User-Centered Design + Generative AI Research Tools: Usability Testing and Implication

Comparing how generative AI facilitates interaction design undergraduate students’ work.

Yi-Fan Chen
Assistant Professor
Farmingdale State College

Research is involved in every step of the user-centered design process. It aims to define users’ needs and pains, uncover user behavior, survey the current state of similar design solutions, aid user experience (UX) design decisions, improve usability, and enhance overall user satisfaction. Designing and conducting various research during the human-centered design process takes time and effort. Recently, increasingly popular generative AI models can generate high-quality images, text, audio, synthetic data, and other types of content. UX professionals have found that generative AI can increase productivity. A Nielsen Norman Group report found UX professionals mainly used generative AI for content editing, research assisting, design assisting, and ideation co-designing (Liu, Zhang, & Budiu, 2023). On the other hand, how much and how effectively UX professionals utilize generative AI in their work is unclear. The current research aims to examine the transformative potential of AI research tools in facilitating user-centered design practices.

To examine AI tools for UX research, this ongoing ethnographic observation study began in the Fall of 2022 when ChatGPT was launched. The research site is an interaction design studio at New York State College. The instructor investigated several generative AI tools to learn each tool’s policies before introducing them to students. Students are encouraged to use the tools for their projects if they identify which tools they used in their projects. The projects served the purpose of comparing how generative AI facilitates interaction design undergraduate students’ work. Preliminary findings include students using ChatGPT to remind them of steps in research method design, such as writing recruitment notes, drafting open-ended and closed-ended questions, and analyzing data. They also use it to assist in user persona development, wireframing, and creating content for case studies. It takes time and experience to prompt useful suggestions. Limitations, implications, and future studies will be further discussed.

This design research is presented at Design Incubation Colloquium 11.1: Boston University on Friday, October 25, 2024.