There were a few posts over on Facebook this weekend about teacher certification. It started with a question - should teacher certification be based on an exam like APCS-A with the teacher correctly answering 85% or greater. A lively discussion followed. There was agreement and disagreement with responses ranging from:

• good idea
• it's too hard
• it's not hard enough

There was also a discussion of the importance of content knowledge vs knowing how to teach in general.

Since there was a lot going on in the thread I thought I'd pull my own thoughts out into this post.

Let's first look at the test. There are a number of states that have a content test for computer science. For some states, I think it's the only requirement. In terms of content, at least the Praxis exam is something like APCS-A on steroids. This is not a good thing. Don't get me wrong, I think a qualified CS teacher should know more than APCS-A on steroids but a single day test is not how you measure deep content knowledge. I know plenty of students who've passed standardized exams - be they regents, AP, or other but they were no way qualified to teach the course. They knew the formulas, recognized the patterns, but didn't really know the material. Is this what we want of our CS teachers? I certainly don't. I'm not hugely opposed to an exam covering somethign like college CS1 content other than my general anti-testing beliefs - it would help to curtail some of the problems that will invariably arise with weak certification programs but it should no way be the actual bar to teach - if it exists, it should be just one standard measure along the way.

Now to that content vs pedagogy thing. What current teachers and in fact most people in the CS Ed space forget is that states move slowly and that we're playing the long game. Right now most CS teachers are converts from other disciplines and relatively few have strong CS backgrounds but over time this will change. As CS becomes more established we'll have young teachers entering the filed with the intent of becoming a CS teacher. If states and teacher prep programs set the content bar low that'll be what we're stuck with for a LONG time. We wouldn't dream of creating a science teacher credential where the teacher hardly knows any science - no way should we do it for CS or any other subject.

The truth is that we need teacher prep programs that ensure that new CS teachers know both the content and how to teach it and that content should not be tied in to any specific curriculum. I wrote more about all of this here.

So what do do with current teachers and current needs? Easy - give them time. Back when I started teaching I was a career changer. I never took an education course, never student taught, nothing. Turns out that Math was a shortage area so I had an option - as long as I had enough math credits I'd get a temporary license and start teaching. Instead of teaching 5 classes, I'd teach 4 and get a mentor. I then had to get a certain number of education credits within a certain period of time. Since I already had the subject credits it was just about learning how to teach. Truth is, I barely survived that first year but things did get better.

New York and other states probably couldn't do the exact same thing. First off, current and potential teachers might not know the content so it wouldn't just be ed courses and second, certification rules have changed big time since NCLB. Still, they could do something similar - approve certification programs like the one I created at Hunter - ones designed for current non-CS teachers and give teachers a period of time to complete it. For career changers, create CS version of programs like the NY Teaching Fellows.

What isn't acceptable is saying that CS teachers don't have to know CS and that career changers who know CS don't have to learn how to teach. Create the programs and give people time. We're playing the long game and the only way we lose is if we cut corners for convenience at the expense of the kids.