Nov 7, 2018
Nearly every public school teacher in the United States receives some form of feedback based on classroom observations, and these assessments are used by policymakers and administrators as indicators of teacher quality and input for professional development purposes. The TNTP Core Teaching Rubric, introduced in 2014, is unlike others in that it is scored based on student behavior rather than teacher actions. In this report, RAND researchers assess whether the TNTP Core Teaching Rubric produces scores that are representative of teachers' overall instructional practices and whether raters' content expertise influence scores on TNTP Core.
For this study, volunteer teachers of mathematics and English language arts (ELA) from more than 20 districts or charter school networks across eight states had their classroom instruction videotaped three times. This sample was complemented with a sample of approximately 100 4th-and 5th-grade teachers drawn from the Measures of Effective Teaching study for whom videotaped instruction was available. Raters trained in the TNTP Core Teaching Rubric then scored the videos from both samples. Analyzing the raters' scores, RAND researchers found a modest relationship between teachers' TNTP Core scores and student achievement gains in ELA, and no statistically or practically significant relationships between TNTP Core scores and math achievement gains. Further, raters struggled to agree on their judgments about the quality of a lesson, suggesting a considerable amount of uncertainty about the extent to which TNTP Core scores represent teachers' overall instructional practices. The report includes a number of recommendations to improve the use of TNTP Core.
Design of Study
Consistency, Bias, and Reliability
Extrapolation and Predictive Validity
Conclusion and Recommendations