This is a good and famous problem, but, to be honest, it don’t think it really hits any of my standards in here in NY.
- Riley Lark has a nice post where he walks through how he handled the Monty Hall problem with his class
- Paul Solomon, on twitter, reports running a big simulation with his classes. Following him, I tried this. I gave a pair of kids 3 cups and a coin, and told them that one would be the “switcher” and the other would be the “stayer.” Kids close their eyes while their partner hides the coin, then they keep track of the data and then we merge it.
- When I tried this in class, something funny happened. The data was all messed up. In a class of 23 with 20 trials per kid, the data had switching and staying at about 50% each. I starting sputtering about how this really, well, probably didn’t happen by chance. Heads were falling down, so I shrugged my shoulders and showed them this thing:
Then two groups admitted that they completely cooked their data. (Which, to be fair, was entirely obvious from their data but there’s no way to win that argument with a kid.)