HAVE YOU been the flags boy? Did you use to know all 190 national flags? Maldives, Suriname, Nepal? Can you sort the stars/moon/sun flags by heart? Which one has a rifle? And an elephant?
When the flags boy grows up, he gets involved with Bayesian inference, causal inference, and machine learning, and ends up scrambling correlation, probability, and causality. When it comes to Medicine, flags boys usually become researchers. Then, they have delirious thoughts about causation, just like the guys linked below, who naively believe in the supernatural effects of buffered crystalloids or aspirin in septic shock.
Recently, I have watched a few flags boys lecturing on causal inference and Bayes on YouTube, and I can testify that mammy combed their hair for the lecture. This image takes me to another digression.
Although possible, I never saw a flags girl. It follows that a flags couple must be very unusual. (Bayesian?)
I can imagine a young woman realizing she is dating the flags boy. She sees his eyes shining bright as he speaks about flags and dinosaurs, and also when he does impressive mental calculations. Now she mentally runs the odds of finding someone else in the relationship market. (Counter-factuality?)
I will give her free advice. The receiving of this advice is a binary variable that the flags boy must incorporate into his “Finding_a_Girlfriend” model.
Here it goes: flags boys usually make loyal, loving husbands and the best fathers. Forget the cool guys and marry the flags boy. You only have to work his way off mammy’s influence.
Well, I know I helped someone.
That felt like pouring bicarbonate into an acidotic patient.
Ok, time to get serious.
You know, I had been an assiduous reader of the Tao Te Ching, and by extension, other ancient Chinese texts. One thing I learned, among many, is that small contradictions may go unnoticed when a new trend is rising, but they carry the germ of decay and eventually will flourish and tear the building down.
Today, I am sacrilegiously questioning a few unwarranted dogmas that underpin the contemporary fads around Causal Inference and Bayesian.
Rebuttals are wanted.
Behind the paywall:
The fatal mistake of a cardinal postulate of causal inference.
The erroneous neutrality assumption.
Posterior probabilities meet real life.