What Should Teacher Evaluations Look Like? Ideas to Consider

Recently, I received an anonymous comment on my blog post entitled “Merit Pay in Education: An Exercise Manipulation and Futility.” I have purposefully chosen to not publish anonymous comments on this blog because I believe firmly that if you have something to say, then you should be willing to divulge your identity. Part of your message is who you are, and hiding your identity is actually hiding part of your message. At any rate, the commenter who calls him or herself “Engaged Parent” seemed to practically dare me to publish his or her comment. I won’t publish it as a comment, but I will share it here with my own commentary because there are some fundamental misconceptions apparent in that comment. Here’s “Engaged Parent’s” comment, then I will respond.

Even though you're going to NOT publish this comment I think it's worth sending it to you anyway. A teacher you are yes? I'm going to assume that.

Over and over again I see the same comments on how we can measure the success of education. It's great that we keep hearing how a merit program doesn't work but what's the alternative? It would be nice (for once) to hear a teacher tell us how a "bad" teacher can get filtered out of the system. I'm sure you will agree that not all teachers are good, there are always some bad apples in the bushel, no way around it. 
I agree, test scores are not always a good judge of how a teacher is doing but it does or it at least should give key indicators to a teacher that maybe what they're doing isn't the best for that particular mix of kids and that another approach might work better. But do they do that? I don't know.
Why are teachers the only ones who don't have to be put under the scrutiny of evaluation? The rest of the world has to go through it.

What I really want to say is I think that a happy medium could be met if careful thought was put in to it.
For instance:
10 % what did your students think of you?
15% what did your students parents think of you?
25% test scores
10% peer scores
15% self score
25% principal scoresIs that not a fair assessment? If not then come up with SOMETHING because personally I'm sick of hearing how teachers don't want to be assessed.

First of all, let me say that yes, I am a teacher, and I don’t think I’ve hidden that fact anywhere on this blog. You can find that information on the blog sidebar. I am currently a principal. As a teacher though, I am more concerned with your logic and misconceptions in your comment than anything else. First of all, you seem to suggest that merit pay should be implemented because we have no alternatives. Following this argument means that the rationale for implementing merit pay lies, not in whether it will work or not, but because we don’t have any other alternatives. That in itself is faulty reasoning. I certainly hope my physician doesn’t implement a treatment or my mechanic doesn’t simply initiate a car repair because he or she says, “What’s the alternative?” without considering the evidence of symptoms and lab test results that identify the problem.  You are hearing criticisms of performance pay because it has been researched and has failed to bring about the improvements sought, which is increased student learning. It has been tried in public education, even in North Carolina with no appreciable effect on the quality of education.

You next assertion seems to be thinly veiled when you state “It would be nice (for once) to hear a teacher tell us how a “bad” teacher can get filtered out of the system.” It appears you have a belief that “bad” teachers never get dismissed. I would agree with you that not all teachers are good, which I gather you really mean effective at helping students learn among the many other things teachers do. But contrary to your belief, they do “get filtered” out of the system, at least in my experience. So your belief that teachers who don’t do their jobs very good somehow are not subject to dismissal is another misconception. Teachers can be dismissed fairly but with “due process,” which means that I as an administrator must thoroughly document my rationale for doing so. Many administrators see “due process” or tenure as many call it, as an obstacle, but due process rights were put in place because historically, our education system was notorious for political firings and reprimands. Teaching historically has been quite political with school board members or even administrators firing teachers for noxious reasons. Teachers have been fired for being pregnant or even so a school board member can then hire a son or nephew. At any rate, it is a misconception that teachers who aren’t doing their jobs can’t be fired. It simply takes leadership and willingness to first try to help that teacher improve, then take the steps necessary to counsel them into another profession.

