I was a part of a special judging contingent at the state science fair over the weekend; the state chapter of the American Statistical Association provided a contingent of volunteer judges to judge the statistical data analysis content of the various exhibits. Fairly heft cash prizes were awarded.
It was a full day because only seven statistics judges showed up and we had to split all 100 or so of the exhibits between us (I got 16). We got a nasty surprise when we showed up for orientation and discovered that only about 20 regular judges had volunteered (which was apparently fewer than previous years), so the fair organizers decided to give half a regular judging load to the special statistics judges on top of our statistics judging load. At least some of my two judging assignments overlapped (luckily), but discovered I was the most oversubscribed judge at the fair with a total of 18 kids I had to judge in two hours.
The kids at the state science fair were all kids who had won their regional science fair, so we were judging the cream of the crop as far as young scientists go. Sadly however, only around maybe 2/3 of the kids knew how to take averages, and very, very few knew how to determine if there was a statistically significant difference between the averages of two groups of data. In fact, we only had 7 shortlisted candidates by the afternoon who showed even an iota of statistical rigor in their experimental design and data analysis. I was (after all the judging was over) pleased to realize that 5 of those 7 kids were girls (although I know for a fact that not a single judge in our group was looking at gender as a factor in our judging...we were simply desperate to find any kid who showed any statistical rigor in their data analysis). The student who won our grand prize was actually a 6th grader who had done a study of gender differences in social networking amongst her school mates. Neither of her parents had gone to college (we always ask discrete questions about the parents to determine if mum or dad was the person who was actually primarily responsible for the work), so it was clear from talking to her and her responses that she, and only she, did the work. The kid devoured psychology texts, and apparently at quite an advanced level because she said she had learned about experimental design and statistics from some of her books; she had developed a wonderful experimental design that blinded her from knowing which data corresponded to which of her classmates, and which classmates her classmates associated themselves with as close friends, friends or acquaintances. Then she applied the Student's t-test to determine whether the average number of close friends for girls was significantly different from that of the boys (it was). And her description of what "statistically significant" means was far better than most graduate students in statistics could give.
And it turned out that the three prize winners and the two "honorable mentions" consisted of all the five girls out of our seven shortlisted candidates. You go girls!!! I wouldn't have been disappointed if it had been 50/50 though...I am happy that there are at least some kids out there, regardless of gender, who have realized that rigor in statistical data analysis is an important part of pretty much any scientific endeavour.
I attended the feedback session for students, parents, and teachers after the fair was over. Sadly I was only one of two judges who showed up at the Junior division feedback session, and walked into a room that was filled to standing room only with kids and adults. The only other judge there was actually an army recruiter (the army has apparently made the canny move of volunteering judges to state science fairs...that way they have a captive audience of intelligent kids to give their recruiting spiel to). The army recruiter guy got up and gave a recruitment spiel. Dead silence followed.
Then I got up and gave a short description of my professional background (including the fact that I was the first person in a very large extended family to ever attend college and that I worked my way through and ultimately got a PhD in physics and that I absolutely adored physics and statistics). Then I wrote "Student's t-test", "variance", and "standard error of the mean" on the blackboard and gave a little talk about how the kids wouldn't learn this stuff in their math class because the No Child Left Behind Act pretty much dictated the math curriculuum and the teachers simply didn't have time to teach statistics (I noted a lot of teacher-looking-types in the crowd nodding their heads at this point). I told the kids to Google the words I had written on the board and/or read up on them in books from the library, and then apply the principles they had learned to next year's science fair projects. Because I wanted to see their already good science fair projects turn into great science fair projects that would win them lots of prizes at the local, state, and national science fair levels such that they would be able to get scholarships to go to university. And not just any university, but great universities like Harvard, MIT, etc. And the way they could do that was to do some extra reading on their own and nurture within themselve their curiousity about the natural world around them and always try to have as much fun as possible while doing it. Because America needed to produce more scientists, and I would be overjoyed if our state became one of the pre-eminent states for producing some of America's best scientists in their generation.
Lots of applause after that. I hope at least some of it stuck. I'll find out at next year's science fair I guess....
