Willingham: 3 key factors in teacher evaluation (beyond the hype of value-added)
Cognitive scientist Daniel Willingham, a psychology professor at the University of Virginia, follow up on his look last week at the Los Angeles Times project that evaluated teachers by using test score data. Willingham is the author of “Why Don’t Students Like School?”
By Daniel Willingham
Last week I criticized the Los Angeles Times for publishing a story that labeled teachers as “highly effective”or “poorest-performing” solely on the basis of a value-added model of student test scores.
I mentioned that the reliability and validity of value-added models is controversial.
Setting aside for a moment the need for a reliable and valid measure, what other considerations ought to come into play when thinking about teacher evaluation?
First, an evaluation system ought to be motivated by one’s goals for schooling. What are we hoping students will get out of school? If teachers are to be held responsible for student outcomes, the sensible thing to do is to begin with the target outcomes, and then use assessments appropriate to those outcomes.
Some target outcomes will likely be hard to measure: that students will have a positive attitude towards learning, for example. But if we begin by listing everything we would ideally want to measure, we will at least have a clear idea of what we can put in an evaluation measure and what we cannot. And we may find ways of satisfactorily measuring constructs we had not considered measurable.
If we start with a test because we have it around for another purpose, we’re doing things backwards. We’re starting with the measure, and then saying “this is important to schooling.”
Second, I think that evaluations should vary for teachers of children at different ages. To my mind, it’s unreasonable to point one’s finger at the teacher of a high school senior and say, “Dan didn’t learn much math this year. That is inevitably your fault.” Where is Dan’s responsibility in this scenario?
I think it is much more reasonable to suppose that a kindergarten student is not responsible for his or her learning. Whether the parents bear some responsibility might be open for discussion.
Third, an evaluation system must confront the issue of the imperfection of any evaluation system.
To take the simplest example, suppose your evaluation amounts to a categorization: “teachers good enough to stay” versus “teachers bad enough to fire.” Two types of errors are possible: retaining bad teachers and firing good teachers.
Either mistake is costly, but you can’t withdraw from evaluation for fear of making a mistake. Even if you do nothing, you’ll still be making mistakes. If you fail to evaluate teachers at all, you’ll never fire a good teacher, but you’ll retain bad teachers.
That’s close to where we are now.
You can’t eliminate errors but you can choose the proportion of each type of error committed. You do that by varying the threshold for firing.
The proportion of each type of mistake that you elect to make would be based on your evaluation of how costly each type of mistake is.
Even with a reliable and valid measure, I’d be leery of labeling teachers “highly effective” or “poorest performing” with a system that did not account for these other three factors.
But history ought to tell us that rushing to a new system would be a mistake. There is not a lot of patience in education policy for the continual refinement of policies. They are put in place and are deemed a success and retained, or deemed a failure and jettisoned.
If a new teacher evaluation system is above some threshold of perceived effectiveness, the basic structures will likely be calcified into practice and unchanged for years, even if it could be improved. We’ll rush on to tweaking some other part of the system.
If the new system is below that threshold, people will recoil from the idea of evaluating teachers (it will be the No Child Left Behind of its day) and it will be that much harder to get any effort off the ground for at least a decade.
There is unprecedented momentum for teacher evaluation. It would be a pity to waste it.
Follow my blog all day, every day by bookmarking washingtonpost.com/answersheet. And for admissions advice, college news and links to campus papers, please check out our Higher Education page at washingtonpost.com/higher-ed Bookmark it!
| August 23, 2010; 11:00 AM ET
Categories: Daniel Willingham, Guest Bloggers, Teacher assessment, Teachers | Tags: daniel willingham, factors in teacher evaluation, how to evaluate teachers, no child left behind, teacher evaluation, teacher evaluation systems, teachers and value added, value added systems
Save & Share: Previous: How to give classrooms a mission
Next: The most expensive public school in America
Posted by: johnt4853 | August 23, 2010 11:56 AM | Report abuse
Posted by: WorriedParent | August 23, 2010 12:56 PM | Report abuse
Posted by: bsallamack | August 23, 2010 2:10 PM | Report abuse
Posted by: bsallamack | August 23, 2010 2:22 PM | Report abuse
Posted by: hainish | August 23, 2010 4:21 PM | Report abuse
Posted by: educationlover54 | August 23, 2010 4:40 PM | Report abuse
Posted by: bsallamack | August 23, 2010 6:21 PM | Report abuse
Posted by: WorriedParent | August 24, 2010 1:04 PM | Report abuse
Posted by: bsallamack | August 24, 2010 3:27 PM | Report abuse
Posted by: TGriggs1 | August 25, 2010 9:43 AM | Report abuse
The comments to this entry are closed.