The XKCD article is a misnomer. Causation is actually verified.
1. You measure a population about whether they know about a fact. You find most people don't know it.
2. You introduce a statistics class to that population that teaches said fact.
3. You measure the population after the class and you find most now people know about said fact.
That is literally proving Causation. But you'll see there's a trick in my wording. If I didn't deliberately introduce that class to a population, and I simply observed an existing population that took that class... well, then in that case I cannot prove causation.
However the character that was asked the question in the XKCD comic, deliberately introduced the statistics class on himself. This deliberate act changes the experiment. HE himself is the experimenter and he controls the parameters and thus he has the knowledge needed to verify causation.
The person who asked the question cannot verify causation, but the person the question was asked too, CAN verify it.
I get the humor behind the comic. But it is technically wrong. Causation is verified. All causative experiments are run this way.
However the character that was asked the question in the XKCD comic, deliberately introduced the statistics class on himself. This deliberate act changes the experiment. HE himself is the experimenter and he controls the parameters and thus he has the knowledge needed to verify causation.
The person who asked the question cannot verify causation, but the person the question was asked too, CAN verify it.
I get the humor behind the comic. But it is technically wrong. Causation is verified. All causative experiments are run this way.