Validity of Observation Ratings in Teacher Evaluations

IESThe Institute of Education Sciences (IES) released a new study recently on the validity of evaluation ratings using an instrument adapted from a well-known teacher observation tool, the Danielson Framework for Teaching. The findings of the study, by Regional Educational Laboratory (REL) West, support the use of a single summative rating to evaluate teachers, but do not support the use of domain or component ratings for evaluation.

This validation study analyzed 2012/13 ratings of 713 teachers in the Washoe County School District in Reno, Nevada. The observation instrument was adapted from the Danielson Framework for Teaching, which groups 22 components of teaching into four domains. While previous studies have examined only a portion of the Danielson Framework, typically focusing on classroom practices, this study examined the full Framework.

This study examined four assumptions underlying Washoe County School District’s proposed interpretation of ratings from the classroom observation instrument:

  • Ratings from the classroom observation instrument differentiate among teachers.
  • Each of the four domain ratings measures a single, cohesive area of teaching practice.
  • Each of the four domains is distinct from the others.
  • Ratings from the classroom observation instrument indicate teacher effectiveness in promoting student learning.

The assumptions were identified through logical analysis of the district’s proposed uses of the observation data. Some of the assumptions were supported by the analyses, and others were not.

The overall findings show that the four domains do not appear to measure distinct aspects of teaching but do support using an average rating from all components. This average rating shows a moderate relationship with student learning, providing some evidence that it may be interpreted as an indicator of teacher effectiveness.

For more detailed information about the effectiveness of the Danielson Framework and its effectiveness, see the full report.