From The Denver Post
By Kevin Simpson
The Denver Post/TNS
October 3, 2017
DENVER — An online chorus voicing anti-vaccine views linking some immunizations to autism has gained momentum via Twitter, a five-year University of Colorado study shows, with most of the negative comments coming from California and several northeastern states — particularly in areas of high affluence and concentrations of new moms.
The study used an algorithm to analyze more than a half-million tweets between 2009 and 2015 that mentioned both autism spectrum disorder and vaccines.
Although anti-vaccine comments nationally became more common over time, the study notes that this does not measure prevailing attitudes on the subject — though it does suggest that the debate rages on.
“It’s a pitfall to look at the data as representative of public opinion or a poll,” said study co-author Chris Vargo, an assistant professor at CU Boulder’s College of Media Communication and Information. “It’s more a measure of, are there vocal anti-vax people in a given area? These are people who are shouting loudly.”
For many years, some have claimed that certain vaccines can lead to autism, though studies haven’t found a causal link, said co-author Theodore Tomeny, an autism researcher at the University of Alabama. “Unfortunately,” he said, “the idea is still very much out there, being promoted by a vocal minority online. That’s problematic because often only one side of the story is being told.”
In Colorado, the Fort Collins area generated a high number of anti-vaccine tweets, according to the study. Nationally, the researchers found the highest incidence of negative tweets coming from Connecticut, Massachusetts, New York and Pennsylvania, in addition to California.
Overall, about half the tweets examined were negative. Vargo noted that about 24 percent of online Americans use Twitter regularly, according to a Pew Research Center report, and the study also took measures to control for “bots” and propaganda by focusing on unique users to avoid a prolific individual account from skewing the data. “Even then,” he said, “it’s a measure of vocal people who are angry and getting on their soapbox.”
The study originated two years ago, when Vargo was working at the University of Alabama. He and Tomeny were talking about how anti-vaccine arguments, despite being refuted over the years, have persisted. They decided to try to understand why this was happening using “big data.”
They paid for access to a compete archive of tweets and set about applying an algorithm to pull out those related to vaccines and autism. Then they grouped them geographically and cross-referenced U.S. Census data, which confirmed their hypothesis that anti-vaccine sentiment corresponds to affluence.
“The No. 1 thing that jumped out in the data is that as income increased in an area, the number of tweets about anti-vaxxing goes up,” Vargo said. “Folks in the middle class tended to see less anti-vax tweets. But as income level went up, the relationship got stronger.”
The authors also noticed that areas of high Twitter activity reacted strongly to news coverage of the subject.
Ultimately, the authors said, the study could help health care providers identify areas where more education on the subject could be helpful. “Identifying clusters of anti-vaccination beliefs can help public health professionals disseminate targeted/tailored interventions to geographic locations and demographic sectors of the population,” the study said.
The study is published in the October issue of Social Science and Medicine, and was co-authored by Sherine El Toukhy, a researcher with the National Institute on Minority Health and Health Disparities.