Troll Watch: Study Shows Older Americans Share The Most Fake News

Jan 13, 2019
Originally published on January 13, 2019 4:13 pm
Copyright 2019 NPR. To see more, visit https://www.npr.org.

SARAH MCCAMMON, HOST:

And now we'll return to our Troll Watch series.

(SOUNDBITE OF MUSIC)

MCCAMMON: That's where we bring you stories of cybersecurity attack, bots and, of course, Internet trolls. This week, we're going to focus on a new study out of Princeton and New York Universities that found older Americans are more likely to share fake news on social media than their younger counterparts. So if you've noticed it's your great aunt or uncle who most often shares dubious news stories on Facebook, well, you might be onto something. Our next guest is Andy Guess, assistant professor of politics and public affairs at Princeton University and the lead author of the study.

Welcome.

ANDY GUESS: Thank you. It's great to be here.

MCCAMMON: So why'd you decide to look into this?

GUESS: Well, I can't say that we were completely clairvoyant in knowing that fake news and online misinformation would be the thing that everyone would talk about after the 2016 presidential election. But it turned out that as more people were discussing this as something that they were concerned about, we realized we were sitting on this treasure trove of data that we'd been collecting the entire year. So it seemed a natural thing to study.

MCCAMMON: And how did you define fake news?

GUESS: We actually used a number of different sources - both peer-reviewed lists that people collected and crowdsource efforts to collect web domains and producers of false and dubious content. So we wanted to really triangulate into as narrow and rigorous a definition of fake news as possible without getting into the sort of gray area of what constitutes truth, what's the difference between news and opinion and that sort of thing.

MCCAMMON: And as you looked over the data that came back, I mean, what are your more significant findings?

GUESS: I think the first big takeaway is that, given the popular narrative around fake news and social media in 2016, the overall prevalence of fake news sharing on Facebook was pretty low. We're talking about something that was driven by a relatively small share of users who just had sort of disproportionate weight in our data.

MCCAMMON: Did your research give you any insight into why older social media users were more likely to share fake news?

GUESS: We can't answer the why question with the data that we collected. What we can do is establish as best as we can that this is a real relationship. And so that relationship is, on average, people who were over 65 were associated with sharing more articles from fake news-producing domains on Facebook in 2016. The fact that this doesn't seem to be driven by older people being more conservative or other factors that are also associated with being older was really striking to us. So we found this independent age effect - independent of partisanship, independent of ideological affiliation, independent of all of these types of factors. Now I think the task is to really adjudicate between the different possible explanations for what's driving this.

MCCAMMON: Yeah. And I don't want to ask you to speculate beyond what the data show, but at the same time, do you wonder if this might have anything to do with younger generations grew up with the Internet, older generations didn't? I mean, how much does that come into this?

GUESS: So there's been a lot of talk about digital media literacy. And there have been a number of efforts around the world to try and boost teaching of digital literacy skills, especially in schools. So we've seen efforts like that, for example, in France. Especially in the past couple of years, as concerns over online misinformation have grown, I think our findings suggest that really a lot of the action is going on not among students and teenagers in schools but among people who are at the other end of the age scale. And so digital literacy education is great, but that's not necessarily going to solve the problem that we identified in our study.

MCCAMMON: So you found that the older person was, the more likely they were to share fake news. And you said this is true regardless of partisan identity or ideological affiliation. But you also did look at partisan affiliation, right? What did you find there?

GUESS: We found a relationship between being more conservative or more Republican and also on average sharing more fake news articles. And that was a clear relationship as well and one that we weren't necessarily as surprised by because we know that most of the false content that was being disseminated during this period was strongly pro-Trump and anti-Clinton in orientation. And it makes sense that people would be more likely to share and engage with content that they're predisposed to agree with.

MCCAMMON: After the 2016 election, there was a lot of finger-pointing at Facebook, at Facebook users sharing fake news. From your findings, do you think older Facebook users helped elect Donald Trump?

GUESS: No, I don't think the evidence that we've shown here, you know, supports such an interpretation. I think that connecting, you know, fake news on social media with the election outcome is highly speculative. And, personally, I find that to be pretty implausible given the weight of the evidence that we've seen so far.

MCCAMMON: Your study did find that a relatively small percentage of people shared fake news. I mean, is there - are there things to feel good about here?

GUESS: So it's true that most people, including, you know, most people over 65 were not doing this. It's also the case that plenty of people were sharing corrective information. So we also find that lots of people were sharing links to fact checks of fake news. So, you know, while it's easy to sort of focus on the problem, and I think the problem is real, it's also the case that potentially more people than the ones who are sharing fake news are also sort of vigilant and paying attention to the possible problems with the information that they're encountering online.

MCCAMMON: That's Andy Guess of Princeton University. He was the lead researcher on this study.

Thanks for talking with us.

GUESS: Thank you very much. Transcript provided by NPR, Copyright NPR.