Mark Weber aka blogger Jersey Jazzman is a veteran teacher and a doctoral candidate at Rutgers University.
He wrote an open letter to a state senator in New Jersey who was angry that Governor Phil Murphy reduced the stakes attached to PARCC testing in relation to teacher evaluation.
State Senator Ruiz mistakenly believes that evaluating teachers by test scores is sound practice. She is wrong.
Weber reviewed the research demonstrating the invalidity of such measures.
By the way, New Jersey is one of the very few states that still mandates the PARCC tests. It originally was offered by 24 states. Only six states and DC still are in that small group.
You might find this to be a valuable resource for understanding why it makes no sense to evaluate teachers by the test scores of their students.
Diane Ravitch's blog 
A good letter and a good analysis. I would add that the use of high stakes testing has a dramatic impact on classroom atmosphere and activity that is not beneficial for student learning. I’m happy with the change in New Jersey though would submit that it doesn’t go far enough.
For Mark Weber.
An excellent letter. If you don’t mind I will share it with a member of the Ohio School Board who is running for re-election and is clearly clueless about the liability risks to the state and districts who still evaluate teachers using EVASS and SLOs. The liability issue may get attention in places where the reasoned arguments (and your wonderful graphics) do not. Also, check VAMBOOZLED for another effort to “rescue” VAM and capture more teachers in that web of deceit.
The Findlay and Dayton Mayors who scheduled and hosted the Gates-sponsored dog and pony show that included Gates’ “Impatient Optimist”, John N. Friedman (co-writes with Chetty), and the right wing
Hoover Institute’s Hanushek, facilitated policy makers’ manipulation. The Mayors, as a mea culpe, should give each state board member (1) the “Layman’s Guide to the ‘Destroy Public Education Movement’ “, by Tultican.com and (2) the Rand report that found Gates’ villainthropy was a $1 bil. failure.
When Sinclair Community College and OSU invited the Gates Foundation to speak at their graduations, they showed the taxpayers
the same respect that ECOT showed the taxpayers.
The whole process of trying to make teachers responsible for test scores is based on bias and false assumptions. Trying to statistically account for all the variables in the teaching-learning process is impossible. Any models that claim to do so are fake and inaccurate, and the motives are political, not academic.
The question I ask, and I don’t have a firm answer (though I have my gut feeling) is this. Even if the tests were accurate, speedy, and reliable, would it be beneficial to use them for high stakes purposes?
Question meant to be philosophical. I can feel the barrage of comments that it’s not possible for tests like these to be accurate, speedy, and reliable.
Not that I disagree with them!
I believe you answered your own question.
An accompanying question might be thus: would it be beneficial to use them for any purpose? If you had a truth machine that would guarentee the evaluation of the truth or falsehood of a situation, would it be good for society to employ that machine? If I have a child whose performance is abysmal, is it good that I make him understand just how much of a failure he is or would it be better for me to try to move him in a more positive direction with soft words that cushion the blow of harsh truth? Is baseball better now that we have instant replay? Congress better for the presence of Cspan?
Let’s assume that the tests are accurate speedy and reliable.
Let’s also assume that we can go faster than light, produce a perpetual motion machine and turn water into wine.
Wouldn’t it be nice?
NO!
If you agree with Campbell’s law, then it is not possible to create an accurate high stakes test. The very nature of “high stakes” will cause many people to game the system through direct test prep or other methods.https://en.wikipedia.org/wiki/Campbell%27s_law
On a philosophical level, I would like to add that if high stakes tests were accurate or reliable, or speedy, Campbell’s Law would still apply. When test scores become the goal, they lose their value.
Sorry, I meant to reply to Ohio Algebra II Teacher.
Oh wait, I did. Dang little phone screen messing with my eyes.
The real, evidence-based case against linking standardized test scores to teacher effectiveness lies not in the nuances of invalid and unreliable methodologies (VAM, SGOs. SLOs) but in the undisputable four (?) year record of FAILURE of these methodologies to improve the very metric that has been the obsession of amateur meddlers. So they have had the accountability system of their choice and it has had zero affect on test scores, zero affect on managing teachers, and a significantly negative impact on classroom climate a student teacher relationships. Why should any student have to work for teachers or schools?