Thursday, July 04, 2013

Why is Science Hard?

A new study (gated)by Ralph and Todd Stinebrickner reports that the relatively low number of science, technology, engineering and math graduates is not due to lack of interest in those subjects but by the fact that students planning to major in them find them too difficult. Matt Yglesias and Tyler Cowen have linked and commented, but I have yet to see anything deep on exactly why this is.

I am absolutely confident that this result will surprise absolutely no one who has taught or studied physics or engineering. Anyone who gets an advanced degree in one of these subjects has noticed that they are better at that sort of thing than most of their contemporaries. For the overwhelming majority of us, we are also aware that we are not nearly as good at it as some other members of the profession. What factors account for these differences?

There is a popular meme that it takes about ten thousand hours of fairly intense work to get good at anything "hard," whether it's chess, piano playing, or physics. I don't think that's wrong, and I would guess that it's at least that number for physics or math, but there will still be big disparities after those ten thousand hours. Certainly there are many who aren't willing to put in the effort, but one can hardly rule out talent as a factor.

Is it possible to be more specific about the elements of that talent? I don't claim to know the answer, especially at the higher levels of achievement, but in physics and probably most STEM subjects, our key test of accomplishment is the ability to solve problems. What does it take to solve a physics problem?

The known laws of physics can be written down rather compactly, so solving problems that doesn't require inventing new laws of physics comes down to figuring out how to apply them. So what does it take? A bag of tricks? Ingenuity? Some kind of magic analytical ability?