You may remember that when the New Teacher Project released its influential Widget Effect report last year, there was a bit of a mini-controversy (“Widgetgate”?) about its dismissal numbers for Toledo, one of the districts studied.
Briefly, the report found that Toledo had very low teacher-dismissal rates, like most of the other districts studied. But then the American Federation of Teachers put out a rather affronted-sounding press release stating that the peer-assistance and -review program in Toledo produces “much better results” than the report stated, and expressing concerns about the data collection.
So did TNTP lowball the dismissal numbers for Toledo by not considering teachers who were dismissed—or resigned—after failing to improve in PAR?
I started doing some digging around the issue and found that there were some inconsistencies in how the data were being classified. Since TNTP and the Toledo Federation of Teachers were engaged in a process of reconciling it, I decided to hold off on doing anything further until the final data were released.
Anyway, TNTP and AFT Toledo have finally audited all the numbers and come up with ones that, apparently, both parties agree on. You can find the full report here. But here are the highlights:
• The initial report found five formal dismissals of nontenured teachers and one of a tenured teacher between 2003 and 2008. The updated data show that two additional nontenured teachers should have been included in the report, for a total of seven. TNTP’s next publication of the data will include the updated figure.
• In addition, the district and the union found that five nontenured and five tenured teachers resigned during this time period (an “informal dismissal”). These data will be reported as a footnote in the next publication of the report.
• Eleven teachers who were dismissed either formally or informally through PAR were long-term substitutes, not full-time classroom teachers, which was the unit of record for the report. I’ll leave you to decide whether you think these teachers should be included in the overall dismissal rates.
Peer review, of course, is one of the features AFT has been promoting in a revised evaluation system. These data don’t really tell us a whole lot more about the relative merits of PAR versus other evaluation models. Even if the dismissal numbers aren’t significantly higher than in other districts, for instance, perhaps the quality of evaluation feedback is better. Perhaps teachers that might have faced performance difficulties have emerged much improved.
Here’s one of the subtexts both for PAR and for evaluation reform in general: What percentage of teachers each year should be identified as underperforming, offered remediation, or ultimately be dismissed for not performing up to snuff? The academic literature in this area is scarce and rather out-of-date, which makes this most complicated of questions even more difficult to answer.
For a longer look at some of these issues,
read my story from last fall.
A version of this news article first appeared in the Teacher Beat blog.