Wednesday, April 9, 2008

Use test scores for tenure? Not a good idea, with these bumblers

So the NY State budget finally was decided, with all the proposed cuts to NYC schools restored, and the promise of CFE maintained, yet all the newspaper editorial boards and bloggers can do is to blather about a provision in the budget that prohibits the use of student test score data in making teacher tenure decisions.

Eduwonkette provides some of the links to the bloggers who are so outraged as to contend that this is the end of the civilized world.

Actually, the final language in the budget bill was a reasonable compromise, in which it was agreed that there will be a two year moratorium while a commission considers how best this information can be utilized to inform tenure decisions.

Evaluating a teacher’s competence on standardized test scores alone is not sufficient, since the gains or losses that any class achieves in scores is often highly erratic from year to year, is partly based on factors such as class size which is quite variable across NYC schools, and the background of students in each class.

Actually, research shows that its not just the current class size that helps determine the rate of learning, but a student's past class sizes, which can change the entire trajectory of his or her academic career.

And what are they going to do about the fact that many of the tests are given in the middle of the year? The DOE's proposed solution is to give last year's teacher half the credit, but that assumes equal effectiveness of all teachers -- which is contrary to the whole point of this exercise - that some teachers are more effective than others.

Moreover, test scores do not tell the whole story. Other evidence of a teacher's skills and value are equally if not more important, including her ability to motivate students, keep them engaged, and guide them in their writing, their projects and all other types of creative learning that cannot be assessed by test scores alone.

Most importantly, it is by now abundantly clear that this statistically illiterate administration cannot be trusted to use this data carefully and intelligently, with a grain of salt and in relation to other critical factors, given their record on merit pay and school grades.

In both cases, they chose to base the results primarily (85%) on test scores, with more than half based upon the essentially unreliable gains or losses in scores over one year alone.

Tying tenure to test scores could have very destructive effects, discouraging teachers from taking on classes of struggling or special ed students, and lead to a further loss of morale, with even more test prep replacing real learning.

A hiatus of two years is a terrific idea since whatever is decided will be implemented by a new administration that will hopefully be more trustworthy with the use of such data. We know that the bunch of bumbling amateurs in charge of our schools now would never be able to figure out how to balance all these factors in an intelligent, humane and constructive fashion.

For more on this issue, including comments from Chancellor Klein, Randi Weingarten and me, see the Channel 2 report here.

1 comment:

  1. This law is not about whether test scores should or should not be used for tenure decisions. The Union simply succeeded in making it more difficult for school adminstrators to dismiss new teachers before they gain the heavy protections of tenure. Any time the Legislature puts an evaluation factor off limits, you give the dismissed employee more ammunition to draw out a long legal battle for the job. Every fired probationer will now argue that the principal's decision was really based on test scores. I understand that tenured employees earn such heavy legal protection but non-tenured probationary employees -- like most of us in the working world -- should be subject to dismissal without cause. Yes, you may lose a few talented probationers but, on the whole, the system benefits from allowing administrators to make non-challengable decisions with new employees and avoiding the expense and inefficiency of protracted litigation.

    ReplyDelete