Using Twitter to Enhance the Social Support of Hispanic and Black Dementia Caregivers (Tweet-SS)

人工智能/机器学习伦理发展和使用行政补充

Administrative Supplement for Ethical Development and Use of AI/ML The prevalence of dementia is higher in Hispanics and Blacks than non-Hispanic Whites. Dementia caregivers often experience social isolation and loneliness as well decreased health status and lower quality of life. Social support interventions have been demonstrated to help mitigate these issues. The expansion of social media use among Hispanics and Blacks, particularly Twitter – a short message service – offers great promise for improving social support for Hispanic and Black dementia caregivers. However, analytic frameworks for analyzing Tweet content and social network structures in Twitter communities are relatively new and evolving rapidly, and there has been limited use of Twitter for intervention delivery. Informed by Social Exchange Theory, the goal of “Using Twitter to Enhance the Social Support of Hispanic and Black Dementia Caregivers” (Tweet-S2) is to advance the science of social support for Hispanics and Black dementia caregivers using Twitter. The intervention comprises daily posting of social support (informational, instrumental, emotional, appraisal) Tweets with embedded web-based information such as videos tailored to Black and Hispanic dementia caregivers, interactions with the caregiver networks by the bilingual project Tweeter, and a monthly Twitter chat hosted by the research team. The aims of the parent award Tweet-S2 are to: 1) Analyze Tweets specific to dementia caregiving in Blacks and compare Tweet content and social network structures with our prior findings in Hispanics, 2) Operationalize a Twitter-based intervention to enhance the social support for Hispanic and Black dementia caregivers, and 3) Evaluate the impact of the Twitter-based intervention on Tweet content and social network structures of Hispanic and Black dementia caregiver networks. This Supplement request in response to NOT-OD-22-065 expands the activities of Aims 1, 2 and 3 of the TweetS2 through the ethical use of our first version Black Tweet detection algorithm developed in Aim 1 to inform the design of culturally sensitive intervention in Aim 2 and recruitment process in Aim 3. The goal of the Supplement is to examine the risks, benefits, and ethical dilemmas inherent to the use of the “Black Twitter”/Black discourse detection algorithm that can occur at all layers of the socio-ecological model to inform the best practices and guidelines for ethical use of this AI/ML algorithm and any related approach developed and applied via similar methods. Using a mixed-methods approach, we will conduct in-depth interviews and an online survey with stakeholders, ethicists, and the public to achieve the goal. The contribution of the proposed study will be to develop an ecological approach to the ethical use of minority detection algorithms in social media in the context of the multidimensional dilemmas that arise when creating interventions to address disparities in dementia caregiving. These findings will inform the development of best practices and guidelines for users of demographic detection algorithm whose research focuses on designing and delivering culturally sensitive interventions regarding various diseases and health conditions.