The primary test of the new teacher evaluation systems implemented throughout the nation over the past 5-10 years is whether they improve teacher and ultimately student performance. Although the kinds of policy evaluations that will address these critical questions are just beginning to surface (e.g., Dee and Wyckoff 2015), among the most important early indicators of how well the new systems are working is their credibility among educators. Put simply, if teachers and administrators don’t believe in the systems, they are unlikely to respond productively to them.
A new report from the Institute of Education Sciences (IES) provides a useful little snapshot of teachers’ opinions of their evaluation systems using a nationally representative survey. It is important to bear in mind that the data are from the 2011-12 Schools and Staffing Survey (SASS) and the 2012-13 Teacher Follow Up Survey, a time in which most of the new evaluations in force today were either still on the drawing board, or in their first year or two of implementation. But the results reported by IES might still serve as a useful baseline going forward.
The primary outcome in this particular analysis is a survey item querying whether teachers were “satisfied” with their evaluation process. And almost four in five respondents either strongly or somewhat agreed that they were satisfied with their evaluation. Of course, satisfaction with an evaluation system does not necessarily signal anything about its potential to improve or capture teacher performance, but it certainly tells us something about teachers’ overall views of how they are evaluated.
The IES authors use models to examine the association between this satisfaction outcome and several teacher and school characteristics. Perhaps unsurprisingly, they found a statistically discernible and meaningful association, all else being equal, between evaluation satisfaction and teachers’ evaluation ratings – i.e., teachers who received the highest ratings were more likely to express satisfaction with the process. Moreover, relatively few teachers reported receiving the lowest ratings (roughly three percent of the sample), and so it is reasonable to expect that satisfaction might have been attenuated had ratings assumed a more even distribution.
There was also a discernible association between satisfaction and an index of principal leadership, which was constructed using several separate survey questions. Teachers who expressed more positive views of their principals were also more likely to indicate being satisfied with their evaluation processes. This association might be interpreted in terms of good leadership translating into effective implementation of evaluations, though such causal inferences are not tested directly.
One additional noteworthy association was found between teachers’ satisfaction with evaluations and whether those systems included test-based productivity measures. Specifically, teachers who were evaluated based in part on test scores (about one quarter of the sample) were less like to express satisfaction with the process. It is certainly plausible to interpret this negative association as indicative of teachers’ skepticism toward test-based measures, such as value-added and other growth model estimates (though, again, causal inferences should be made with caution).
Finally, it bears noting one additional finding, which is actually the lack of any association between evaluation satisfaction and virtually all other characteristics besides the few mentioned above. In other words, satisifaction was not discernibly related to any teacher characteristics, including experience or grade level taught, nor to any school level characteristics, such as the proportion of students eligible for subsidized lunch assistance or on special education plans.
Overall, then, this analysis relies on a potentially useful but necessarily limited outcome (satisfaction with the evaluation process), and the data are from a time when most of the new teacher evaluation systems were either still being designed or had just come online. Still, most of the attention paid to the “results” of these new systems has been on the distribution of ratings they yield, whereas an equally important early result is how teachers view the new system. Going forward, it will be very interesting to see these kinds of results, both nationally and in different jurisdictions.