This is the first of what will (I hope) be weekly posts to this blog, each one aimed at exploring a relevant issue in schools, whether ANCS or all schools. I’ll do my best to watch for any comments and respond as I can.
This week’s post is on the new teacher evaluation system that will be required for all public schools–including charter schools–starting in 2014-15: the Teacher Keys Effectiveness System (TKES). I won’t get into all of the nitty-gritty details here, but if you are interested, the Georgia DOE has posted a great deal of information on its website. Last week I attended a three-day training on TKES facilitated by the DOE, and, in my opinion, there are some fundamental flaws in this system that parents and all who care about schools should know about just as much as teachers should because of the potential impact it will have on what gets taught and who will teach it.
First, though, there are some positives to TKES. A significant portion of the evaluation system is focused on giving teachers feedback in ten different “domains”, such as professional knowledge, instructional strategies, academically challenging environment, and communication. Good schools have mechanisms in place for teachers to continuously reflect upon and receive feedback on these elements of good teaching–for ANCS, our process of support and evaluation does not look much different than this aspect of TKES and you won’t get much argument from me on attempting to make this the norm for all schools.
The real issues lie in the other half of TKES. You see, each teacher will eventually receive a score–known as a Teacher Effectiveness Measure (TEM)–that is comprised 50% of data from observations of teaching practice and student surveys and 50% of student growth data. It’s this student growth measure that is much more problematic than fancy DOE powerpoint presentations make it out to be. To be fair, the Georgia DOE–like many other states–has developed TKES in rapid response to meeting criteria set by the U.S. DOE in exchange for millions of federal “Race to the Top” dollars. That cash-strapped school districts are being asked to meet a seemingly higher bar for desperately-needed funding than the criteria private banks and mortgage lenders who are partially to blame for these school districts being cash-strapped in the first place had to meet for bailout money is a blog post for someone else–the fact of the matter is that the student growth measure soon to be required in teacher evaluations is problematic in at least two main ways:
Growth measures attempt to simplify something that is just not that simple: It would be nice to think that by taking a student’s scores on the CRCT in two consecutive years and running them through an advanced statistical model we would then be left with a precise measure of how much of an impact that student’s teacher had. Common sense would tell you, though, that there are numerous factors that might influence a student’s performance on a standardized test and his growth (or lack thereof) from one year to the next–living situation, conditions on test day, influence of other teachers, and on and on. And it’s not just common sense, but plenty of research that confirms the complexities involved in calculating growth and value-added measures and warns against using them for a large portion of high-stakes purposes like a teacher’s evaluation.
Defining growth narrowly emphasizes some skills and knowledge to the exclusion of others: True story: all teachers in Georgia will be required to sign an assurance that “all standards within the course are taught with the appropriate level of time/emphasis. No course standards are taught to the exclusion of other standards for the sole purpose of [student growth measure] attainment.” That the DOE feels compelled to make teachers agree to such a statement is telling. It suggests that individuals, when faced with one element of their annual reviews being tied so closely to one type of data, might do all in their power to improve that data. If that’s not the case, then why set the system up that way? Maybe it’s because there is a recognition that this would result in schools becoming very dull places. Not only will teachers of “tested subjects” given a student growth measure, but so too will teachers of the arts, P.E., music, and the like. These teachers will be measured against how far students “grow” on a pre and post-test in their subjects. So, for example, a performing arts teacher will be measured on how many more parts of the stage her students know at the end of the year, or a P.E. teacher will know he’s made a difference in the lives of young people by whether or not more students can identify the proper form for a push-up by the time the class ends. Those are actual measures from actual assessments used last year as a part of a TKES pilot. I would certainly hope teachers of these classes would spend most of their teaching working to inspire students to learn how to act or how to become more physically fit. But, it appears as if their “effectiveness” will now get to be tied to a multiple-choice test that measures none of students’ growth in those skills.
So what then could be done instead? I offer a few thoughts, each of them readily available and supported by others in education:
Don’t try to boil teacher performance down to a single score: At ANCS, we do not give students numeric grades. Instead, we evaluate their performance against standards and give them meaningful, actionable feedback based on their performance. I’m not sure why we would do any differently for professional educators. In light of the statistical concerns cited above, a singles “TEM” score would seem to tells us little about that teacher and is, quite frankly, patronizing.
Include other measures of student achievement and growth in teacher evaluations: Is there a place for growth measures in assessing teacher effectiveness? Certainly, and especially if the concerns above are more clearly understood and accounted for. But what about incorporating student performance on real tasks of writing, reading, critical thinking, and performing? Such assessments have their own limitations but are much more aligned to what we hope our students get out of school. At the DOE training I attended, the facilitator said that Georgia received $400 million to develop the TKES system. Are you telling me that we couldn’t have developed a more comprehensive system that measures student growth in multiple ways with that kind of money? States such as Minnesota, New York, and Vermont have had success in developing authentic student performance tasks that measure student achievement across schools.
Support teacher effectiveness in ways that have been proven to make a difference: If the whole point of this evaluation system is to increase teacher effectiveness, then it should really be coupled with aggressive efforts to improve the working conditions of teachers. A number of large-scale research studies have shown that using student growth measures in evaluations and attempting to incentivize teachers with monetary bonuses for student growth have negligible to no real impact on improving teaching and learning in schools, while making schools better places for teachers–stronger leadership, positive school climate, appropriate instructional resources–and supporting collaboration among new and skilled teachers are highly predictive of better teaching and student achievement.
In the end, if we want to improve teacher effectiveness, we need policies and systems that are thoughtful and nuanced. To do otherwise will likely lead to truly effective teachers leaving the profession.