Still another misconception from you comment is that teachers aren’t under the “scrutiny of evaluations.” I have been an educator in North Carolina for 25 years, and I have always been subject to evaluations, so the idea that teachers aren’t somehow evaluated just isn’t true. Teachers in my state have been evaluated for years, and some are dismissed as a result of those evaluations. When you hear all the criticism of current evaluations systems, those do not come from the desire to avoid evaluations; it comes from the desire to have those evaluations be fair. You yourself acknowledge that test scores alone are not always a “good judge” of teacher, but there are much deeper issues with using test scores as a part of evaluation, such as the fact that more 70 percent of courses in a high school do not have these state tests. Teachers that I know aren’t trying to avoid being evaluated as you suggest; they simply want those evaluations to be fair and just, as any employee in any line of work would want.

Now, let’s take a look at the evaluation system you suggest. You provide several interesting sources of evidence for teacher evaluations. First of all, you would base 10% of the evaluation on “what students thought of the teacher.”  I hope you aren’t suggesting that students rate the teacher as a person. I suspect you are really suggesting that students are somehow surveyed on what kind of job they think their teachers are doing. This is reasonable in some ways, and I’ll agree with you. But there obviously has to be careful attention paid to the survey instrument so that its questions get at the heart of instructional prowess and not opinions about the teacher personally. The question then becomes, what do students know about a teacher’s instructional ability? The answer to that question could be used as a basis for survey questions. Our school has used student surveys for over 5 years, and one quite common problem is that the responses can sometimes be about anything other than the teaching. The data is useful, however, because it can help us make changes where there are genuine complaints and issues.

The second source of evidence for teacher evaluation you suggest is “what the parents think of their teachers.” Again, that seems reasonable, if what you are suggesting is a parent survey that focuses on a teacher’s ability to teach. For obvious reasons, the survey would need to move beyond asking the parent what they thought of the teacher as an instructor. But there are some issues that would need to be ironed out. For example, one issue that would need to be dealt with would be the fact that few parents witness the actual teaching teachers do in the classroom. They get most of their information about what happens in the classroom secondhand. I can tell you as an administrator that quite often  that secondhand information isn’t entirely accurate. A student often goes home and tells a parent one story about something that’s happened in their classroom or to them or what a teacher has done. That parent next calls the principal, clearly angry, until they hear what really happened. Using parent opinions about teacher practice would be especially difficult since they do not witness teaching in action; they have to rely on hearsay, which as we know is unacceptable in court rooms. Parent surveys could focus, however, on the parts of teaching that they do directly witness, such as teacher to home communication. At any rate, if parent surveys were to be used in teacher evaluations, they would need to be much more than simply asking a parent what they thought of his or her teacher.

The third source of evidence you suggest for teacher evaluations are test scores. That also seems reasonable until you try to implement that practice which we’ve done in North Carolina.  The issues are many. Some of the tests are inferior and are of questionable quality. Then there’s the fact that not all subjects are tested, which leads to questions like: Do you evaluate only those in tested areas, or do you develop and administer tests in every single subject area, which is quite costly in terms of time and money? In addition, some tests were designed in a way that they make poor choices for use in evaluating teachers. Tests like ACT and SAT were created for an entirely different purpose rather than measuring teacher quality. Additionally, tests designed for assessing student achievement have never been proved valid for assessing teacher quality. Other issues with test scores used as part of teacher evaluations? How do you separate the effects of previous years teachers on this year’s teacher scores? The list of issues with using tests as evidence of teacher effectiveness is lengthy, that’s why teachers and myself question the practice. A test is a test is a test is how much of the world sees it, but most of us who’ve been in education know tests aren’t as simple as we wish they were.

The fourth source of evidence for teacher evaluations that you suggest are “peer scores” which I suspect you actually mean peer ratings. Peer ratings seem to make sense too, after all, who would know better than a teacher’s peers whether or not they are effectively teaching. But the problem with this one is similar to the parent one: there are few teachers who actually know how effectively a teacher is teaching, unless of course you set up a system by which they peer observe. That would work, but it offers logistical problems such as finding time and means for each teacher to observe each other in action. Currently, we use peer observations in evaluations as a part of our process so your suggestion isn’t far from current reality. There is one additional issue with peer ratings too; teachers are often very reluctant to honestly rate their peers because quite often, they have to work with these individuals, and they are often their friends. You have to be pretty naive about human nature to see peer ratings working very well.

