Time to expand the analogy a bit, I think:

Sometimes students cheat on their exams.

And sometimes the students that usually don’t do very well will surprise you with good performances.

Ronnard:

You’re absolutely right. Major education research reports have criticized the U.S. school system as being “a mile wide and an inch deep” – there are too many things being taught, in not enough detail.

It takes students time to learn a new concept, and if we try to pack in too many things, they end up not remembering any of it.

You’re also right that we want students that can solve problems. It’s less clear how to do that, although student projects seem to help, especially if the projects resemble what real professionals in that field do. In this context, doing scientific research should really help students learn science.

It should be sufficient, if they would get an overview about a topic, learn the most significant methods and facts and then start directly to do some original work themselves.

Lets take astronomy as an example:

Would it not be possible, that a student gets a short overview of lets say 6 months and thereafter does do some research work, like doing observations with telescopes himself and doing some theoretical calculations. Eventually using some finished software programs.

Thereafter 6 other months with teaching and so on.

The question is, what our societies want to get from their students: An army of uniformly trained people or a lot of highly motivated and creative people, who are eager to solve problems ?