Does Geo-co-location Matter? A Case Study of Public Health Conversations during COVID-19

TL;DR

The study investigates whether geographic co-location between public health experts and the online public influences engagement on Twitter during COVID-19. By constructing a large dataset of PHE tweets and replies and inferring user locations with Carmen, the authors compare geo-co-located (gcl) vs non-gcl interactions, finding higher engagement for gcl conversations, especially on masking, lockdowns, and education, and among academic/medical professionals. LIWC and sentiment analyses reveal that gcl replies tend to be more positive and emotionally expressive when authors share personal experiences, while non-gcl replies skew toward negative emotions; topics like vaccines show different patterns. The findings imply that leveraging geo-localization can enhance public health messaging, with practical implications for localized outreach and discourse management in online health communications.

Abstract

Social media platforms like Twitter (now X) have been pivotal in information dissemination and public engagement, especially during COVID-19. A key goal for public health experts was to encourage prosocial behavior that could impact local outcomes such as masking and social distancing. Given the importance of local news and guidance during COVID-19, the objective of our research is to analyze the effect of localized engagement, on social media conversations. This study examines the impact of geographic co-location, as a proxy for localized engagement between public health experts (PHEs) and the public, on social media. We analyze a Twitter conversation dataset from January 2020 to November 2021, comprising over 19 K tweets from nearly five hundred PHEs, along with approximately 800 K replies from 350 K participants. Our findings reveal that geo-co-location is associated with higher engagement rates, especially in conversations on topics including masking, lockdowns, and education, and in conversations with academic and medical professionals. Lexical features associated with emotion and personal experiences were more common in geo-co-located contexts. This research provides insights into how geographic co-location influences social media engagement and can inform strategies to improve public health messaging.

Figures (4)

• Figure 1: Distribution of gcl and non-gcl engagement rate. Tweets with $0\%$ engagement rate are not shown in the plot.
• Figure 2: How gcl replies and non-gcl with specific LIWC categories correlate with LIWC features in the original Tweet. Each column is a logistic regression. Each cell is the coefficient. $p < 0.1$ is marked with *, $p < 0.05$ with **, and $p < 0.01$ with ***.
• Figure 3: Most distinctive n-grams ($n\in \{1,2,3\}$): tweets with $0\%$gcl engagement rate (bottom $~68\%$ of the tweets) vs. (a) the top $10\%$ percentile of tweets ($3.4\%$gcl engagement rate) and (b) the top $20\%$ percentile of tweets ($1.0\%$gcl engagement rate) . Only n-grams with $99.5\%$ confidence level are colored ($z > 2.57$)
• Figure 4: How gcl replies and non-gcl with specific LIWC categories correlate with LIWC features in the original Tweet. Each column is a logistic regression. Each cell is the coefficient. $p < 0.1$ is marked with *, $p < 0.05$ with **, and $p < 0.01$ with ***. Only selected LIWC categories are shown for demonstration purposes.