Author Type

Graduate Student

Date of Award

Summer 7-24-2026

Document Type

Thesis

Publication Status

Version of Record

Submission Date

August 2026

Department

Communication and Multimedia Studies

College Granting Degree

Dorothy F. Schmidt College of Arts and Letters

Department Granting Degree

Communication and Multimedia Studies

Degree Name

Master of Arts (MA)

Thesis/Dissertation Advisor [Chair]

Jaime Robb

Abstract

This study presents a longitudinal, two-condition, between-subjects experiment in which partner identity is disclosed as either human or AI in text-based chat; however, all participants (N = 60) interacted with the same chatbot. Informed by both Social Information Processing Theory (SIPT; Walther, 1995) and the Computers Are Social Actors Paradigm (CASA; Nass & Moon, 2000), this design draws on the former’s longitudinal measurement tradition to empirically test whether its predictions hold, using the latter as both a framework and procedure for investigating chatbot interactions. The findings inform both theories, indicating that perceived partner identity significantly affects perceptions of relational communication with chatbots, establishing identity salience as a boundary condition for CASA’s predictions. However, SIPT’s paramount claim that perceptual changes occur over repeated interactions and cue exchange was not supported. Together, these findings position identity salience as a condition that governs whether social responses to chatbots mirror those of human interaction.

Download

Included in

Communication Commons

Share

COinS