How Do We Rate Teacher Effectiveness?

US Secretary of Education Arne Duncan is changing his talking points from last month's KIPP charters, Teach for America, and Value-Added Data Tracking Systems. At least, that seems to be the case, judging from his recent interview with the Washington Post.*

A snippet, followed by a question I hope you'll address in comments:

Duncan said he will encourage states to adopt achievement standards that give a clear picture of whether U.S. students are prepared to compete with global peers. And the funding will help states create better tests to show whether students are on track for college.

Duncan said the Obama administration aims to support performance pay to reward good teaching, individually and schoolwide. Beyond standardized test scores, Duncan mentioned classroom observation, parent and student surveys, and attendance as ways to help rate teacher effectiveness.

Question 1: In your view, what would these "ways to help rate teacher effectiveness" look like?

I'll share my thoughts on that.

First, it may surprise some of you to hear that this teacher wishes performance was rewarded. I've worked with enough complacent types to know the frustration that comes from seniority-only pay and privilege. So, for starters:

We’d have to define “merit” to include instruction for a broader set of skills than rote memorization: analysis, synthesis, collaboration, evalutation/critical thinking, problem-solving, creativity. This is not the opposite of the “fact-based, right/wrong, multiple choice” testing that NCLB and the College Board/AP/SAT pushes, but what you might call the upward extension of it. Mastery of facts is the beginning, not the end, of the assessment for meritorious teaching and learning. (click "read more" to continue...)

If we start there, that means teacher merit is measured by the types of projects that are assigned in the classroom - not by the standardized testing industry - and by the performance of students who complete these projects. This further means that said teacher measurement is performed not centrally, but locally - or perhaps by boards consisting of local and central judges. (I know that “central” is vague.)

My thinking is that if teachers were rewarded for designing learning activities that measured positively against a checklist of such higher-order thinking traits - and crucially, that the measurement was based not on a single unit, but on a portfolio of all units assigned throughout the semester or year (this eliminates the dog-and-pony show liability of single principal evaluations) - then the best teachers would be rewarded with higher pay, while the worst ones would have an incentive to change their practice for the better. Teaching to the test wouldn’t be the goal any more; teaching to higher instructional standards would be.

As for what those higher instructional standards would look like, we need look no further than Linda Darling-Hammond for answers. Her presentation lays the groundwork for such guidelines.

Your thoughts?

--

*For a less generous take on Duncan's WaPo interview, see the Daily Howler.

Image on Flickr by Dean Shareski