The fifth source of evidence for teacher evaluations you suggest are “self scores” which I would suspect you mean self ratings. Believe it or not, we already ask our teachers to do “self-assessments.” While these aren’t directly connecting to a teacher’s final rating, they are used as a part of the evaluation process early in the year. Teachers use them to honestly look at how they stand with our teacher evaluation, and many use it as a basis for their professional development plan. Of course it’s like any self-assessment; it’s only as useful as the amount of honestly employed in its completion. The main problem with self-ratings is if high stakes are tied to it in any way, how noble do you think someone would be to give themselves a poor rating if it could impact their job status?

Your final source of evidence for teacher evaluations you suggest are “principal scores” which I take to mean principal ratings. In North Carolina, this is practice. Principals complete a summative evaluation of ratings on teachers at the conclusion of every year. This is in turn used to suggest professional development for the next year. Principal ratings are certainly not without issues as well. Most of us have worked for bosses who were tyrants and completely incapable of making fair and partial judgments. Because of this danger alone, principal ratings have issues too.

I would say “Engaged Parent” you do get one principle right in your evaluation suggestion; we do need multiple sources of evidence for teacher evaluations. The issue is simply deciding which ones will effectively improve instruction for our students. We do need multiple sources of evidence to evaluate teachers, but if we choose the incorrect evidence, then our teachers don’t teach well and what we set out to do which was improve student learning never happens. We can’t afford to get teacher evaluation wrong because in the end our children will suffer.

So, “Engaged Parent” we do have evaluations of teachers. We also use much of the evidence you suggest above, and there’s talk of adding other sources as well. Teachers as a rule in North Carolina do not complain about being evaluated. Like any employee they do complain about evaluated unfairly, and that is a great deal of the criticism you’re hearing. I would suggest that since you have a great interest in teacher evaluations, you might want to read a book called Evaluating America’s Teachers Mission Possible? by W. James Popham. Popham goes to great length to argue what a good, fair and valid teacher evaluation would look like. He even goes through about every single source of evidence you suggest and points out the problems with them and suggests how those problems might be alleviated. It’s a good read for anyone, educator or non-educator, looking into the evaluation of teachers.

NC to Test Every Subject K-12 and Tie Teacher & Principal Evaluations to Test Scores

In a meeting this past Monday I attended, representatives from the North Carolina Department of Public Instruction provided educators with a presentation describing how our state is adding a sixth standard to our teacher evaluation and an eighth standard to our principal evaluation directly tying those evaluations to test scores. What I discovered at that meeting was that the standards proposed are worded innocuously and can hardly be questioned. For example the teacher standard reads:

"Teachers contribute to the academic success of students. The work of the teacher results in acceptable, measurable progress for students based on established performance expectations using appropriate data to demonstrate growth."

I would think just about all teachers hope that what they're doing is contributing to the academic success of their students. The big difference in opinions among educators, however, is perhaps what this "academic success" is and whether growth measured by a test score is accurate. The principal standard is also written in this hard-argue-against language:

"ACADEMIC ACHIEVEMENT LEADERSHIP: School executives will contribute to the academic success of students. The work of the school executive will result in acceptable, measurable progress for students based on established performance expectations using appropriate data to demonstrate growth."

