RESEARCH OVERVIEW

Increasing global mobility creates a significant design challenge in supporting international students, who must rapidly adapt to unfamiliar academic, social, and cultural contexts while managing intense emotional and academic demands. This adaptation is a continuous, high-stakes process of sense-making.

While Artificial Intelligence (AI) and Large Language Models (LLMs) have emerged as crucial, daily digital resources—used for tasks ranging from interpreting academic feedback and social cues to managing stress—they often function as generalized tools that are not intentionally designed to address the unique cultural and cognitive load faced by this specific population.

The core problem is that we lack an understanding of the experiential reality of how international students are actually incorporating these generic AI tools into their complex, high-pressure adaptation journeys. Without this insight, designers and developers risk creating unhelpful, or even detrimental, digital supports that fail to mediate the cross-cultural challenges effectively.

This research, therefore, aims to investigate the real-world roles and experiential limitations of LLMs in the daily adaptation of international students, establishing a foundation of user context necessary for the intentional design of specialized, culturally sensitive AI tools.

RESEARCH QUESTIONS

1. What everyday challenges do international students try to address using AI tools?
2. What motivates international students to turn to AI tools instead of existing support?
3. What perceptions and feelings do international students have toward using AI tools?
4. What aspects of these experiences and perceptions could inform the design of a AI-based solutions that aligns with students’ values and needs?

CHALLENGES

Conducting research on international students’ use of generative AI presents unique challenges. In this project, we identified key obstacles and adopted strategies to address them effectively:

1. Diverse Backgrounds and Contexts: International students come from a wide range of cultural, educational, and linguistic backgrounds, which can make designing standardized research instruments and interpreting results consistently difficult. To manage this, we focused on a single demographic—Indian international students. While this introduced some bias, it allowed for deeper, culturally informed insights, leveraging the researcher’s own experience as an Indian international student.
2. Self-Reporting and Behavioral Gaps: Students may underreport or overreport their use of GenAI due to social desirability, fear of academic consequences, or unclear understanding of the tools’ capabilities. To capture authentic behavior, we designed data collection methods that encouraged participants to share experiences organically, rather than recalling abstract or fragmented memories.
3. Ethical and Privacy Concerns: Researching GenAI use involves sensitive information, including academic work, personal communication, and mental health. We prioritized methods that minimized risk of exposure, allowing participants to engage without worrying about their identities being revealed or their actions being judged.

MY RESEARCH APPROACH

As Project Lead and Primary Researcher:

1. Designed an experimental study: Developed a 90-minute session based on the principles of participatory structured storytelling, inquiry, and persona creation.
2. Facilitated the study: Conducted sessions with 30 Indian international students across 5 focus groups, recruited using purposive sampling.
3. Human-in-the-loop data curation: Reviewed and curated a dataset of 500+ student narratives.
4. Conducted thematic analysis: Synthesized 239 narratives into six thematic domains, revealing the key challenges that students address using Generative AI.
5. Identified key features: Determined four features that make Generative AI tools particularly appealing to international students.
6. Evaluated user preferences: Analyzed what students like and dislike about the tools.
7. Explored future tool design: Investigated how students envision the design of next-generation AI tools to better meet their needs.

INITIAL FINDIGS & DESING IMPLICATIONS

The findings from this study have directly informed a set of 15+ design recommendations for reimagining LLM tools to better support international student success, currently under review.

IMPACT & NEXT STEPS

The findings from this study have directly informed a set of 15+ design recommendations for reimagining LLM tools to better support international student success, currently under review.