Hiroshima University · 郝天時

Tianshi Hao

Specially Appointed Assistant Professor, PhD

AI in language education, studied through learning process data

Portrait of Tianshi Hao
BeijingLos AngelesHiroshimaBorn and raisedUSC, PepperdineHiroshima University, 2026–

Tianshi Hao is a Specially Appointed Assistant Professor at the Institute for Foreign Language Research and Education, Hiroshima University, Japan. He works across second language writing, the learning sciences, and learning analytics. His research asks what learning still happens on the student’s side when generative AI can draft, summarize, explain, and revise on demand, and how that learning can be traced in process data rather than in self-report alone. His primary methods are quantitative ethnography and epistemic network analysis. A further strand of his work is systematic reviews and AI-supported evidence synthesis.

At Hiroshima he teaches English courses in which students learn to write with AI, and he leads the development of generative AI guidelines for English education. He holds a PhD in Global Leadership and Change from Pepperdine University (2024), where his research focused on educational technology and virtual leadership. He also holds an MAT in TESOL from the University of Southern California (2020). He is the author of Play to Learn: Use Video Games to Learn a New Language (2020).

Research

My work follows three connected lines:

AI in language education

How generative AI changes the way students write, read, and learn in a second language, and how universities should respond, from classroom practice to institutional guidelines. His 2025 article “Spontaneous use of generative artificial intelligence in preparing for college English test: GenAI adopting factors, learner motivation, and self-efficacy” appeared in Technology in Language Teaching & Learning.

Quantitative ethnography

Using epistemic network analysis (ENA) to trace learning and identity patterns in process data, including how learners interact with AI tools and what separates productive use from shortcut use. The ICQE 2026 chapter (Springer) uses ENA to map patterns of generative AI use during source-based writing.

Evidence synthesis

Systematic reviews that map evidence across education and the social sciences, and AI-supported tools that make synthesis faster, including work on an NIH-funded project at Hyperthesis LLC.

Featured Publications

International Conference on Quantitative Ethnography (ICQE) Proceedings, Springer Communications in Computer and Information Science (CCIS) vol. 3227, Chapter 16 · 2026

Patterns of Generative AI Use Associated with Higher- and Lower-Quality Source-Based English Writing

Tianshi Hao, Paul McGachey, Marshall Higa

A quantitative ethnography pilot study of how Japanese first-year undergraduates used ChatGPT during a source-based English argumentative writing task, and which patterns of use were associated with higher- and lower-quality writing.

Read more

All publications

Teaching

At Hiroshima University I teach undergraduate English courses in which students learn how to write with AI. I also run Writing English with AI, On Your Own Terms, a 10-session course open to university staff, faculty and students, and I organize faculty development sessions on the digitalization of English education and AI literacy.

Contact

The best way to reach me is by email at haots2308@gmail.com. I am glad to connect with colleagues working on AI in education, second language acquisition, the learning sciences, learning analytics, quantitative ethnography and systematic review, including co-authoring, ENA collaborations and invited talks.