...A few things. But not everything.
IMPACT, Washington D.C.'s teacher evaluation system, has received a lot of press lately. There's
Sam Dillon's piece, which gives the perspective of a terrified first-year teacher afraid for her job, and then
Bill Kerlina's confessional in the Washington Post about why he, a young principle, quit his job.
I've never been a Michelle Rhee fan (IMPACT is her baby) but I'm here to defend it, kinda. First, some background. IMPACT is an evaluation system that distinguishes between teachers with and without "value added student achievement data," a.k.a., 4th through 8th grade reading and math standardized test scores.
"Group 2" teachers are those for whom a definitive value added statistic cannot be computed because they teach something other than 4th through 8th grade reading or math. So these are the social studies teachers, science teachers, lower elementary reading and math teachers, high school reading and math teachers, etc. According to the
IMPACT handbook, Group 2 teachers are evaluated by in-classroom observations (75% of the total score), non-value added student achievement data (10%), value added student achievement data (5%), school/community involvement (10%), and professionalism (evaluated qualitatively). There are five in-classroom observations per year, three by the principal and two by master educators.
Common criticisms of IMPACT are as follows:
(1) People don't like the merit pay component.
(2) People don't like the fact that IMPACT scores can be used to fire bad teachers
(3) Support is not given to struggling teachers, only punitive measures
(4) Test scores are a factor, especially for Group 1 teachers
(5) No forgiveness/leniency for teachers in high-needs schools
As for (1) and (4), my main complaint is that the pet program of people who frequently use a battle cry of "Not Research Based!" to disparage their opponents lacks a basis in research, i.e. it is Not Research Based. In short, I'll be sympathetic to the critics of merit pay until it's proven that merit pay actually improves educational outcomes for low-income students. (And I do mean ACTUAL educational outcomes. I don't want no
inflated-through-mindless-test prep faux-achievement achievement, I don't want no
cheating scandals.)
But, (2) and (5) don't really bother me. Having seen and heard stories of teachers who were retained year after year, given two billion second chances to learn classroom management, etc., I have to think that it takes much more thought and effort to fire a teacher than to retain a teacher, even if said teacher is really, really bad. As for (5), it makes my skin boil to think that we should have lower standards for teachers in high needs schools than for teachers working in high-income neighborhoods. More is at stake for the students in high needs schools--we need better teaching, which means high standards for instruction. It wasn't clear from the Sam Dillon piece as to whether the teachers who were censured for failing to notice their students eating and texting during their lessons were in high needs environments. But, it shouldn't matter. It makes no difference how dazzling your history lesson is if students are engrossed by other things. If you're a teacher, you have to devote half your attention to what you're saying (which is why you have to know that material darn well) and half your attention to what your students are doing. While outlining the four causes of the American Revolution, or fielding a question, or explaining the assignment, or cold calling students, or whatever, the teacher should be circulating throughout the classroom, and without ceremony, taking the cell phone, or the sandwich, and slipping it into his/her pocket. (Yes, it's gross to put a sandwich in your pocket. So goes the life of K-12 teacher.)
Let's talk about (3). Lack of support was main complaint of Bill Kerlina, who to be quite frank, came across as a major crybaby. But, I contend that the IMPACT system is a step in the direction of more support, and better support, more so than its alternatives, which include the status quo. I really, really like IMPACT's emphasis on in-classroom observations. It forces principals to get into the classroom (the eternal temptation for school leaders to be beholden to paperwork in one's office), and also, creates a network of expert teachers who can see precisely what teachers are doing well and poorly. Test scores don't tell you much, and sometimes give false information about a teacher's quality. But, it's pretty hard to fake it for an entire lesson. Short of threatening your students with corporal punishment on a to-be-named date in the future if they don't behave, it's impossible to abracadabra a structured, well-functioning classroom. Such is only the product of months of deliberate, dedicated work. Perhaps right now, mentor teachers are coming into the classroom, filling out their forms, and leaving without giving teachers much by way of feedback. But, the first-year science teacher in Dillon's article certainly sounded like she got feedback. And the teachers who allowed texting and eating? Sounds like they got their feedback too, except they didn't like it.
In short, unlike standardized tests, in-classroom evaluations answer a ton of questions about teacher performance. Are they well-prepared? Are they able to simultaneously deliver material and manage student behavior? Are they interesting to listen to? Do they explain concepts clearly? Are they providing students with age- and developmentally-appropriate material? How's their rapport with their students? How receptive are they to student questions? Do they have ways to meaningfully gauge whether their students are learning the material, or is their style all lecture/homework? Are they accommodating students with special needs? Do they seem like they know what they're doing? Yes, most likely the system needs to be honed. Maybe make it so that low-performing teachers have ways to be in contact with their mentors--according to my calculations, mentor teachers evaluate about one teacher per day, so there should be plenty of time for extra support. But the movement from standardized test-based evaluation to in-classroom evaluation can't be a bad thing....
Except that this is not quite the trajectory. Remember when I said that there were Group 1 and Group 2 teachers? Group 1 teacher evaluations are 50% value added student achievement data, 35% in-classroom observation, etc., etc. According to the IMPACT brochure, the goal is to shift from a Group 2 evaluation framework to a Group 1 evaluation framework:
Over the next few years, we will be implementing developmentally appropriate standardized assessments for students in kindergarten, first grade, and second grade. We will also be adding end-of-course exams for secondary English, math, science, and social studies. As these assessments are rolled out, more teachers will be moved from Group 2 into Group 1.
If things unfold as proposed, I officially retract all compliments, and predict a bleak future of
cheating,
cheating, and
more cheating. Excuse the cheesy turn of phrase, but a Group 2 evaluation framework is the only way IMPACT can make a positive impact.