A fascinating study and article by the LA Times discusses the role of value-added analysis in teacher evaluation.  This is a controversial subject, at least among teachers unions and politicians.  The role of value-added evaluation is presented most often as an input in to firing decisions, with the question being one of proportion relative to other inputs in to the evaluation.  This does education measurement a disservice.

A section of the article contrasts two math teachers at the same Los Angeles area elementary school.  One teacher's students regularly show significant year-to-year increases.  The other teacher's students regularly show significant decreases. As the article describes,

"Told of The Times' findings, Smith expressed mild surprise.

'Obviously what I need to do is to look at what I'm doing and take some steps to make sure something changes.'"

A lot of the discussion in education right now is about "getting rid" of bad teachers as part of school reform efforts.  Without question, more effective teachers are key in increasing student achievement. But unless I'm missing something, there aren't exactly thousands of qualified people out there trying to get in to the teaching profession, but being turned away for want of an opportunity.  Firing all the bad teachers just shifts them elsewhere.

Measurement can be used in many ways.  But I've yet to see an organization benefit greatly from using it punitively.  Measurement's highest and best use is in the identification of opportunities for improvement and in guiding the development of solutions.  While some teachers probably don't belong in the classroom, I have to believe that the greatest benefit of the increasing use of value-added data in education is in the training and development of current and future teachers.  Teachers themselves recognize this.  It would be nice to see less public discussion of how data can be used to remove bad teachers and more about how to turn bad teachers in to good ones.