(PS8-71) Examining race and ethnic group differences in openness to donating personal data from digital devices for mental health research among college students

Abstract Text:

Introduction: The ubiquity of technology use among youth offers researchers the potential for rich insights into human behavior. Digital data donation is a collaborative approach whereby participants upload data collected by digital platforms for the benefit of research, which may ultimately inform treatment and clinical care. Given the novelty and privacy implications, it is important to understand how technology users feel about data donation, particularly groups historically underrepresented in research and who frequently experience health inequities. This study explores racial and ethnic identity group differences in college students’ comfort about: 1) sharing different forms of their digital data, and 2) with whom the data will be shared.

Methods: Participants (N = 335; Meanage = 19.2, SDage = 1.2 years; 67.2% women, 28.7% men, 2.8% non-cisgender; 62.4% White; 12.5% East Asian, 9.6% South Asian, 7.5% Other Racial Backgrounds, 5.7% Black/African; 81.8% non-Latinx/Hispanic, 8.1% Latinx/Hispanic) at the University of Virginia (UVA) completed an Openness to Data Donation Questionnaire, reporting their comfort level regarding sharing 9 forms of data on a 5-point scale (1 = Very Comfortable to 5 = Very Uncomfortable) in 4 different scenarios (that varied by who could access the data). Participants were asked to assume that ordinary anonymization produces were followed, with data stored securely. Exploratory ANOVAs and t-tests investigated racial and ethnic identity group differences in levels of comfort with data sharing.

Results: Preliminary results revealed mostly comparable comfort levels across racial and ethnic identities. Specifically, when asked about four scenarios where their data would be shared, participants across racial and ethnic identities were most comfortable sharing summary statistics with UVA researchers, and participants were least comfortable sharing data for a public database. Across the nine data types, participants were most comfortable sharing social media data (i.e., YouTube, Facebook, Instagram, TikTok) and Google search queries, and least comfortable sharing text messages, ranging from M = 3.7 (SD = 1.1) for South Asian students to M = 3.9 (SD = 0.9) for East Asian students. Importantly, when examining racial identity group differences in openness to sharing data, effect sizes were very small, suggesting similar comfort levels across the fives racial groups. Similarly, effect sizes were very small or small when comparing non-Hispanic to Hispanic ethnic group differences.

Conclusions: It is noteworthy that participant comfort levels were broadly similar across racial and ethnic groups. Future studies should recruit larger numbers of racial and ethnic minoritized participants across broader age groups and settings, and consider the impact of intersecting identities along with social determinants of health and digital acceptability and literacy.