The Influence of Grading Reliability and Grading Leniency Metrics on Student Evaluations of Teaching

• In Event: Faculty Evaluative Approaches

Abstract

Millet (2018) introduces the grading leniency and reliability metrics of teaching style. Millet
motivates the interpretation of grading reliability as an important dimension of instructor
effectiveness. The present study replicates Millet’s work for a new institution. This replication
finds that much of Millet’s results extend across institutions with some departures. Millet found
that grading reliability increased monotonically as leniency fell. In replication, grading reliability
peaks when leniency is near its institutional average. This study makes a significant contribution
by introducing an external mechanism to validate the interpretation of grading reliability and
grading leniency as important indicators of instructor effectiveness. Including Student
Evaluations of Teaching (SET), the study finds that grading reliability and leniency are important
predictors of instructor ratings.

Authors

• Celeste M Calkins, University of Nevada - Las Vegas

• John (Skip) Crooker, University of Nevada - Las Vegas

• Qingmin Shi, University of Nevada - Las Vegas