I read a description of this phenomenon in a book (can't remember which one).
Basically, if a problem is presented in a dry, academic format, it's MUCH more difficult to solve than a problem presented in a real-world format.
The example from the book was something like this:
Problem presented in academic format - 10% of people are able to solve it:
You have 3 cards. One side of each card has letters, the other side has numbers. 2 cards are showing - one shows an A, the other shows a 3. How many cards do you have to turn over in order to figure out if all of the numbers are even?
Problem presented in real-world format - 95% of people are able to solve it:
There are 3 people in a bar. One of them is drinking a coke. Another one is a senior citizen. How many people's ID's do you need to check, to see if they're of legal drinking age?
This isn't exactly right, but it was along these lines. You'd remember if you read it as well, it was a very striking and very believable difference in people's ability to solve the exact same problem (logically) but just in different formats.
Can somebody remind me of where I found this, and what it's called, maybe the book it came from or something?