The first class test happened yesterday. I always worry something will go wrong on the first test, when the students are nervous and unforgiving. But technically it went without a glitch, amazing since I had to write it a week ago under great time stress. And yesterday I did not have room for error. In Phoeniz, AZ, at 5.30 am I checked the test was live and then it was 2 pm in Ann Arbor, MI before I could next check the action (and then panic: less than a third of the class had made a start!). At 9.30 pm when I next got a chance to look, I was just glad there were results. But now, the day after, I see those results are not great.

The overall class average was 71%, or 75% if the 11 no-shows were discounted. No one got everything right, but on my ask-28-questions-grade-out-of-25 algorithm, four students got 100% (one of those got 26/28 right). Nine students got an A, 7 an A-, 19 B+, 17 B, 31 B-, 25 C+, 17 C, 47 D and 16 fails (class size = 199). To see a graphic of the distribution, click here.

However I look at those numbers, they’re not good. I’m pleased a few students did very well – it shows it was possible. But more students got unacceptably low grades (defined in my books as D or less) and fewer achieved an A or A- than in any previous class at this point (2013, 2012, 2011, 2010). The class mean is in line with 2010, the lowest score ever.

Why so bad? Am I teaching worse? Is it because the class is almost entirely freshmen this year? Good thing the grading algorithms in this class are set up to reward improvement. That means all of us – students and me – can improve, and there will be no legacy of this. It might even help kick all our butts. Just what I’ve always wanted at this point.