Attitude Networks and COVID-19
Within the context of the COVID-19 pandemic, mapping divisions in health attitudes and behaviors can be important for understanding how effective a region’s COVID-19 response will be. As we have seen, successfully flattening the curve requires coordinated actions and trust in the process from all groups that make up a region (a college campus, a city, a state, a country, etc.). The authors of the study, “Mapping public health responses with attitude networks: the emergence of opinion‐based groups in the UK’s early COVID‐19 response phase,” wanted to decipher if rifts in public health behaviors emerged in the U.K. over 3 crucial weeks in March when the pandemic was more evidently taking hold in Europe.
In class we have explored bipartite graphs which exist where a network has two different types of nodes and edges exist only between nodes of different types. Using this type of graph, the authors were able to showcase two different representations of attitude networks: one where nodes represent different individuals and a second where nodes represent different attitudes.
In the first type of graph, an edge between two nodes exists if the two individuals share many of the same attitudes. Going from T1 to T3, one can notice that two clusters emerge. The authors identify the less dense cluster as those who are skeptical of public health initiatives and the more dense cluster as those who trust in their value (Maher et al., 2020).
Figure 1: Individuals as nodes
In the second type of graph, an edge exists based on the number of people who do or don’t share the same attitudes. Blue edges exist where many people agree with the two connected nodes and red edges exist where many people do not agree with both of the connected nodes. While right from the start there are two distinct clusters, from T1 to T3 there become fewer blue edges between the two clusters while edges between nodes within each cluster become stronger (Maher et al., 2020). For example, over the 3 weeks in March a stronger alignment emerged between those who trust scientists and those believe vaccines are important for children while a negative alignment emerged between those who trust scientists and those who believe in the work of traditional healers (refer to Figure 2).
Figure 2: Attitudes as nodes
Attitude networks have become increasingly fascinating to study given the context of a global pandemic. This study done in the U.K could easily be replicated with Cornell’s Ithaca campus community. For example, it would have been extremely informative for the university administration to track COVID-19 opinions and behaviors during the periods of arrival testing, the beginning of surveillance testing when clusters had emerged, and a few weeks into surveillance testing as cases dropped significantly. Being able to graph how the attitudes of individuals in our community shift as circumstances change provides an avenue for tracking if individuals are becoming more lenient with social distancing practices as the immediate threat of infection subsides. Equipped with this knowledge, the university could know if stricter messaging is required as we settle into the semester.
Reference:
Maher, P.J., MacCarron, P. and Quayle, M. (2020), Mapping public health responses with attitude networks: the emergence of opinion‐based groups in the UK’s early COVID‐19 response phase. Br. J. Soc. Psychol., 59: 641-652. doi:10.1111/bjso.12396