Here's one. You can watch the video and click on what you think the right answer is to see if you got it right.
What is interesting about this problem is that 90% of people get it wrong. But I think that is not surprising. Our brains generally don't strive for efficiency which is what this problem is asking you to do (least number of cards turned over to verify the proposition). Flipping a card costs nothing so why waste brain power on figuring the least amount of cards to turn over?
But also interesting is that the largest reason people get this wrong is is because it deals with symbols (colors and numbers). If the same question is re-couched into something we have familiarity with, then the percentage of correct results shoot up. This was shown by another set of psychologists in 1982 when they rephrased this problem into one that involved asking if 4 people at a bar were old enough to drink (21 years old). The right answer went up to 75%. There are two competing explanations of why this contextual change, resulted in dramatically different results.
The first belief is that we have two competing analysis systems in our brain. One that is very good and analytical and one that is more intuitive. If you got the answer above wrong then your more intuitive answering system likely made that choice of answering. At least that's the belief.
This makes some sense to me. Going into this I knew, as you probably did, that this was going to be a trick question in some sense. And so my brain first tried to structure the problem as my engineering degree and training at McKinsey has taught me to do. I reread the question to be sure I understood what was being asked and then systematically went through each card to determine if it had any bearing on the answer. My answer is right. But you can see how this is a laborious process when we might be introduced to many choices like this in a day. Studies show that the more analytical system degrades with age.
The other answer to why people get Mason's problem wrong and the drinking one right is that we are socially conditioned to quickly understand if someone is breaking the rules. I find this funny but I also think it also carries some weight as an explanation. I'm not sure if the wording of the Mason problem can be changed to make it easy (75% hit rate) but not involving something that invokes our sense of a social contract to test this theory.