"Mommy Blogs" and the Vaccination Exemption Narrative: Results From A Machine-Learning Approach for Story Aggregation on Parenting Social Media Sites

"Although substantial work has considered the role of social media, particularly Twitter, in discussions of vaccination and other health care–related issues, there has been little work on describing the underlying structure of these discussions and the role of persuasive storytelling..."

This study sought to develop an automated and scalable machine-learning method for story aggregation on social media sites dedicated to discussions of parenting: so-called "mommy blogs". Although straightforward data mining techniques such as topic modeling exist for determining what parents are talking about on these sites, the researchers here sought to discover the aggregate narrative frameworks through which parents exchange experiences and commentary - that is, how they are talking about those topics. The study also sought to characterise temporal trends in these narrative frameworks on the sites over the study period. That is, the work shifts attention away from bursts of activity on sites such as Twitter to a complementary examination of long-term conversations that evolve over many months and even years, with a primary focus on the emerging and endemic narrative frameworks that inform these conversations and that can shape beliefs and decisions.

This research analysed 1.99 million posts contributed by 40,056 users and viewed 20.12 million times indexed from 2 popular parenting sites over a period of 105 months ending in 2012. As mothers are on the "frontline" of discussions about the health of their infant children, these sites offer information about how they approach decisions related to vaccination. Both draw members from a wide range of backgrounds with broad geographic diversity, although largely from the United States and Canada.

For the 2 parenting sites, the researchers determined the topics of discussion and the stories circulating in those discussions through an automated content analysis process, which is explained in detail in the paper. In brief, they started by computing dominant topics in the forums using 2 different probabilistic approaches, Latent Dirichlet Allocation (LDA) and Contextual Random Walk Traps (CRWT). To understand how people talked about the discovered topics, they then developed a story model, the actant-relationship context model, and used it to extract the underlying stories from posts across the entire set of 1.99 million discussion posts, recognising that forum posts frequently include only parts of stories or comments on story parts as opposed to complete stories. They conceptualise story parts as relationships among "actants". In their model, to generate a social media post, a user picks a set of actants and draws from the distribution of relationships among those actants. The user then composes the post according to the outcomes in the first step. In a social media corpus, the underlying probabilistic model including both the primary actants and their contextual relationships is hidden. Consequently, the researchers' task was to estimate this hidden model from the posts. They accomplished this through a computationally scalable estimation algorithm, described in the paper.

The methods developed for this study allow for the discovery of stories circulating informally on social media sites. The system can detect the presence, persistence, and pervasiveness of story signals on otherwise very noisy sites, aggregate these story signals into a narrative framework, and provide a mechanism for tracing the emergence of specific strategies endorsed in these stories that parents might adopt to counteract perceived health-related threats. In short, the analysis revealed that in most vaccination stories from the sites analysed, it is taken for granted that vaccines and not vaccine-preventable diseases (VPDs) pose a threat to children. Because vaccines are seen as a threat, parents focus on sharing successful strategies for avoiding them, with exemption being the foremost among these strategies. That is, the researchers found a strong narrative framework related to exemption seeking and a culture of distrust of government and medical institutions. Various posts reinforced part of the narrative framework graph in which parents, medical professionals, and religious institutions emerged as key nodes, and exemption seeking emerged as an important "edge" (relationship between nodes). In the aggregate story, parents used religion or belief to acquire exemptions to protect their children from vaccines that are required by schools or government institutions, but (allegedly) cause adverse reactions such as autism, pain, compromised immunity, and even death. Although parents joined and left the discussion forums over time, discussions and stories about exemptions were persistent and robust to these membership changes.

Any new parent joining these sites, irrespective of their orientation to vaccination, is exposed to stories that activate the narrative framework of vaccination as threat and exemption as strategy. Given the 90-9-1 rule of social media, where 90% of visitors simply read without commenting (9%) or contributing (1%), it is very likely that the narrative framework is reaching a much larger audience than simple user statistics suggest. The 1.99 million posts studied had an aggregate view count of more than 20 million views from registered users (unregistered users could view the posts, but their views were not recorded and therefore not tabulated). Even for parents who may not have initially believed that vaccines are harmful, the persistent circulation of stories about the potential harmfulness of vaccinations and the efficacy of the strategy of exemption to protect children from this alleged threat could convert some parents to embracing these beliefs.

The researchers discuss the persuasive nature of personal experience narrative, noting that storytelling plays a central role in exposing people to ideas and converting people to particular beliefs. People are inclined to believe first-hand accounts from members of their community, which is captured by social network theory's establishment of a strong tendency toward homophily in online communities. They believe that the personal stories highly popular on these sites make use of the shared trust developed in online forums and thus act as an ideal method for converting nodes to the beliefs encoded in those narratives. In this study, the persistence of the exemption signal suggests a broad-scale susceptibility in these networks to the exemption strategy for dealing with the "vaccination threat". Ultimately, their goal is to contribute to a system that monitors health care–related websites for emerging beliefs and attitudes, and that recognises the power of narrative to persuade and create communities of like-minded individuals.