Paper-Pencil and Computer Adaptive Test Formats: Comparing Student Psychometrics on a Mathematical Problem-Solving Assessment

• In Event: Virtual Poster Session
  In Virtual iPresentation Only: Division H - Section 3: Assessment in Schools Virtual Poster Session

Abstract

Assessments are a critical pedagogical tool used in education but often require significant testing time and effort. Our study examined whether a computer adaptive version of an existing paper-pencil mathematical problem-solving test could improve delivery efficiency and student psychometric performance through a more targeted testing experience. Results showed improved student test score reliability, higher separation, and smaller standard errors of measurement compared to the static test, with reduced administration time. CAT also functioned effectively using dichotomously scored constructed-response items rather than multiple-choice questions. Findings provide evidence that CAT can enhance score precision and efficiency for this mathematical problem-solving test, further supporting broader CAT applications in educational measurement contexts.

Authors

• Toni A. May, Binghamton University - SUNY

• Gregory E. Stone, University of Toledo

• Jonathan D. Bostic, Bowling Green State University

• Gabriel Matney, Bowling Green State University

• Connor J. Sondergeld, MetriKs

• Timothy D. Folger, Binghamton University - SUNY

• Yiyun Fan, Binghamton University - SUNY