How to grade a teacher

As the recent job evaluation of Los Angeles Unified Supt. John Deasy showed, test scores and other metrics can be a useful addition to the assessment process — as long as they’re not allowed to substitute for the bigger, more meaningful picture.

Progress should be measured. Data matter. But rigid adherence to them is counterproductive. Deasy had set lofty goals for himself in 2012-13, promising, among other things, to boost student test scores, graduation rates and attendance rates by specific amounts. But his bosses at the school board clearly understood that even if he didn’t meet all of his ambitious standards — and he didn’t — things were moving in the right direction and great effort was going into improving academic achievement. And that’s what needs to be measured, both for teachers and superintendents.

State and local teachers unions — including the California Teachers Assn. and United Teachers Los Angeles — have fought vigorously against using student test scores in teacher evaluations. And on the state level, the unions’ victory is nearly complete. California has missed out on federal grants and failed to receive a waiver from the federal No Child Left Behind Act because of its ongoing refusal to link test scores with teachers’ job performance. Faced with a lawsuit by school reformers, Los Angeles Unified has developed a mild system for including the scores in its teacher ratings.

It’s a shame that the issue has become so polarized, and it’s as much the fault of reformers as of unions. Because of an obsessive push by the Obama administration, some states have given far too much weight to test scores in measuring teachers’ performance. Parents know better. They and even their children are aware of who the beloved teachers are, the ones who have a gift for explaining and infusing lessons with energy and creativity, the ones who dispense encouraging words and extra help as needed and who instill in their students a desire to learn.

Test scores add to the picture by showing whether students have gained proficiency or lost ground. But their usefulness is limited and should be treated that way. They are better at measuring the strongest and weakest teachers rather than the many who fall in the middle. And little if anything can be gleaned from a single year of scores; it’s the trend over several years that reveals meaningful information about an individual teacher’s performance. In general, test scores should be used to help hardworking teachers improve rather than to fire them.

Randi Weingarten, head of the American Federation of Teachers, has favored the limited and responsible inclusion of test scores in teacher evaluations. Instead of rebuffing all talk about the scores’ possible uses, lawmakers should be devising policies to make tests and other data part of the process without overstressing their importance. Good teachers should have nothing to fear from a fuller picture of their skills that includes a few metrics.