The final Institute of Education Sciences report is out on the success of comprehensive induction services modeled after those of the New Teacher Center and the Educational Testing Service compared to standard district induction fare. If you have a sense of déjà vu about the report, it's explained by the fact that two reports on the study have already been released with first year and second year results.

Contrary to the findings in the two earlier reports, a positive effect on student performance on standardized reading and mathematics tests was noted, but only for teachers who had two years of comprehensive induction. Even so, the intensity of induction (which varied considerably within and between treatment groups) does not seem to correlate with student progress. It's disappointing that comprehensive induction didn't show a straightforward "treatment-response" relationship, where the intensity of induction was correlated with the magnitude of student progress.

Moreover, consistent with the findings that were reported earlier, there were no differences among the three groups of teachers in terms of attrition rates.

One finding provides grounds for caution about educational social engineering: selecting mentors who matched teachers in terms of race/ethnicity or grade-level appeared to reduce retention rates, rather than raise them as one might hypothesize.

Although not a resounding endorsement, the final report's finding that students of teachers who had two years of induction did benefit academically is significant news and considerably different from the findings of the two earlier reports. This prompted our general suggestion last month that interim results not be broadcast until the completion of a study. We received an immediate rejoinder from Steven Glazerman, who is coincidentally the Mathematica project lead for both this study and the study on Chicago's TAP program we covered last month. The full text of Glazerman's letter to the editor appears below.