![Sample our Education journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5932/TOC-Subject-Sample-2021-Education.jpg"})
Abstract
This study was designed to assess the impact of providing teachers and principals with performance feedback on instructional practice, principal leadership, and student achievement. The teacher and principal performance feedback that was the focus of the study’s intervention incorporated promising features of educator performance measures identified by recent research. The impact of the 2-year intervention was assessed through a school-level randomized experiment conducted in 127 elementary and middle schools from eight districts. The study found that providing educators with performance feedback had positive impacts on teachers’ classroom practice, principal leadership, and student achievement in mathematics but not in reading.
Notes
1 Perhaps the most important of these policies was The American Recovery and Reinvestment Act of 2009, which contained the Race to the Top competition. For the Race to the Top competition, the Great Teachers and Leaders area accounted for the largest share of points (138 of 500), mostly for systems that measure educator effectiveness and for human resource policies that use such information in tenure review, nonrenewal decisions, and other decisions. The U.S. Department of Education’s (Citation2010) blueprint for the reauthorization of Elementary and Secondary Education Act (ESEA) emphasized the same approach. Because of challenges with passing a reauthorized ESEA, the Obama administration pursued these same goals using the Flexibility Waiver provisions of the existing ESEA.
2 The authors noted that some teachers did not undergo TES on schedule and explained why that was unlikely to bias their results. (See especially page 3644 and Footnote 18 in Taylor & Tyler, [Citation2012].)
3 In years when teachers did not undergo a TES evaluation, they received a perfunctory annual evaluation with nearly all teachers receiving a “passing” rating. Essentially teachers were not evaluated in non-TES years (see Footnote 7 in Taylor & Tyler, Citation2012).
4 Under the checklist evaluation system, probationary teachers were evaluated annually, and tenured teachers rated “excellent” or “superior”—ratings received by nearly all tenured teachers—were evaluated every two years (Steinberg & Sartain, Citation2015).
5 This distribution of effort was intended to engage principals in the implementation of the performance measure without overburdening them. Using multiple observers to rate the same teacher was also likely to produce a more reliable end-of-year average rating, compared with using a single observer for each teacher (Ho & Kane, Citation2013).
6 Teachers of grades K-3 in treatment schools also participated in the study. This was done mainly to promote schoolwide engagement in the implementation of intervention components focusing on feedback on classroom practice and principal leadership. These teachers were not included in the main study analyses, however, because student assessment data needed for the feedback on student growth based on value-added scores are not available for grades K-3. Moreover, data needed to assess the intervention’s impact on student achievement are not available for grades K-2.
7 Since the videotaping focused on teacher actions rather than individual students, we were not required to obtain parental consent for the video-recorded classroom observations. However, we did give parents the opportunity to opt their child out of the videotaping. Students whose parents opted them out were not removed from the class during the recording, but were instead seated in a section of the classroom that the camera did not capture. The number of such students was minimal and did not differ significantly by study condition.
8 The different aspects of classroom practice are officially referred to as “dimensions” in the CLASS system and “components” in the FFT system. For simplicity, we use the term “dimensions” for both systems in this article.
9 Note that the inter-rater reliability of video-recorded lessons pertains to the amount of measurement error due to coders in an individual teacher’s CLASS/FFT scores for one observed lesson. The CLASS/FFT scores also contain measurement error due to the sampling of lessons to observe. Both sources of error are largely averaged out when group means based on individual teachers’ scores are compared. For the impact estimates on teachers’ classroom practice based on comparisons of group means, the relevant metric for measurement errors is the standard error of the group mean difference rather than the reliability of an individual teacher’s score based on one or two observed lessons.
10 According to Hill et al. (Citation2008), the average annual gain in mathematics is about 0.42 standard deviations for students in grades 4–8. Thus, an impact of 0.05 standard deviations translates to about 0.05/0.42 or 0.11 of a year’s achievement gain. Assuming a 36-week school year, this implies that the impact corresponds to about four weeks of learning.
11 Our achievement impact models only controlled for prior achievement in the same subject area as the outcome. As a sensitivity analysis, we re-estimated the impact models using prior achievement in both reading/ELA and mathematics as controls. The estimates were similar in magnitude, but the Year 2 impact on mathematics achievement became statistically significant (p < .05).
12 It is also worth noting that the 15 percent of the treatment teachers did not attend the training webinars.
13 This finding mirrors the finding from the New Teacher Project which revealed that more than 99 percent of teachers were rated satisfactory in districts using binary ratings and 94 percent received one of the top two ratings in districts using a broader range of ratings (Weisberg et al., Citation2009).