At the annual meeting of Pennsylvania AFT, the leaders of the union called on the legislature to eliminate the test-based teacher evaluation system. Because of the inducements offered by Race to the Top, almost every state spent many millions to design a new teacher evaluation process, based on Arne Duncan’s insistence that such a system would weed out “bad” teachers. Behind that assumption is the wacky belief that bad teachers cause low test scores.
Last year, the first year of the new system, 98.2% of teachers were rated satisfactory or higher.
This year, 97% of Pittsburgh’s teachers were rated proficient or distinguished. The statewide figures for this year are not yet available.
“AFT Pennsylvania president Ted Kirsch said, “The law was based on a false narrative that low-performing schools exist primarily because of ineffective teachers, which is not the case. There are many factors involved in student success that are not given the proper weight under Pennsylvania’s new teacher evaluation system. The result is a system that gives high marks to educators working in well-funded schools with few disadvantaged students and penalizes teachers who take the tough assignments in under-funded schools with large concentrations of students from low-income families or with special needs or English language learners.”
“The release stated the delegates want a system that is “transparent and understandable by teachers and the community“ and is “primarily a professional growth system that supports teachers in their development and differentiates evaluation for new and experienced teachers to ensure that new teachers who are in need of support are not driven away.”
Diane Ravitch's blog 
Statistical Russian roulette witch hunt couched as valid science, boo.
They should read through the Teachers College analysis of VAM to use in their defense for a more valid form of evaluation. VAM formulae can be tweaked and twisted to get the type of the result desired by the evaluator. They are junk metrics that mean very little and do nothing to improve outcomes for students.
Demanding transparency is key to ending the worst abuses of “reform” because stuff like VAM can not withstand open scrutiny (which is precisely why the creators and users of VAMs refuse to release detailed methods, source code and data)
Even if it does not violate any laws or the Constitution (which is not at all clear), that teachers have been “evaluated” by such “black box” methods and in some cases lost their jobs as a result is highly unethical.
These teachers are placed in a Kafkaesque situation in which they can not defend themselves and therefore have their due process rights violated.
The craziest thing about VAM is that there is this silly predictability formula which tells what the computer thinks your test scores will be. It becomes part of your value added score too. I had wonderful test scores last year, but I had to take a three year average which brought my score down because of one “down” year. To make it even more painful, I had a 91 percent passage rate on this “down” year. It is so unfair. I am so thankful to be close to retirement. This battered profession is not worth the horrible stress. I will always do my best for my students, but I will never look back. Value added teachers have it so hard. Just because a student may fall from a 468 to a 462 on two different tests on two different years, does that make me an incompetent teacher?