Monday, August 19

Teacher evaluations need to factor in more than one indicator


Current system relies on single indicator, telling incomplete story about instructors' effectiveness

BY Jose-Felipe Martinez

There is no question that teachers should be evaluated much more thoroughly than they currently are.

The evaluation systems in place in most large school districts are bureaucratic rituals that serve no useful purpose for districts or teachers.

At the same time, there is reason to be skeptical about overreliance on student test scores for evaluating teachers.
In August, the Los Angeles Times published a database of about 6,000 elementary school teachers’ value-added scores, which are based on student performance on standardized tests.

Value-added models judge teacher effectiveness in terms of student progress over time, avoiding the obvious problem of comparing teachers who taught a very different mix of students.
However, this does not mean that we can evaluate teachers merely by looking at their value-added scores.

Experts agree that value-added models are not “causal” indicators of teacher effectiveness. The models do not say a teacher “caused” students to progress at a certain rate, only that students progressed at a certain rate on average.

Other aspects of teaching should factor in: Evaluation of instructional practice through observation, video or portfolios is needed to illuminate the “black box” and provide the critical formative component (value-added analyses don’t tell us what teachers need to improve on).

Measures of a teacher’s pedagogical knowledge are important.
Principals’ assessments seem reasonable to include, as are parental reports. Used in combination, all these sources of information would give us a much more solid basis to evaluate teachers and to help them improve their craft.

The rationale for value-added analysis as a policy instrument also needs to be discussed. Publishing value-added model estimates for individual teachers is misguided. There are concerns about the statistical models, teacher privacy, and standardized test scores as measures of student learning.
James Stigler, a UCLA psychology professor, makes an apt analogy to folks chasing the stock market.

Value-added scores can vary considerably across years and subjects; a teacher deemed “effective” in 2009 may rank lower in 2010, or vice versa.

It is also unclear how schools and districts are supposed to handle a deluge of parents fighting to get their child assigned to the “most effective” teacher.

I think we would all agree that teaching is a complex, multidimensional task. Consequently, evaluating teaching requires considering various sources of information.

While the notion that teachers are the least evaluated of public servants may be accurate, it is also true that most organizations do not evaluate employees on the basis of a single “effectiveness” indicator.

Even the reporters in the recent Los Angeles Times piece acknowledged the expert consensus that value-added analysis should never be used as the sole measure of teacher effectiveness, which makes their decision to then publicly identify teachers based on that one measure all the more puzzling and troublesome.

A comprehensive system of teacher evaluation would help improve the education all children receive.

It could be used to identify outstanding teachers. It could also help those who need to improve aspects of their practice, while offering a clear pathway for sanctioning and eventually discharging teachers who consistently underperform.

Jose-Felipe Martinez is an assistant professor of social research methodology at the UCLA Graduate School of Education and Information Studies.

Share on FacebookTweet about this on TwitterEmail this to someoneShare on Google+Share on Reddit

Comments are supposed to create a forum for thoughtful, respectful community discussion. Please be nice. View our full comments policy here.