Brenda Wilkerson, Director of CS and IT education for Chicago public schools was one of the keynote speakers at this year's CSTA conference. During her talk, she made a comment about it taking three times through to get it right so if you're working hard and struggling in your first year of teaching CS, it's not the time to give up. Learning to teach takes time.

She's absolutely correct. I saw this in myself and I've seen this in many other young teachers. First year through, your job is to survive. This is when you rely on colleagues lesson plans, it you can you shadow another teacher and in general do your best to give what you can to your students while making it through the semester.

Then, I often see something interesting. Year two is frequently a step back. It was for me and I've seen this in many young teachers in many subject areas. I'm guessing this happens because we feel we have a year under our belt and then we overreach while trying to be innovative and more effective.

In year three things start to improve again and then, if the teacher continues to work at their craft, there's steady progress for years to come. At the point, the teachers seem to better understand how to experiment and grow as a teacher in a safer way while also understanding that it's OK to have a bad day or even a bad unit - there's usually time to recover.

It's also interesting that while this progression is most obvious in new teachers it also seems to take place with experienced teachers teaching new courses.

All this means that teaching CS or any new subject takes time, effort, and patience. It means young teachers shouldn't beat themselves up and that supervisors need to be supportive and give the time and resources to allow teachers to succeed.

It also means that schools with high teacher churn - notably charter chains create environments that are anything but conducive to allowing society to build a cadre of expert teachers. It works for these charter chains because they're exploiting the fact that they require a small number of teachers as compared to large urban public schools so can burn through teachers like gasoline and the fact that they're judged only by standardized tests. The model is not sustainable if privatizers get their wish and charters become the national norm.

Finally, it's another reason for people to wise up to "leadership" programs like Teach for America where candidates come in for a couple years before moving on to "better" opportunities.

So, if the third time's a charm, how about the flip side? After how long a period of time should we expect CS teachers to really know their subject area? I'd say three years is also probably right. Year one, the teacher is just surviving but after that, I'd expect a teacher to want to master their subject. When I taught math early on in my career, I felt overmatched. Particularly at Stuyvesant. I was a CS major, not a math major. What did I do? I sat in on colleagues classes and did self study over the summer. Of course by the time I got up to speed, at least to my standards, I was teaching compute science.

So, if you're a new CS teacher, why not take CS101 and data structures over the summer instead of doing another scripted PD? Between local and community colleges, at least in urban areas, this shouldn't be a problem.

So, is third time a charm? Three strikes and you're out? A bit of both?

We have to give our new CS teachers the time and support to learn to teach and to learn the subject area but at the same time, I think it's reasonable to require that they do so given the resources.