Speech

Using readily available social media data to describe sentiments towards transmission-reducing behaviours during the Covid pandemic

Details

Citation

Maltinsky W, Manuf S, Den Daas C, Ozakinci G, Gaitens H & Swingler K (2022) Using readily available social media data to describe sentiments towards transmission-reducing behaviours during the Covid pandemic., 23.08.2022-27.08.2022.

Abstract
Background: At the outset of the pandemic, in absence of a Covid-19 medical intervention, transmission reducing behaviours (TRB) were the only way to prevent transmission. The Scottish Government enforced key behavioral measures including mandatory lockdown, 2-meter distancing, and wearing of face-coverings. The aim of this paper is to understand TRB Scottish public sentiments and behaviours in relation to government guidelines during the first year of the COVID-19 pandemic. Methods: A government public health timeline was constructed highlighting key dates/announcements in Scotland. This timeline was superimposed on TRB Google search trends and TRB social media twitter mining for sentiment analysis conducted between 01/03/2020 and 31/03/2021. Population TRB behavioural adherence data was collated from the CHARIS project. Results: The increased Google TRB search results, tweet sentiments, and adherence to TRBs demonstrated a pattern consistent with changing guidelines. Extensions in lockdown lengths were met with negative sentiments witnessed in April 2020, January 2021 and February 2021, in addition to mandatory face coverings in the workplace (October 2020). Positive sentiments were associated with easing of rules, such as phased return to campus learning in July 2020 and August 2020, and the reduction in the self-isolation period in December 2020. Self-report TRB data indicate high adherence to guidelines. Conclusions: While sentiments wavered in polarity, behaviour consistently was adherent even as guidelines changed. Sentiment analysis of social media used in conjunction with self-reported behavioral data plotted against key time frames, can enable a deeper understanding of public perceptions which may assist future public health guideline announcements