PD26-11 Beyond the Bubble Sheet: Shaping the Future of Assessment with AI

Session Type: Professional Development Course

Abstract

The goal of this course is to prepare education professionals to responsibly adopt AI within assessment systems that are both technologically advanced and equity-driven. This training offers a practical, hands-on introduction to AI-enhanced assessment systems, focusing on four key areas: understanding and using large language models (LLMs), exploring ethical considerations for using AI-based assessments, using advanced machine learning for cheating detection, and assessing evolving definitions of validity. The course blends presentations, live demonstrations, guided coding activities, and small-group collaboration. By the end of the course, participants will be equipped with tools, strategies, and insights to responsibly apply AI in their own educational contexts, supporting more equitable, valid, and forward-looking assessment practices. Designed for a wide range of educational professionals, including graduate students, early-career scholars, researchers, education administrators, and policymakers, the course is accessible to those with basic knowledge of assessment and general familiarity with AI concepts. No programming experience is required. Support will be provided for varying levels of technical proficiency. All materials, including open-source code notebooks, datasets, slides, and prompts, will be distributed in advance.

Sub Unit

• Professional Development and Training Committee

Directors

• Xinhui Xiong, ExamRoom AI

• Mark D. Shermis, Performance Assessment Analytics, LLC

Instructors

• Christopher Ormerod, Cambium Assessment

• Xinhui Xiong, ExamRoom AI

• Matthew S. Johnson, Educational Testing Service

• Hong Jiao, University of Maryland

• Jiawei Xiong, Curriculum Associates

• Javier Suárez-Álvarez, University of Massachusetts - Amherst

• Sergio Araneda, Caveon Test Security