Apparently it's a zeitgeist thing, because the same day I posted my reflection of teacher evaluation, Charlotte Danielson was also taking a look at the issue.
Danielson misses some points here and there (for instance, referencing "the Widget Effect," a pseudo report from TNTP that enjoys a life far beyond the value of its content), but she gets one point absolutely correct:
There is also little consensus on how the profession should define "good teaching."
And then there's the money quote that launched a thousand tweets:
I'm deeply troubled by the transformation of teaching from a complex profession requiring nuanced judgment to the performance of certain behaviors that can be ticked off on a checklist. In fact, I (and many others in the academic and policy communities) believe it's time for a major rethinking of how we structure teacher evaluation to ensure that teachers, as professionals, can benefit from numerous opportunities to continually refine their craft.
It's a sentiment that has been expressed by tens of thousands of teachers, but when it comes from the woman whose brand is on many an evaluation form, it gets attention. However, I'm a little less excited about her following paragraph.
Simultaneously, it's essential to acknowledge the fundamental policy imperative: Schools must be able to ensure good teaching. Public schools are, after all, public institutions, operating with public funds. The public has a right to expect good teaching. Every superintendent, or state commissioner, must be able to say, with confidence: "Everyone who teaches here is good. Here's how we know: We have a system."
The public has a right expect good teaching? Absolutely. We can prove the quality of the teaching with a system. I'm dubious. I am not a systems guy-- systems have huge value as long as we recognize their limits, but the minute we start imagining that a system can somehow do better than human judgment, I think we're in trouble. All systems are ways to impose the judgment of a few humans on the behavior of a large number of other humans, and therefor systems have the effect of moving judgment and decision-making further away from the place where the actual rubber meets the real road.
Danielson tosses out a number-- 6% of teachers are in need of remediation. That seems high to me, but it also seems made up, so we'll let that slide, because part of her point is that a personnel system should be built around the vast majority of non-sucky teachers. To her, that means focusing on "professional learning" rather than ratings.
She offers four truths about professional learning that need to be folded in. 1) It requires active intellectual engagement. 2) It can only occur in an atmosphere of trust. 3) Both challenge and support, and a career-long process of support which is never "finished" to which I say yes, yes, yes. I've said it a zillion times-- every good and great teacher I know can give you a list of things they still need to work on. 4) Policy makers must acknowledge that top down, butt in seats, assigned reading traditional PD doesn't do jack.
And she has some preliminary thoughts on what a personnel policy might look like:
* must id underperforming teachers and promote professional learning
* should include a step from probationary to "continuing status"
* differentiated, according to employment status, with different rules for novice teachers and different roles for experienced ones.
* experienced teachers should still be evaluated now and then
Peter Cunningham (top dawg at Education Post) connected some dots between my piece and Danielson's. He reduced my post to a list of possible purposes for evaluation:
- To find bad teachers.
- To find good teachers.
- To guide and support teachers.
- To compare teachers.
- To let the taxpayers know whether or not they’re getting their money’s worth.
- To give teachers a clear set of expectations.
- To make the complex look simple.
Into the mix, Cunningham throws Chicago writer Mike Dumke, who gets to ask the next logical question which is, if teachers don't want to be accountable for test scores, what do they want to be accountable for?
My first response is to suggest that we reframe the question-- instead of asking for the best ways for teachers to prove they're doing their job well, we would do better looking for the best way to find out if the teacher is doing a good job. It may seem like nit-picking, but to me, it's the difference between a boss who judges you by checking to see what you're doing and a boss who makes you stop doing your job so that you can go attend a meeting on doing your job.
Not all purposes on my list are created equal. #4 is useless and a waste of our time. #7 is destructive and also best avoided.
#3 is more important than #1 and #2, both of which assume that goodness and badness are static qualities in a teacher. They aren't. Some days I teach really well. Some days I teach okay. Some days I don't do well at all. Additionally, I teach some students far better than I teach others--not because of will or skill or desire, but for the same reason I am a great partner for one woman and a lousy partner for some others. Teaching is a human relationship, and beyond good/bad or right/wrong, we have the combination of two specific humans.
In other words, instead of trying to find good (or bad) teachers, we should be trying to help teachers teach well.
Are there teachers who fall so far on one end or the other that we can go ahead and slap a good or bad label on them? Sure (I have had both as student teachers). But most of us fall in the middle, always striving and working to do our job a little better.
So what should we look for? How would I answer my version of Dumke's question?
I've tried to answer this before, and someday I'll work out all the details and launch my million-dollar consulting business. But here's the basic process that I envision:
1) Bring together, in person or by technomeans, a huge number of various stakeholders from the community-- parents, grandparents, employers, graduates, elected officials, business leaders, students, teachers themselves. Give them a hefty, robust list of teacher qualities, skills and behaviors. Have them determine which ones to include and how much to weight them.
2) Now you have generated both a job description and an evaluation form (and, if you're lucky and your local leadership is good, you have also generated some good, thoughtful discussion about what folks want from their schools and their teachers).
3) Just as you have generated a custom evaluation format for your schools and teachers, you will need a custom method for evaluating the items that you have selected. Some may be, by their nature, easy to hit (eg if your community wants to include "students get good test scores," that one will be easy to cover). Many, by their nature, will be a matter of opinion and observation. If, for instance, something like "maintains a safe and supportive classroom environment" is chosen, you aren't going to be able to administer a testing instrument that generates a Classroom Environment Rating number.
However, one of my premises is that most of the people in the school community already know the things you want to find out. Students know who the nurturing teachers are and who the hard-but-fair teachers are and who the mean-and-scary-for-no-good-reason teachers are. Former students know, and they know what turned out to be useful in the long run. Parents know. Other teachers know. Most of what you want to know is already known-- what is needed is an instrument that taps into that knowledge.
Cunningham suggests that a student rating system would be susceptible to teachers essentially sucking up for a good review-- but I believe that if you are asking the students to rate specifics, you'll get far better results. In other words, "Is your teacher good?" is a question that will yield lousy results. But "Does your teacher help you understand hard things?" will get a better answer from students even if that student hates that teacher.
4) Do not, no matter how great the temptation, hire outside consultants to do the work. Maybe to manage the structure and data involved. And somebody will need to come up with a master list/menu. Perhaps that's how I'll finance my eventual retirement. But do not let non-stakeholders in the local system try to tell locals what they are supposed to want from their schools. The state will need to look over local shoulders to see what's being done and whether it is being done well and without tilting the table and with broad community involvement.
And do not rush. I am guessing it will take at least a year-- at least-- to get the first go-round ready to go. And then you'll have to periodically review it as you find gaps or issues or time goes by and your local stakeholder population changes.
Some folks will not like this system because it absolutely fails in terms of making schools and teachers comparable across lines, and it lends itself more to a descriptive, qualitative rating than a simple one-dimensional stack-and-rank. But it could generate real data of actual use to the teachers and district themselves. As a bonus, it could also elevate and involve the voices of the community, giving them a real feeling of involvement and ownership in their district.