Analyzing the Katie Couric effect on the vaccine conversation

On December 4th, 2013, Katie Couric gave the HPV vaccine center stage during a segment on her talk show, Katie. The segment, entitled “The HPV Controversy,” was 20 minutes long, but ignited a digital firestorm between pro- and anti-vaccine voices that raged for days after the stage lights went dark.

In partnership with Global Prairie, the entire online conversation surrounding this Katie segment was digitally captured using DataFarm. This powerful social media analytics tool allowed the vibrant, real-time conversation to become a tangible data set to explore. Our analysis revealed unexpected insights for those who use the Internet to promote public health messaging, with the potential to help improve and refine online efforts.

Here is an exclusive look at the social media conversation surrounding the Katie episode, as it was captured from November 30th to December 21st.

Analyzing the Katie Couric effect on the vaccine conversation

Figure 1: Total conversation around HPV and Katie Couric

Analyzing the Katie Couric effect on the vaccine conversation

Figure 2: Media sources and post title word cloud

*For information clarity, all word cloud images excluded the terms: hpv, #hpv, rt, katie, couric, @katiecouric, vaccines, vaccine, Gardasil.

A total of 12,049 posts were captured during this 22-day window. Figure 1 shows 2 distinct areas of conversation occurring in relation to 2 specific events. The first spike occurred around the actual airing of the Katie segment. The second spike followed the release of Couric’s statement that included her personal reflections on the content of the segment.

During this time period, the analysis revealed there were 40 online publications that were the primary catalysts driving social media conversation. For the purposes of this project, having 50 or more social media posts linking to or sharing the article defined this influential content. The text of the 40 articles was reviewed and subsequently defined as being either pro-vaccine (24 articles) or anti-vaccine (12 articles). These 36 pro- and anti-vaccine articles were responsible for 7,317 posts on Facebook and Twitter. The remaining 4 articles were interpreted as vaccine-neutral, or were written by the Katie editorial/marketing team or Couric, and were excluded from analysis.

Using DataFarm, 2 groups (those who shared the pro- or anti-vaccine articles) could be individually examined. The results of this analysis are shown below.

Analyzing the Katie Couric effect on the vaccine conversation

Figure 3: Pro-vaccine and anti-vaccine conversation generated by influential publications

Analyzing the Katie Couric effect on the vaccine conversation

Analyzing the Katie Couric effect on the vaccine conversation

Figure 4: Pro-vaccine media sources and post title word cloud

Analyzing the Katie Couric effect on the vaccine conversation

Analyzing the Katie Couric effect on the vaccine conversation

Figure 5: Anti-vaccine media sources and post title word cloud

As figure 3 shows, the pro-vaccine message was the first to be amplified around the time of Katie airing. Interestingly, the social media channel was Twitter. These conversations were primarily heavy-handed criticism of the inaccuracies within the episode’s content, including challenges to the journalistic integrity of Couric herself. Shortly after Couric’s statement, however, the anti-vaccine community’s conversation rose quickly, and nearly exclusively, on Facebook. Users who shared the anti-vaccine publications came to the defense of the episode, applauding its efforts.

Critical analysis of the online conversation surrounding this specific media event led to the following insights:

1. Conversational real estate. It was no surprise that the online conversation about this media event was primarily held on Twitter and Facebook. However, the analysis revealed an unexpected and exaggerated divide between the content shared on each of those social networks. The pro-vaccine message was primarily shared on Twitter, while the anti-vaccine message was more visible on Facebook. The cause of this virtual segregation is unknown and will be the subject of forthcoming analysis. Meanwhile, being aware of the relative isolation of each group’s online location should be an important consideration when discussing critical public health issues on digital forums.

2. Anger leads to action. This distinct timing of the 2 opposing viewpoints shows each community’s rise to speak in defense of their position. The language is aggressive and full of finger-pointing. Emotion reigns.

3. Intensity of purpose. The impact of the pro-vaccine voice was swift and severe, likely contributing to Couric’s decision to write a statement 6 days after the episode aired. However, of the 7,317 Twitter and Facebook posts captured, the anti-vaccine message was actually shared 20% more, despite having half of the written content and representing an opinion held by a significant minority of Americans. This reflects the vigor and digital impact of a passionate anti-vaccine community. It also shines light onto the pathetic shortfall of pro-vaccine message amplification.

