Sunday, July 26, 2009

Arne Duncan's Right--At Least Once

Heh. I had exactly this discussion with a friend--at that time he was a teacher--about 30 years ago.

President Barack Obama and Duncan on Friday unveiled proposed rules on how the money will be awarded. One of the firmest: "To be eligible under this program, a state must not have any legal, statutory or regulatory barriers to linking student achievement or student growth data to teachers for the purpose of teacher and principal evaluation."

Wisconsin is one of the few states that have such a rule, right there in state law.

Or, as Duncan put it in a New York Times interview: "Believe it or not, several states, including New York, Wisconsin and California, have laws that create a firewall between students and teacher data. I think that's simply ridiculous. We need to know what is and is not working and why."


His notion, generally opposed by, among others, teachers unions, is that teachers and their bosses should be held accountable for student success (or lack thereof). The idea applies not only to teacher evaluations but the controversial subject of merit pay for teachers.

The objection often raised is that teachers have no control over the home environment of their students, and that environment is absolutely critical to continued learning.

But that assumes the "snapshot" evaluation of a student.

Any evaluation of a given teacher/student must take into account that student's past history in achievement, just as it should take into account a teacher's past history with other students. A simple example:

Johnny didn't learn to read well in first grade. He only scored 45%. If his second- and third- grade teachers manage to keep Johnny at 45%, or improve him to 55%, they've done just fine. However, if after 2nd and 3rd grade, Johnny's score falls to 35%, or 25%, then is something wrong with either the 2nd or 3rd grade teacher?

Maybe. But before we shoot Teacher Sally in the head over Johnny, we also look at Teacher Sally's track record with all her OTHER students during the period in question. If she's done well with the others (maintaining or improving scores), then Johnny has a problem.

If her other students, by and large, have also showed deterioration, then Sally needs to go.

Seems that Arne has the right idea.

1 comment:

John Tenny, Ph.D. said...

I agree that having the information on all the contributing factors is important to avoid making the wrong decision. The test scores can tell us when things are going wrong, but not why. Data on Johnny's health, attendance at school, etc (parent factors) added to school safety, quality of materials and classrooms, breath of curriculum offered, etc (school district factors) AND the implementation of proven best teaching practices (teacher factors) will help in determining the why.

Fortunately, the teacher implementation of best practices is now easy to track. The eCOVE Observation Software is an easy to use tool that provides objective data on both the teacher's practices and student behaviors. Unfortunately, the process seems to be focused on just the outcome measures. When we match that up with teachers, it will do a great job of assigning blame, but not of much help in fixing the problem. We need all the data to do that.