Science education is sort of like the weather -- everyone discusses it but nobody seems to know what to do about it. From Ars Technica comes an article that says stressing science education in middle and high school is bad timing. What we really need to do is start educating them in kindergarten.
They're right on the mark when they say that even very young kids "figure out" how the world works -- and they're often wrong about it or confused about how things work. The direct method of teaching (where it's explained what's going on) is the best for the youngest kids when it comes to bringing concepts to them. But without the "magic" of the curiosity-inducing "discovery method" of teaching (where students devise, execute, and interpret their own experiments), some of the fundamentals can be lost.
And here teachers are sitting on the horns of a dilemma. It's pretty easy to outline a "direct" method of teaching (just lecture.) "Socratic methods" where you lead them through it with leading questions usually only work best for Socrates -- the rest of us end up with an audience who asks questions we don't expect. Discovery methods are very good -- but only once you've taught the kids how to do a proper experiment.
The best solution seems to be to combine all three -- give a short lecture, ask leading questions, have them explore.
...of course, this assumes you've got a perfect class and that Rudy and Maxie will NOT decide to throw spitballs in the middle of the lesson just to see if they can annoy each other.