Both of these standards make sense on the surface. There's no educator alive who would argue that teachers and principals are not responsible for the achievement of their students. In my years as an educator, there's not a day that passes where concern about whether our students are learning what we're asking them to do isn't on my mind. The problem is not with these standards, but it is with the North Carolina Department of Public Instruction's interpretation of what "student achievement" is. Under their plan, student achievement = growth on standardized test scores. While that is a nice, neat simplification of what achievement is, it ignores all learning that can't be tested with standardized testing. By interpreting student achievement as test score growth, the state is simply making testing in North Carolina even higher-stakes than before. With this test-emphasis, we will be well on our way to becoming test-prep factories that don't churn out educated students, but excellent test-takers. North Carolina is going to subject students to tests early and often, not to measure student progress, but to measure how well the teacher is doing. This betrays an underlying, but mistaken belief that tests can be used to tell you how teachers and principals are doing. North Carolina has defined "effective teaching" and "effective school administration" as simply growth demonstrated (by whatever model they can create) by test score performance.

Ultimately, the meeting I attended was billed as an opportunity for educators like myself to provide "feedback" to North Carolina Department of Public Instruction personnel on this proposed teacher evaluation change. However, in practice, it seemed more of a here's-what-we're-going-to-do-to-you session, but we want to give the "appearance that we're listening to educators." (That's a tactic perfected by Arne Duncan.) During the session, when educators expressed concerns about what the state is planning to do, they were often cut off by DPI personnel who interrupted to defend the state's plans rather than sincerely allowing educators to voice their concerns and listening. They felt the need to stifle honest opinion by cutting off people while they were speaking. Instead of being an opportunity to express concerns, it was an opportunity for the state of North Carolina to try to summarily dismiss those concerns.

To this educator, there are two equally frightening things about this whole test-centered approach to teacher and principal evaluations that I and other educators at this meeting tried to bring up, but the North Carolina Department of Public Instruction representatives quickly let us know that we were wrong.

• First of all, North Carolina is planning to evaluate teachers and administrators  by using test scores this spring even though they have not decided what growth model they are going to use, nor do they know exactly how this evaluation is going to take place. What's worse, they will not know at least until February of this year when the State Board has an opportunity to hear the first reading. In typical bureaucratic fashion, they are rolling out a program and implementing it even before it is finalized and clearly defined. Historically, I can recall two other examples of North Carolina Department of Public Instruction initiatives that were rolled out before they were defined well. These were when the state during the 1990's rolled out the ABC Accountability Model, and in the early 2000s, when the state rolled out the NC WISE student data system used throughout the state. Both of these were rolled out without any deep thought on its practical application in the districts. They were both rolled out bugs and all, and classroom teachers and educators had to suffer while the state got its act together. Now, it appears our State Department of Public Instruction is doing it yet again, except this time there are much higher stakes tied to it. North Carolina will be evaluating educators' careers based on something that isn't even clearly defined, and won't be until three-fourths of the year is completed.
• Secondly, North Carolina is also planning to create tests for every single subject taught, administer these tests, and use the results, not to see how students are doing, but to see if teachers and principals are able to raise test scores. Recently, Charlotte-Mecklinburg proposed this "test-everything-that-moves" approach, and it blew up in the administration's face so bad, they had to take money from Bill Gates and the Broad foundation to hire public relations personnel to try to sell it. If something smells so bad that you have to repackage it to sell, then perhaps there's something fundamentally wrong with it. In spite of the fact that educators at this meeting expressed concern over how much more time we will spend testing under this proposal, and how problematic it will be to implement, the Department of Public Instruction personnel at this meeting dismissed these concerns outright repeatedly, and told us how wrong we were. 

As an administrator and educator, I understand the need for accountability. I understand the need for testing to see how our students are doing. I even cynically understand why our state is doing this. They are not doing it because it's what's best for our kids, because how could testing students' every move be beneficial? No, the state of North Carolina is doing this so that they can keep their Race to the Top money, plain and simple. As a 22 year veteran educator, I've seen education measures come and go. Most of those have been benign and simply were discontinued, and no one noticed their passing. However, this time I'm afraid it's going to be quite different. The state of North Carolina is using a program that is not fully defined yet, and a strategy that "tests-everything-with-a-pulse" that is going to destroy public education in this state. It is going to turn our schools into the test-prep factories that Diane Ravitch has spoken about so eloquently so many times.

Note: Here's the link to the presentation used if you would like to see it for yourself.