Comparing Pre and Post-Lock Down Emotional Valence from Twitter to Gain Insights to Refine Interventions
Full Title: Comparing Pre and Post-Lock Down Emotional Valence from Twitter to Gain Insights to Refine Interventions for Hispanic and African American Family Caregivers of Persons with Dementia
Video
Team Information
Team Members
Frederick Sun, Doctorate Candidate, Vagelos College of Physicians and Surgeons, Columbia University
Faculty Advisor: Sunmoo Yoon, Associate Research Scientist in the Department of Medicine, Vagelos College of Physicians and Surgeons, Columbia University
Abstract
News media have frequently described dementia patients as COVID-19’s hidden victims, and family dementia caregivers have been physically and psychologically affected during the pandemic, citing loss of support and increased loneliness. On the other hand, positive caregiving experiences during the pandemic also were reported.1 Unfortunately, it remains largely unknown if these positive experiences extended to those who were already suffering socio-economic and structural disadvantages.1 Afinn algorithm helps to detect the affective state of users of social media, proposing a total emotional valence score.2 The purpose of this study was to detect topics and emotional valence to understand emotional distress as a foundation for developing Twitter-based interventions for dementia caregivers.
The purpose of this study was to detect topics and sentiment in the Tweet corpus to understand emotional distress as a foundation for developing Twitter-based interventions for Hispanic and African American dementia caregivers. We randomly extracted Tweets mentioning dementia/Alzheimer’s caregiving related terms (n= 58,094Tweets) from Aug 23, 2019 to Sep, 14, 2020 via an API. We applied natural language processing to identify topics and sentiments from the Tweet corpus and compared emotional valence scores of pre (through 2019) and post COVID-19 (2020-). The mean emotional valence score decreased significantly from 1.18 (SD 1.57; range -7.1 to 7.9) to 0.86 (SD 1.57; range -5.5 to 6.85) after COVID-19 (difference -0.32 CI: -0.35, -0.29). Interestingly, topics related to caregiver emotional distress (e.g., depression, helpless, stigma, lonely, elder abuse), and caregiver coping (e.g, resilience, love, reading books and poems) increased, and late stage of dementia caregiving (e.g., nursing home placement, hospice, palliative care) decreased in prevalence. Application of topic modeling and sentiment analysis of streaming social media Twitter provides the foundation for research insights regarding Alzheimer’s caregiving mental health needs for family caregivers of a person with dementia.
Team Lead Contact
Frederick Sun: qq2127@columbia.edu
