Journal of the American Geriatrics Society · 2020 · 17 citations · 23 references
MedicineNeurologyCommunity-based Participatory ResearchDementiaGerontologyDementia CareOnline Health CommunitiesSocial MediaSocial Media CommunicationSocial Media Health CommunicationAlzheimer Disease DementiaAlzheimer’s Disease CareCommunity ForumsMixed Methods AnalysisMixed Methods Research DesignResearch Data CurationSocial Media CommunitiesSocial Media Dissemination
BACKGROUND/OBJECTIVES: Social media platforms are promising sources for large quantities of participant-driven research data and circumvent some common challenges when conducting dementia research. This study provides a summary of key considerations and recommendations about using these platforms as research tools for dementia. DESIGN: Mixed methods. SETTING: Alzheimer's Society's online Dementia Talking Point forum from inception to April 17, 2018, and Twitter in February and March 2018. PARTICIPANTS: All users of Dementia Talking Point who posted in subforums labeled "I have dementia" and "I care for a person with dementia," and Twitter users whose posts contained the keywords "dementia," "Alzheimer," or "Alzheimer's." MEASUREMENTS: We quantified the average daily number of dementia-related posts on each platform and number of words per post. Guided by a codebook, we conducted thematic content analysis of 5% of the 15,513 posts collected from Dementia Talking Point, and 10% of the 25,948 comprehensible posts from Twitter containing "dementia," "Alzheimer," or "Alzheimer's." We also summarized research-relevant characteristics inherent to platforms and posts. RESULTS: On average, Dementia Talking Point provided less than two new daily dementia-related posts with 213.5 to 241.5 words, compared with 7,883 new daily Twitter posts with 14.5 words. Persons with dementia (PWDs) commonly shared dementia-related concerns (75.7%), experiences (68.6%), and requests for, as well as offers of, information and support (44.3% and 38.6%, respectively). Caregivers commonly shared caregiving experience (67.0%) and requests for information and support (52.5%). Most common dementia-related Twitter posts were derogatory use of the term dementia (14.5%), advocacy, fundraising, and awareness (11.6%), and research dissemination (8.0%). Recommendations about these platforms' unique technical and ethical considerations are outlined. CONCLUSIONS: Understanding the priorities of PWDs and their caregivers remains important to understand how clinicians can best support them. This study will help clinicians and researcher to better leverage online health forums and Twitter for such dementia-related information.
23
A Survey of Techniques for Event Detection in Twitter
Atefeh Farzindar, Wael Khreich · Computational Intelligence · 2013 · 711 citations