Their star course, called “Modern Applied Statistics: Learning,” started a decade ago with 30 students. Its current enrollment just closed off at 190. “We try to give them long and difficult homework assignments,” Mr. Hastie says. “Nothing works.”
Now read that last sentence again and ask yourself this question: “is this how we get people into STEM?” I think not.
Two more quotes:
The two men also teach a two-day course for businesspeople called “Statistical Learning and Data Mining” that costs $1,450 and attracts a broad range of data-laden people. “We had two guys from Hong Kong who taught a course in horse race prediction,” Mr. Tibshirani says. “One of them came back and told us they’re making $10 million a year by modeling the last-minute betting.”
Are the only people who ever get to learn something like statistics from a teacher that is designing the course for educational impact (as opposed to status-affirmation) those that spend an enormous amount of money or those that spend nothing at all? How weird is that?
About half of the Stanford stat professors have joint appointments with other departments, including economics, human biology and environmental science. “Statistics is unusual,” Mr. Hastie notes. “It’s a service field to other disciplines. It doesn’t rely on its own work. It needs others.”
This last quote is interesting. Statistical techniques are tools and are not particularly useful without domain knowledge. But they still need to be taught by a specialist, probably.
But these guys/gals get down and muck it up with real data every day and for that they need to join forces with others. It’s an applied discipline.
via Jim Lynch