Archive for January 21, 2011
Beth Simon made an excellent recommendation after my report on my first Peer Instruction lesson: Was it really a bad question that students misinterpreted? Why not ask the students? You would expect that students would most likely give me the answer on this survey that they thought I wanted. This was the first slide of the day.
Here’s the distribution of responses:
I did several more “clicker” questions today in lecture, and I’m getting a better sense of what works and what doesn’t work. (Something that doesn’t work: My <expletive deleted> Lenovo TabletPC that refused to wake up at the start of class, requiring me to reboot, and losing 10 minutes of lecture! ARGH!) I asked students to write in a piece of code today (rewrite a FOR loop as a WHILE loop). The answers were actually pretty good, but the writing took a long time. I won’t do that often.
One of the general insights I’m getting is about the large variance in the class. Here’s another question I asked in class (before the Java nitpickers let loose — we’re using DrJava, they’ve seen that code works fine without semi-colons, and in fact, we had just done these three lines verbatim with variable “fred” instead of “mabel”):
And the responses:
Most of the class grokked this one, but 5 of the 22 students who responded (some told me after class that they didn’t even respond) are pretty confused. That’s over 20%.
I chatted with several of the students after class today. They’re very confused, despite having read the first two chapters of the book (they claim) and taken the quiz. (I’m using out-of-class Video Quizzes, where students watch a videotape of me using Java, then answer questions about it.) My main insight into their confusion: After only one semester of CS classes, reading code is not an automatized skill. That’s not surprising, but it’s not something that I’d thought much about. The students told me that they’re metaphorically “sounding out” the code. They’ve thinking through what’s a method (and translating that into a MATLAB or Python “function”) and what’s a class and what’s valid Java syntax with semi-colons. That’s taking them time, and sometimes, they’re responding before they’re really confident about what they read.
Peer Instruction is taking me extra time: To get the slides onto Ubiquitous Presenter, to only present from my TabletPC, to write questions and insert them into slides, and to take time from lecture (for students to answer, to discuss, to respond again). I still think it’s worthwhile, and I plan to continue trying it.
I’ve been hearing a lot about the new COMPETES law and how much it helps education programs. This article raises a serious concern. The new education programs are mandates on NSF, and with the new Congress, increases in NSF’s budget are unlikely. A possible (likely?) outcome, then, is that research programs would be cut back to fund the new mandates.
The reauthorization also tells NSF to begin several new education initiatives. One asks NSF to replicate the successful UTEACH program at the University of Texas, Austin, that trains STEM majors to become science and math teachers in public schools. Others would encourage high school students to help university scientists collect data for NSF-funded projects, allow for a competitive grants program to support research on improving graduate education, and create industry internships for undergraduates in STEM fields. All of them would require new funding, however—some $10 million a year for the UTEACH replication, for example—meaning that NSF officials are extremely unlikely to move ahead unless Congress appropriates the money for that particular activity.