Fairness, humanism, and ethics in oral communication assessment using generative AI | LARC 2025

Plenary Speakers:
Carol A. Chapelle (Iowa State University)
Presentation: https://drive.google.com/file/d/18tU6...
Gary J. Ockey (Iowa State University)
Presentation: https://drive.google.com/file/d/1kGyQ...

Today’s widespread access to generative AI (GenAI) amplifies the need for dedicated study of technology in language assessment. Of particular note are the opportunities GenAI may create for excellence in the assessment of oral communication. The dialogic assessment tasks created using avatars and GenAI represent an important step in a series of developments that have led to the use of diverse tasks, delivery modes, and scoring procedures for assessing oral communication. We will discuss some of our current collaborative projects that use emerging technologies to contribute to these developments by assessing oral communication with a goal of creating fair, humanistic, and ethical oral communication assessments. We will then provide examples of how the values based in concerns for fairness, humanism, and ethics ripple through all aspects of GenAI-mediated language assessment including creating test tasks, rating test takers’ essays in placement testing, and giving feedback on students’ written assignments. As practices and proposals in language assessment for generating and evaluating language are underway, we will suggest that validation research on GenAI-mediated language assessment should be shaped in part by the interests of creating fair, humanistic, and ethical assessments.