Here's a front-page story in today's NYT that's actually, by and large, a thoughtful and balanced look at the Impact teacher evaluation system that Michelle Rhee set up in DC, designed and run by TFA alum and National Teacher of the Year, Jason Kamras.  My main problem with it is the silly, inflammatory headline, "Teacher Grades: Pass or Be Fired".

165 Washington teachers were fired last year based on a pioneering evaluation system that places significant emphasis on classroom observations; next month, 200 to 600 of the city's 4,200 educators are expected to get similar bad news, in the nation's highest rate of dismissal for poor performance.

The evaluation system, known as Impact, is disliked by many unionized teachers but has become a model for many educators. Spurred by President Obama and his $5 billion Race to the Top grant competition, some 20 states, including New York, and thousands of school districts are overhauling the way they grade teachers, and many have sent people to study Impact.

Its admirers say the system, a centerpiece of the tempestuous three-year tenure of Washington's former schools chancellor, Michelle Rhee, has brought clear teaching standards to a district that lacked them and is setting a new standard by establishing dismissal as a consequence of ineffective teaching.

But some educators say it is better at sorting and firing teachers than at helping struggling ones; they note that the system does not consider socioeconomic factors in most cases and that last year 35 percent of the teachers in the city's wealthiest area, Ward 3, were rated highly effective, compared with 5 percent in Ward 8, the poorest.

A few comments:

 

- The fact that the system has found that "35 percent of the teachers in the city's wealthiest area, Ward 3, were rated highly effective, compared with 5 percent in Ward 8, the poorest" isn't evidence that the system is wrong – it's simply revealing what anyone who knows anything knows: "good" schools (with the richest, whitest kids) get the best teachers and schools serving the most disadvantaged kids – the ones who need great teachers to have any chance in life – instead get the worst.

 

- This is exactly right, one of many measures needed to draw better teachers to tougher schools: "Washington's compensation system offers bigger bonuses ($25,000 versus $12,500) and salary enhancements in high-poverty schools"

 

- Gee, what a shocker that a lousy teacher thinks Impact is unfair:

Another teacher who expects to lose her job next month because of low ratings said at a public hearing that evaluators picked apart her seventh-grade geography lessons, making criticisms she considered trivial. During the most recent observation, her evaluator subtracted points because she had failed to notice a girl eating during class, the teacher said.

"I'm 25 years in the system, and before, I always got outstanding ratings," she said. "How can you go overnight from outstanding to minimally effective?"

The fact that this teacher got "outstanding" (let's be honest: "satisfactory") ratings for so many years simply highlights the outrageous absurdity that 99% of teachers everywhere are rated satisfactory, even when NO LEARNING is going on in the schools.

 

I don't know this teacher of course – maybe the Impact system did evaluate her unfairly, which raises an important question: no system is perfect – this is why I oppose the death penalty, because I'm not willing to accept even 1 innocent person being executed for every 99 guilty people who are – so what error rate are we willing to accept?  In the case of the ultimate penalty, I won't accept even a 1% error rate, but I'm certainly willing to accept getting rid of 99 horrible teachers, even if I knew that 1 good teacher would be unfairly fired.  I'd feel badly about that teacher, but he/she could surely get another job, if he/she is a good teacher – and even if this wasn't the case, the good for students of removing 99 horrible teachers far outweighs in my mind that harm to 1 teacher.  But what if the error rate were 5%?  10%?  20%?

 

These are tough questions, with no easy answers, made harder by the fact that NOBODY knows what the actual error rate of Impact (or any other system) really is.  My instinct is less than 10%, but I can't say for sure that that number is right.  But I do know for sure that the answer Randi, Ravitch, Strauss, etc. would give – 0% or at least less than 1% – is wrong.  They view everything through the lens of teachers/adults, so they consistently come up with the wrong answers for kids.

--------------------------

Teacher Grades: Pass or Be Fired

By SAM DILLON

Published: June 27, 2011

www.nytimes.com/2011/06/28/education/28evals.html