Let say I want to compare two methods and see which one delivers best “life results”. I would need to ensure that the sample I select for both are actually representative of the method.

So let’s assume that on one hand I have “public school attendees” and on the other I have “unschooled”. Public school attendees (within the same country) received very similar education to each other due to standardized curriculum and such. The “unschooled” might have received extremely different kinds of “non school education”

One one hand you have a biased sample, on the other you have a noisy sample. Hard to make sense of it.

Unschooling means different things to different people. Extremely hard to make sense of the value of it all, if any.

The Illuminati: New World Order card game.

It’s quite amazing to see some of the cards now, 30 years after the game was first released.