Next time someone says, "correlation isn't causation" I now can say, "didn't you keep reading in the stats textbook? The part about causal inference?"
Anyhow, the justifications remind me of Bayesian statistics, where you take prior knowledge and that allows you to say and test something more than "associations do/don't exist" #pywhy
https://www.youtube.com/results?search_query=causal+inference+pywhy
I wish I knew how many data points this requires. (30? 100s per explanatory variable? Tens of thousands to millions like machine learning?)
♡ 1 ↻ 0