In the upcoming part 2 of this post, additional insights surrounding this digital conversation will be shared.

Natasha Burgert is a pediatrician who blogs at KC Kids Doc.

email

Comments are moderated before they are published. Please read the comment policy.

  • edpullenmd

    No surprise here that the anti-vaccine element is more passionate, uses facebook more, it’s a more personal and emotional media outlet than twitter. Convincing anti-vaccine advocates by using facts, education and logic is simply not going to have any impact. Some issues like abortion, immunizations, and religion are just not based on logic, rather on deep personal belief and values that are not going go be modified by logical arguments.

    • Natasha Burgert

      Completely agree, Dr. Pullen. Does this work give you any insights as to how pro-vaccine advocates can be more effective in their messaging? Would love to hear your thoughts. NB

  • Dorit Reiss

    Thank you. I’m looking forward to your further analysis for the reasons behind these differences.

  • http://www.facebook.com/nurseswhovaccinate MelodyRN

    While the anti-vax misinformation was more likely to be shared by the public, I do find it comforting that the pro-vaccine message had the most impact. I’m curious to find out what it is about the anti-vax message that appeals to the public. What I do remember about this whole situation was that the anti-vaxers weren’t really active until they were angry about Ms. Couric apologizing. That’s when I took notice of their campaigns.

  • http://onhealthtech.blogspot.com Margalit Gur-Arie

    I think there is something missing here. The number of posted messages is not as relevant as the number of people that actually read those messages. If some small group of passionate folks kept tweeting one way or another, largely to themselves, the effects are most likely negligible and no different than what occurred before social media was around. The only difference is that now we can capture gossip and analyze it (and amplify it?). Case in point, I don’t do Facebook, but I am often on Twitter and I have no recollection of reading anything about this.

    As to the subject itself, I am not sure why we are surprised about the resurgence of illiterate response to science (which goes far beyond vaccines), seeing how poverty is increasing and our education system is failing.

    • Karen Leuenberger

      I agree that most of the social media activity boils down to preaching to the choir. But I’m not sure science literacy is going to actually get very far. Most people with anti-vaccination views tend to be college educated and with at least upper middle class incomes: http://www.immunizationinfo.org/science/demographics-unvaccinated-children It’s the undervaccinated (partially vaccinated) who are more likely to be at the poverty level – seems to point to an access problem, not an education problem, in this group. Unfortunately, strongly held ideology (on any side of the spectrum) tends to made unbiased analysis of facts difficult: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2319992 . This is an emotional issue. I think it will be countered best by story – narrative. Stories in the news, in movies, on TV, of what it is that we’re *not* afraid of anymore because we don’t see it – those diseases we vaccinate against.

  • http://onhealthtech.blogspot.com Margalit Gur-Arie

    Thank you.

  • Natasha Burgert

    Thanks, Tom. This work is very exciting, and relevant to anyone who wants to improve online health messaging. Will look forward to continued work with this data set to find some answers to the great questions readers have posed. NB

  • Dorit Reiss

    Was there a trigger for the timing?
    Will you do any content analysis – e.g. use of sources, types of arguments?
    I’m curious about your finding about the Twitter/Facebook difference, but my impression is you’re already looking at that.
    Are you also examining regular news outlets?

    • https://twitter.com/TomPeddicord Tom Peddicord PharmD

      I will let Dr Burgert fully answer the timing question as we at Global Prairie became aware of this through our conversations on the topic with her. Following these discussions, we established a monitoring platform on the topic to enable us to analyze the conversation. It should be noted that this project has been purely exploratory.

      In response to your second question, DataFarm monitors all major news outlets, the social conversation, open source information, and can incorporate enterprise (proprietary) information as well. The ability for DataFarm to cut through the “social clutter” allowed us to pinpoint the conversation around the Katie Couric program and analyze this in a unique way that had not been shared prior to this blog post to our collective knowledge.

      As you would expect, news releases from major organizations (Forbes, CNN, etc…) are typically the biggest drivers of the conversation, but this is not always the case. For example, within this data set, there is a Facebook posting from July 2013 that continues to drive a a large proportion of the conversation in the anti-vaccine community. We are currently analyzing this and hope to share more in future blog postings with Dr Burgert.

      Thank you for the questions!

  • Dorit Reiss

    And thanks for asking.

Most Popular