Loads of writing has passed since last week’s agreement in New York on teacher evaluations. See the following for a very brief sample:

• The extent to which local bargaining would still determine evaluation specifics.
• Andrew Cuomo’s defense of 40% from test scores in comparison with other states.
• Carol Corbett Burris’s argument for why the formula agreed to puts test scores in charge of the entire evaluation.
• Diane Ravitch’s criticism of the agreement.

Then other stuff continues to roll in on teacher evaluations, from the latest on Tennessee’s creaky implementation to Aaron Pallas’ argument that use of value-added scores must include error bands and the next month’s Kappan article on teacher evaluation. The New York state and the New York City agreements each have attracted interest on one point each.

First, the New York state Department of Ed press release sparked quite a bit of attention with the following claim:

Teachers rated ineffective on student performance based on objective assessments must be rated ineffective overall.

I’ve read through the relevant legislation, and I think that Corbett Burris has the gist right on mechanics: the state is mandating that a teacher get at least 4 out of 40 points from the test-based parts of evaluation to avoid an ineffective rating. Will that have as much of an effect as she worries? Because the locally-based part of this (half of the 40 points) will be developed in collective bargaining, it should be possible to craft local agreements on evaluation that essentially operationalizes “your students showed you didn’t completely suck” in a way that is reasonable to achieve. That doesn’t mean it’s guaranteed, but it should be possible, and this is the sort of practical experimentation that is much better than what Florida and some other jurisdictions have mandated.¹ I could be very picky about methods, but we need to see what develops in practice. If I were a teacher, I might be more worried about needing to achieve at least 25 points out of 60 on the non-test-based part of evaluation to avoid an ineffective rating…

The second point of conversation has been the agreement that UFT can appeal a minority of evaluations to three-person panels when the union suspects the evaluation is based on non-performance grounds (harassment, etc.). This was part of a larger agreement that is probably more important in practice: after an ineffective rating, a teacher is provided a “validator” for the next year to observe the teacher three times. This validator’s ratings can either agree or disagree with the principal’s evaluation, and disagreement with the principal does not change the prior evaluation but keeps the burden of proof on the city in a termination hearing. If the independent validator agrees with the principal, then the burden of proof switches to the teacher in a termination hearing. This package is a big deal: limited appeal of evaluations, validation, and the switching of burdens in formal hearings. It is a big deal in labor relations for an employer to allow a binding process that can change evaluations. It is a big deal in education to have evaluations in schools that can trump (and potentially embarrass) a school principal. It is a big deal to craft a mechanism that balances due-process protections against evaluation procedures.

There are huge questions about the implementation of both the state and the local agreements. At the state level, local agreements shaping the 20% from local achievement measures are going to be crucial. Is it impossible for certain types of teachers (and teachers of certain populations) to achieve good ratings? And don’t forget the 60% that isn’t mandated to be tied to student achievement measures; those will be hard, too. If there is evident unfairness in many districts, then the state agreement will have failed.

At the local level in New York City, implementation will show whether the balancing of interests in the agreement works. From a union standpoint, I like the ability to appeal even a small proportion of evaluations, as it provides a significant check against abusive supervisors. The validator/burden-swapping mechanism is a check on everyone, and it will be hard to argue that teachers are always against accountability when the largest teachers union local in the U.S. just made an agreement that said, “K-12 teaching is not guaranteed for life; we want due process and a check on unfair evaluations.”

What is not mentioned in the agreement is questions of practice — in particular, the practice I have heard about in NYC where principals delay permanent-status decisions for arbitrary reasons (usually framed informally as insufficient test scores). But if a principal is screwing around in that way, I suspect there are now mechanisms to make visible what principals would prefer to keep hidden. How UFT uses them is going to be a fascinating thing to watch.

Notes

1. I also don’t think Corbett Burris’ point is politically viable: “you just need to get 10% of the available points; how is that unreasonable?” is hard to argue against. [↩]

Related posts:

1. A short review of Teacher Evaluation 2.0
2. Combining qualitative and quantitative evidence for teacher evaluation: What does “predominant” mean?
3. Proposed ground rules on teacher evaluation and test discussion