Students' Perceptions of Learning Analytics for Mental Health Support: Qualitative Study.

Background: Poor mental health among higher education students is a global public health concern. Learning analytics, which involves collecting and analyzing big data to support learning, could detect changes in behavior, learning patterns, as well as mental health and well-being. This could help inform mental health interventions in university settings. However, research has yet to explore students' perspectives on using learning analytics for mental health and well-being purposes. Objective: This study aimed to explore students' perspectives on using learning analytics to support students' mental health and well-being at university. Methods: Semistructured interviews were conducted online using Microsoft Teams between June and July 2023. Participants were identified through university student unions, social media, and snowball sampling. In total, 3 university students aged 20-26 years joined our team and formed our student advisory group (SAG). They informed the design, analysis, and dissemination stages of the research cycle. Braun and Clarke's approach guided our thematic analysis. Data were triangulated by comparing codes from 2 transcripts across 2 independent researchers over a 2-hour online meeting. A coding framework was cocreated with the SAG to code the remaining transcripts and ensure data saturation. Themes were finalized and presented in a thematic map during a 2-hour meeting with the SAG and 2 researchers. Results: In total, 15 participants were interviewed. We identified three main themes: (1) potential of learning analytics for mental health and well-being innovation, (2) student involvement in decision-making regarding learning analytics, and (3) integration of learning analytics with existing support. Despite being initially unaware, students recognized the potential of using learning analytics as a monitoring and early intervention tool to support university students' mental health. However, students raised concerns regarding data reliability and identified several ethical issues, such as privacy and lack of transparency. They also expressed the need to be involved in decision-making regarding learning analytics design, practices, and policies. Overall, students welcomed the possible integration of learning analytics with the existing university support. Conclusions: This is the first qualitative study to explore students' perceptions of using learning analytics to support student mental health and well-being. Students' generally positive attitudes toward learning analytics suggest that this tool could be effectively integrated into the existing university support systems. Considering the ethical concerns raised by students, our findings suggest the need to bring the student voice into learning analytics development and implementation.