Overview #
Optimising Public Health Messaging: Analysis of Type, Sentiment, and Source examines how tweet type, sentiment and source shaped public engagement with COVID-19 messages in Ireland and the UK. The research shows how social media analysis can help public health teams design messages that are more timely, targeted and responsive to public concerns.
Abstract #
Objective: This study examined how the type, sentiment and source of COVID-19 tweets influenced public engagement in Ireland and the United Kingdom. It responds to a gap in understanding which combinations of message content, tone and sender are most effective for public health communication on social media.
Approach: The research analysed tweets posted between 1 January 2020 and 31 March 2022. The team combined manual content analysis, abductive coding and machine-learning classification to code each tweet by user category, message type and sentiment.
Analysis: The dataset was organised into 12 user categories, six message categories and three sentiment classes. A zero-inflated negative binomial model was then used to identify which combinations were associated with higher engagement.
Findings: Personal stories and positive messages generated the strongest engagement overall, although the most effective content varied by source. Known individuals and influencers gained the most engagement from humorous tweets, while health professionals performed best with advocacy, personal stories or statements, and humour-based content. Health institutes received higher engagement when using advocacy, personal stories or statements, and positive sentiment. Personal stories or statements accounted for 22% of tweets but 27% of engagement, while shock, disgust or fear-based messages made up 32% of tweets and 21% of engagement. Informative or educational messages were common, representing 33% of tweets, but attracted 16% of engagement.
The study concludes that public health communicators can improve engagement by tailoring message type and tone to the source, rather than relying on a single communication style.
Access #
Sana Parveen – University College Dublin (sana.parveen@ucd.ie)
Location #
UK, Ireland
Keywords #
Public health communication, COVID-19, infectious diseases, healthcare information, social media, infoveillance, WSMC conference 2025
