Researchers analyze Facebook comments on vaccine beliefs

In January 2016, Kate Faasse, a psychology lecturer at the University of New South Wales in Sydney, Australia was tagged by a friend in a comments thread on a public Facebook post by that company’s co-founder Mark Zuckerberg. The post showed a photo of Zuckerberg holding his baby daughter with the caption, “Doctor’s visit – time for vaccines!” As Faasse read the comments, a thought occurred to her – why not use linguistic software to analyze the commentary and test hypotheses about the views of pro and anti-vaccine supporters. It would be a different approach from the traditional population sample testing processes employed by researchers.

Faasse contacted another friend, fellow researcher and psychology professor Leslie Martin stationed more than 7,600 miles away across the Pacific Ocean at La Sierra University. The two, along with La Sierra student researcher Casey Chatman, embarked on a six-month project to study the words individuals were using on Facebook in the back-and-forth debate about vaccines and whether they are safe. The scientists discovered, contrary to their initial hypothesis, that the anti-vaccine commenters used significantly greater analytical thinking in promoting their viewpoints, while the pro-vaccination crowd employed more emotionally-charged, anxious language.

The team’s findings were published in November in the pre-eminent international peer-reviewed journal Vaccines. ScienceDaily and Epoch Times also reported on the findings. The aim of the study, the scientists stated in their article, was to analyze types of vaccine debate language and arguments used within the same conversational context toward better understanding thought processes and shoring up future efforts aimed at spurring the use of vaccines.

Following approval from the University of New South Wales ethics board, the researchers began using Linguistic Inquiry and Word Count software developed by their colleague James Pennebaker to analyze types of phrases and rates of specific words used in the Zuckerberg vaccine post comments. 

The researchers studied 1,489 replies which posted within one week to a comment on the Zuckerberg post that garnered 49,000 likes. The researchers calculated the percentage of specific words categorized under Anxiety, Anger, Family, Risk, Work, Death and other topics and compared the percentage between pro-vaccination, anti-vaccination and unrelated, or control comments.

“The words that people use can provide important insights into their thoughts and emotions,” the researchers stated.

The scientists concluded that the more anxious nature of the pro-vaccination posts may indicate “a greater awareness and understanding of the scientific data and thus a greater cause to worry, especially about their own families whom they believe may be harmed by the failure of others to accept vaccinations.” Similarly, the anti-vaccine commentators may express less anxiety and appear more logically analytical “because they do not realize the erroneous nature of their beliefs, their flawed interpretation of the scientific evidence, nor the health risks to themselves and society inherent in their refusal of vaccinations.” The scientists attributed the primary source of anti-vaccine arguments to inaccurate or incomplete information on the internet.

Seemingly sound arguments posited by the anti-vaccination commentators on the Facebook post may “make anti-vaccination arguments particularly compelling for uncertain parents seeking information about childhood vaccinations,” the scientists said.

The research occurs as vaccine refusal rates are on the rise and have been linked to preventable infectious disease outbreaks. An article in Live Science earlier this year cites a study published in the Journal of the American Medical Association showing that of the 1,400 measles cases occurring in the United States between 2000 and 2015, more than half occurred in people who were not vaccinated against the disease, and a third occurred with people who had an “unknown history” of vaccination. “In developed countries, a large proportion of these infections occur among the intentionally unvaccinated,” Martin and Faasse said.

One of lessons confirmed through the unconventional social media analysis is that scientists may not have a firm grasp on the publics’ level of scientific knowledge. 

“The general public needs more help in understanding the scientific process, probability and risk,” said Martin. 

Potential applications stemming from the Facebook research could include apps for kids and curriculum focused on critical thinking skills, the researchers said. Also, the use of social media and website data in analysis could be applied to a broad range of health issues, said Faasse. 

By May 2016, the Zuckerberg post had received approximately 3.4 million likes and 184,000 comments, affording a significant opportunity for analysis. Social media provides a forum in which participants are less guarded in their responses than in more traditional scientific surveys and processes, the scientists stated. “It seems to be the forum where people are free to say [what they think],” said Martin.

Faasse and Martin met at an Innovations in Health Psychology conference four years ago in the Cook Islands and discovered common interests, including running. Most recently, Martin completed the Superfrog Half Ironman in San Diego. The two scientists do daily morning runs when attending conferences together.

In carrying out the analysis of the Facebook comments, the pair communicated mainly by e-mail and Skype. The duo is working on several other projects including how people conceptualize public health threats and how perceptions of risk change as media messages shift. They are recruiting participants through social media, but are not pursuing additional linguistic analyses, said Martin.