Friday, October 18, 2013

Fifth grade teachers were good predictors of their fourth grade test scores

From Problems with the use of student test scores to evaluate teachers by Richard Rothstein, Helen F. Ladd, et al. The authors are questioning whether value-added modeling (VAM) of teacher contribution to improved learning outcomes is well enough grounded for determining pay and promotions. As evidence:
A study designed to test this question used VAM methods to assign effects to teachers after controlling for other factors, but applied the model backwards to see if credible results were obtained. Surprisingly, it found that students’ fifth grade teachers were good predictors of their fourth grade test scores. Inasmuch as a student’s later fifth grade teacher cannot possibly have influenced that student’s fourth grade performance, this curious result can only mean that VAM results are based on factors other than teachers’ actual effectiveness.

VAM’s instability can result from differences in the characteristics of students assigned to particular teachers in a particular year, from small samples of students (made even less representative in schools serving disadvantaged students by high rates of student mobility), from other influences on student learning both inside and outside school, and from tests that are poorly lined up with the curriculum teachers are expected to cover, or that do not measure the full range of achievement of students in the class.
When you first read it, this sounds like a death knell to the credibility of VAM. But on reflection, I am not so sure. It depends on what they were "controlling for" in terms of "other factors." If a school has a disposition to channel children (and I suspect most schools do, whether explicitly or not), then as a child is identified as academically talented, they will be channeled to more demanding teachers the next year. If children are randomly assigned to teachers from one year to the next, then Rothstein et al's test is a good indicator that there is something amiss. If, on the other hand, there is any sort of channeling going on, explicit or not, then it seems to me that you should be able to find a correlation between the VAM quality of later teachers and the results achieved in earlier grades.

It is a long report that is skeptical of the use of VAM. I scanned the rest of the report to see if the issue of channeling (and its effects on precasting) was addressed anywhere else, but I could not find any reference. I am left to conclude that what was presented as a damning piece of evidence against VAM is actually just a data point that lacks context and may not be indicating what they want it to say. I.e. it seems like there is manipulation of information to support a conclusion already in place.

No comments:

Post a Comment