn as much as I do, and most people don't get that chance

I hope this list of resources is at least a bit of help to people who are interested but didn't know where to start!

PS Diving into the causal inference lit has helped me realize a lot of the questions I care most about are prediction-focused

If you're interested in a primer on prediction-focused approaches, check out this paper from @talyarkoni
ncbi.nlm.nih.gov/pmc/articles/P…