Consider this one next time instead of the airplane:

This is quite cool

From my understanding, the idea behind this distinction is based on the fact that the mathematical structure behind DAGs is “not able” to express causality. For example, the three DAGs in the attached image have the same restrictions on the observed data: (2/3)

From the many sources on DAGs, I can only recall Hernán & Robins (What If?) - attached image - and Pearl (Causality) distinguishing between DAGs and *causal* DAGs. I guess I won’t say much about what Pearl wrote because, at the end of the day, I just ended up more confused. (1/3)

The psychometrics (SEM) literature has 2 issues I dislike a lot: (i) assumptions not clearly stated and (ii) absence of proofs when very much needed (in my opinion). As an example of the former (in its simplest form possible), we have the beginning of the third chapter in Bartholomew et al. (2011):

Any comments?

Sharing a PhD position from my department:

Things that make me genuinely happy: wake up to a sunny day in Gent and go to class in which Stijn basically taught the content of his article