I'm reading "Why don't students like school?" The book is quite good because it shows certain cognitive principles, (mostly based on how human memory works) that make teaching (and learning) easier.
Nevertheless, I find certain things disturbing. For example, the author says that to teach abstract ideas you need to show a lot of examples:
To Help Student Comprehension, Provide Examples and Ask Students to Compare Them.
Let's use the Newton's principle and make the rule universal. That is, imagine that students become really good learners, by exploring examples and comparing them to each other. How the hell are they going to learn proofs? No wonder that even the best high school students when they go to university have no concept of proving their ideas logically, instead of by "here's another example."