Teacher observations comprise an important part of my new role as principal. I enjoy being back in the classroom and am in awe of the talent I see on a daily basis. I can only imagine how much stronger of an educator I will be when I return to full-time teaching one day.
I have struggled a little, though, in determining if my evaluation system is effective enough for the teachers I observe. I conduct one to three (depending on tenure status) scheduled, formative observations and one summative evaluation per faculty member each school year. I use pre- and post-observation conferences and have specific requirements about lesson plans for the observation.
My reasoning behind this is that, because I am only observing a full class period a few times, I want to see what the teacher is capable of on her best day. I don't want to observe a teacher proctoring a test, for example, because a teacher would not need much guidance in how to improve that type of activity. I hope that with additional, unscheduled, informal walk-through observations I am able to identify potential areas of weakness to address with a faculty member as well.
I also wonder how much of a teacher's daily experience is truly represented during my visits—either formal or informal. Effective classroom management is a key piece to delivering instruction, and when the principal is in the room, classroom management is no longer an issue. In addition, it's only logical that some teachers will modify their instruction and comportment when their supervisor is present. For example, a teacher might send text messages during class with some frequency but would never do so when an administrator is there.
I hope to ultimately add supplementary types of observations, such as peer evaluations, and have teachers film themselves to critique their own performance. I think if we use such types of evaluation in addition to my observations, and not use them punitively, teachers would be more open to exposing challenge areas. Feedback from a colleague and self-critiques are less intimidating than documented feedback from an administrator that goes into a teacher's file.
I have also read about quantitative observations through which a supervisor uses a categorical frequency instrument to measure how often a behavior exists. Behaviors can include a teacher providing adequate wait time before calling on a student, a teacher asking higher-level thinking questions, or students being off-task. I imagine if this experiment were conducted among teachers rather than by a supervisor, there would be greater confidence in exposing a weakness a teacher wants to improve.