This paper has been accepted and is currently in production.
It will appear shortly on 10.2196/106984
The final accepted version (not copyedited yet) is in this tab.
Generative AI use for mental health purposes among college students: A cross-sectional survey
ABSTRACT
Background:
Mental health problems are prevalent among college students. In conjunction, growing use of AI among this population, including for mental health and emotional well-being support, requires further understanding of how these tools are being used.
Objective:
This study aimed to evaluate AI use for mental health purposes among college students. This included quantitative evaluation of predictors of AI use for mental health purposes and qualitative analysis of what participants may be seeking from AI related to their current mental health.
Methods:
Undergraduate students (N = 205) from a large public university in the United States completed a one-time survey assessing their use of AI for mental health purposes, current mental health symptoms, perceived social support, loneliness, well-being, and attitudes toward AI. Hierarchical logistic regression was used to assess predictors of AI use. Four groups of predictors, consisting of demographic characteristics (step 1; dichotomous sex, age, race), mental health symptoms (step 2; depression, anxiety, psychosis risk, lifetime therapy, perceived stress), social and well-being (step 3; flourishing, loneliness, perceived social support), and attitudes toward AI (step 4) were evaluated. Participant-written AI prompts were analyzed using an inductive coding approach in which codes were developed through affinity mapping and consensus discussion. Each prompt was then independently coded by three coders and could receive up to two codes.
Results:
Overall, 73/205 (35.6%) of respondents endorsed past use of AI for mental health and emotional well-being purposes. After accounting for all focal predictors, more positive attitudes toward AI (P < .001) remained significantly associated with increased likelihood of reporting AI use for mental health purposes. In contrast, predictors related to demographic characteristics, current mental health symptoms, and well-being were not significantly associated with use. Individuals who endorsed AI use for mental and emotional well-being reported being more likely to seek out support from online forums/support groups, self-help books/podcasts, mental health applications, and expectedly, generative AI (Ps < .05) and less likely to seek out support from family members (P = .050). Six themes were identified in participant-developed AI prompts, with use of AI as a guide, a listener, and an information navigation/search engine as the most commonly identified functions.
Conclusions:
Positive attitudes toward AI were the strongest predictor of AI use for mental health purposes. Students described various goals when using AI for mental health, most often seeking guidance, a listening space, or information and coping skills. Researchers, clinicians, and those designing and regulating these tools should distinguish among the different functions that AI serves for users rather than treating AI use as a single behavior. Given the varied uses of AI for mental health purposes in this population, greater research focused on the outcomes associated with these uses is needed.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.