I can’t find the original Gadfly post that has Barnett Berry so upset*, but to the extent it was skewed, Barnett’s response is hardly any better. Apparently he’s back to his old any weapon to hand approach on teacher quality issues (here he is more recently playing fast and loose on Teach For America, again). Despite his protestations, the Harris – Sass study (pdf) Barnett discusses offered little in the way of great news about the effectiveness of National Board Certified Teachers. And, for Barnett to chide Gadfly for ignoring caveats while he touts the mentoring finding as powerful news is, as the kids would say, LOL funny, and provides a nice window into the method:
Relevant snippets from Harris – Sass on the study overall and then the mentoring part:
…The efficacy of NBPTS as a tool to improve student learning appears questionable. The pre-certification signaling effect might warrant both the identification and additional compensation of NBCTs, but the possibility that the effect does not exist post-certification is an important drawback. Also, while the extensive application process has the potential to raise the productivity of teachers, the results here provide little evidence of gains in human capital. In addition to the potential benefits, it is important to consider the substantial costs that go into the certification—teacher time, NBPTS administration and direct financial incentives—as well as other possible means of accomplishing the same objective of increased teacher effectiveness. NBPTS certification provides a distinctive mixture of certification, preparation and merit pay, but that does not necessarily make it a more cost-effective policy compared to other options
….There is some evidence of positive spillovers of NBPTS mentor teachers on other teachers, in both reading and math, but only with the FCAT-NRT student achievement test. Such positive evidence is not surprising in math where NBPTS mentors are more effective with their own students, but it is somewhat surprising in reading where such direct effects are not present. It is possible that the presence of any type of mentor induces teachers to spend more time improving their teaching. Therefore, the effects we estimate may be the effects of mentoring in general and not the effects of NBPTS mentoring per se. To our knowledge, this is the first study to present direct evidence regarding these spillover effects.
On Barnett’s point about choice of test, read the paper, important caveats there, too, on the research on this issue overall and this study. I still remain in the mend it don’t end it camp, but that’s increasingly despite rather than because of the National Board’s most vocal defenders.
*Update: Helpful and more web savvy readers have sent it along, you can read by clicking here.