Establishing Multilevel Benchmarks for Small-Scale, Complex Examinations Using Objective Standard Setting

• In Event: Saturday Roundtable Session 11:25 am
  In Roundtable Session: Assessment Design: Diverse Principles, Benchmark Evaluations, and Validation Insights (Table 25)

Abstract

Assessments are an important part of the teaching and learning. Test results are routinely provided to teachers and students, but how useful are those reports? Do they provide the teacher with actionable information to inform instruction? This study was conducted to determine whether the establishment of multiple performance levels may be accomplished using a hybrid form of Objective Standard Setting (OSS), which maintains psychometric integrity whilst providing useful information to educators. The proposed hybrid model blends an OSS content-based criterion with a performance-level description defined through use of standard error (SEM) based confidence intervals. Results suggest that a hybrid approach retains the criterion-importance of OSS while simultaneously providing rich information about student mastery and performance, useful for adjusting instruction.

Authors

• Gregory E. Stone, University of Toledo

• Toni A. May, Binghamton University - SUNY

• Kristin L.K. Koskey, Drexel University

• Jonathan D. Bostic, Bowling Green State University

• Gabriel Matney, Bowling Green State University