What If Teachers Designed Teacher Evaluation?

[Sharing a letter I wrote to my TLN friend and recent guest blogger, Susan (Ernie) Rambo].

Hey Ernie,

I'm fairly impressed with the work of the work of our colleagues in the Denver New Millennium Initiative on their new report on teacher evaluation. The report couldn't have come at a better time for me, as I'm involved with work on teacher evaluation right now on several levels both within my state and nationally. This is one of the most articulate and most provocative set of teacher views on this topic that I've seen so far. Clearly, the group is working with an eye to the political realities of their own state, but also with their hearts towards what's best for students, and their thinking grounded in the wisdom of practice informed by carefully analyzed research.

For example, the four issues (p.6) they identify as critical to developing an effective teacher evaluation system, to me highlight the practical wisdom teachers bring to this discussion, especially, item #2: Qualifications and training of the evaluators. Their treatment of this point later in the report reminds me of how evaluators are trained for the National Board of Professional Teaching Standards scoring process, and I suspect they drew heavily from that model.

As teachers, the group also recognizes the dual role of an effective evaluation system (p. 10), and this is a point that needs more stress in the current education reform discussion. Schools and districts cannot use a "fire-and-replace" approach to reach the goal of consistently high quality teachers for all children. Helping teachers to identify strengths and weaknesses and providing ongoing support to address those weaknesses and share those strengths should be the primary focus of our teacher evaluation systems. Doing so would not only lessen the possibility of incompetent teachers remaining in the profession, but significantly reduce the amount of mediocre, overly scripted, or trial-and-error teaching occurring in schools across the country.

Also, as a member of the Teaching 2030 TeacherSolutions team, I especially appreciate the Denver group's suggestion about the role of "professional guilds" in the teacher evaluation process. One of the reasons I pursued National Board certification myself was to get serious feedback from highly accomplished peers on my work, and their suggestions for how I could improve my work. I love Vinnie Basile's projection (p.10) of how such a guild might work. This is one of the key roles I envision for teachers' unions as they reconfigure themselves to serve our profession in the new century.

I wonder how many policymakers or others will understand just how powerful The Goal Attainment Process described on pg. 8 is for actually reaching the goal of putting a quality teacher in every classroom? The suggested evaluation process would require teachers to demonstrate a critical professional pedagogical skill: the ability to develop, analyze, and use assessments at the classroom level. In this step, teachers would have to demonstrate how they connect what's going on in individual classrooms to the larger school and district goals. Also, the suggestion for a statewide assessment database (p.8) of teacher made assessments and how to handle the use of that database, particularly in the non-standardize tested areas, is brilliant and squares with practices from our most advanced competing nations in the area of education.

Of course, the most contentious issue around teacher evaluation is the use (or not) of student standardized test data to determine teacher performance. I've noticed how many teachers not only don't oppose, but expect the use of student test data as part of their professional evaluations. This contradicts many of the media-touted education reformers who claim teachers are afraid of these tests for our own selfish reasons. However, like most teachers, the Denver NMI also point out the very real dangers and limitations of excessive testing and over-reliance on test data, even as they (and I) look forward to "more robust student assessments" (5). The group seems to take the position that neither the tests nor the data is the problem, but rather how both are being misused.

The group's response to current attempts at using value-added measures in determining teacher performance, also, shows how thoughtful and fearless these teachers are. Despite the clearly documented flaws in these measures, the group doesn't advocate throwing out test data completely, but rather calls for more appropriate and effective ways to use it, and they emphasize that can only be done in partnership with teachers. It's a professional and principled stance; after all, standardized tests have been around a long time and some educators have found them more useful than others. Unfortunately, such student-centered logic may be hard to hear over the racket of racing to reform and the cheering of a billion-dollar testing industry.

Just as I finished writing this, I noticed fellow TLN blogger Bill Ferriter had posted his take on the report, so I hope this signals the start of much more discussion both within TLN and around the nation on the ideas Denver NMI has put forward. As many states and districts rush to "do something" about teacher evaluation, it's truly important that professional teachers step up and bring much needed light to this heated debate. Can't wait to hear what you see in the report, and where you think it might take us.

Best,

Comments

[Mike_Bostock]

It surprises many of us in the UK that teachers' careers can be put at risk by simplistic assumptions about the link between effective teaching and pupils' test scores. If all schools had an identical intake then we might expect differences in its outputs to be representative of differences in school provision. But if not, then we no longer have a point of reference, i.e. there is no constant in the implied mathematical formula that is being used.

In the UK we have similar problems in comparing the work of schools. The UK schools' inspection agency, Ofsted, largely use criteria to judge the work of schools and teachers that are fair and sensible, but there are still some rough edges around evaluating pupils' achievement in relation to their starting points. Worthy efforts to 'contextualise' the achievements of disadvantaged groups of pupils have recently been played down in recognition that some schools get great results with these same disadvantaged groups of pupils. The big problem lies in finding how to recognise the work of teachers and schools who educate pupils where "success is harder to achieve" (Dr John Dunford, 2010) yet succeed in doing so, for these will be the very best schools in the land. UK school 'League Tables' rarely show this. Schools that select the most advantage pupils inevitably populate the top layers of such tables. But a new category, 'Outstanding schools', is proving more effective in recognising the successes of schools working in different contexts.

I have written about some of the issues associated with using school performance data in my blog/book at www.mikebostock.wordpress.com. In our work with over 200 UK secondary schools we have developed some helpful approaches that allow schools to recognise the good work that their teachers do, and for teachers themselves to be able to gather evidence of the impact of their teaching. So what have we learned that might help the debate that is taking place in the US?

Well, as a first step we need to evaluate teaching separately from its impact, using objective criteria. Then we should evaluate the impact of teaching by making a series of comparisons, including comparisons with pupil achievement in similar schools. Then we should us common sense logic to equate one to the other. There are no simplistic formulas to be found to do this last step.

A significant technique that we have developed is looking at 'Within School Variation'. This is a big topic and an even bigger issue for UK schools because school variation, or the 'inconsistency in the impact of provision', is higher in UK secondary schools than in any other OECD country. The technique works because it has found a constant. It would take space here to explain, but it is possible to look at school variation in regards to a constant measure for 'average teaching', a measure that will apply to any type of school.

Methods which support Intelligent Accountability will mark the next step in the development of teacher and school evaluation. Teachers are sitting on the prime source of evidence to explain the impact of their work. The message here is to seize the initiative on using performance data. Quality assurance is a story that schools should tell for themselves.