Read Latest Issue now

Francesca Tripodi, Data & Society: “How we get information is very flawed”

Is online and algorithm-driven news distorting our ability to see the world accurately? Dr Francesca Tripodi researched it in the American south

Sociologist and media scholar Dr Francesca Tripodi has been studying how the current trends in online and algorithm-driven news are magnifying inequality and distorting our ability to accurately assess our world.

Tripodi is a postdoctoral scholar at Data & Society where she studies how partisan groups interact with media and the role that community plays in deciding what is ‘real’ news. Her Data & Society report, “Searching for Alternative Facts” was published in May and is available for download.

Data & Society is a non-profit think tank that studies how Big Data impacts society on many different levels. Based in New York, the organisation tries to make sense of the double-edge sword that is transformative digital technology.

FEED: Can you start by telling us about your work?

Dr Francesca Tripodi: I study inequality in participatory media spaces. My dissertation looked at some discrimination that was happening on Wikipedia (“Non-notable? Deletion, Devaluation, and Discrimination on Wikipedia”, American Journal of Sociology) and also in a now defunct app called Yik Yak (“What Colleges Might Lose by Banning Yik Yak”, The Chronicle of Higher Education).

After finishing my doctorate in sociology I was fortunate to receive a postdoctoral fellowship with Data & Society for the past year.

FEED: And what has been your role at Data & Society?

FT: I have been part of the Media Manipulation Initiative team. We’re looking at a wide variety of questions – from something as urgent as concerted disinformation attacks to generally thinking about how we find information we can trust.

There are other researchers at Data & Society who are studying artificial intelligence or human rights or how technology affects the labour market. It’s a robust institution looking at a wide variety of issues concerning data.

In the past year, under the very generous funding and overarching advice of Data & Society founder Danah Boyd, we crafted a research project on the connection between partisanship and how we access information.

FEED: And this resulted in your report “Searching for Alternative Facts”. How did that come about?

FT: We initially thought about how I could put myself in a few different states in the US – say, a blue state, a red state and a ‘purple’ state. In the US, blue states are those that tend to go to the Democrats, red states go to the Republicans, and purple states are the swing states in between.

Unfortunately, ethnography takes a really long time, and only having one year, I had to home in on one specific area. The state of Virginia was having a very high profile governor’s race, and Virginia is now considered a swing state, even though it was red for quite some time. So we decided to focus my research inside the state of Virginia.

I was able to gain access to some really amazing people inside a women’s Republicans group and a college Republicans group who knew I was doing a research project and graciously allowed me to sit in on their monthly meetings and go with them to their events, in addition to allowing me to do observation at barbecues and fundraisers. I also went to the election night party and did in-depth interviews with people closely associated with them.

I wanted to get at a central research question, which was : how do you go about looking for news and information that you trust?

FEED: And what did you find?

FT: I was inside Republican organisations, but most of the people I spoke to identified as ‘conservative’ rather than Republican. The groups I was embedded with were in different cities, but there were similar rituals in all of the groups. And there was the influence of a Christian world view which, I think, does affect critical interrogation of media.

FEED: What kinds of sources did they use to get their news and information?

FT: They were actually looking at a wide variety of sources. Fox was obviously a very important one, and within Fox, Sean Hannity, Parker Carlson and Fox & Friends were the most commonly discussed. Online, there was the Daily Bell, which is a daily email news summary from the Heritage Foundation, and The Daily Signal, which is another Heritage Foundation outlet

There were also YouTube channels like Prager University or The Rubin Report, which tended to be watched by the younger people I spoke with, as well as podcasts from The Daily Wire. And in terms of major publications, there were The Financial Times and The Wall Street Journal.

A lot of the young people I spoke with got their information from their phone. They subscribed to a variety of news alerts, either through Apple News or downloading various news apps, and they subscribed to a fairly wide variety. A lot of the younger university students said they regularly read The New York Times and liked its long-form journalism. Markedly not on the list were MSNBC or The Washington Post.

FEED: Did they get any news from any non-US sources?

FT: Interestingly, Russia Today (RT) and Al Jazeera were also sources they talked about.

FEED: We talk about finding news sources we can trust. But what does ‘trust’ mean in this context? When people are looking for sources they can trust, are they looking for new information or just looking to have their world view confirmed?

FT: The people I spoke to were telling me they were looking for just the facts. They said that the mainstream media has become too emotional, that it is favouring emotion over intelligence and feelings over facts.

But what I argue is that facts are not neutral, and there are ways you can present ‘the facts’ that can sway arguments. It’s very complicated, but I think there are conservative media producers in the US who are promoting the story that the mainstream media is a bunch of emotion and not fact-based. Then these media producers are saying ‘We just present the facts’, but are very careful and very meticulous about the facts that they present.

It scares me how closed off we seem to be from any opinion that differs from our own

FEED: Your study was around American conservatism. But what is your understanding of how media tunnel vision happens in other parts of the political spectrum?

FT: Unfortunately, that was the limitation of just having a year-long project. I gained access to Republican groups straightaway and quickly realised that was all I was going to be able to actively observe and understand. But in future I want to continue to explore this idea of partisanship and trust in news and information more broadly.

I studied people who identify as US conservatives, but I don’t think this is exclusive to conservatives: I don’t think many of us understand the implications of how what we search on really matters and how the key words we put into Google might unintentionally reaffirm our existing beliefs – and in some cases, take those beliefs to a more extreme place.

FEED: Google has become the world’s major portal for accessing information. What did your research reveal about its use for getting accurate news?

FT: What I argue in the section of my report called ‘Googling for Truth’ is that Google as a platform can be really great for finding the local pizza place or a historical date, but what worries me is people are turning to it to find information about who they should vote for and what news they should be paying the most attention to. They think that Google is a neutral purveyor of information, but users don’t understand how what you put into Google dramatically alters what you get out.

For example, in the Virginia election for governor there was a campaign ad that accused the Democratic candidate Ralph Northam of squandering $1.4 million on a fake Chinese company. If you Googled ‘Northam fake Chinese company’, you got returns from The Washington Post, an editorial in the Richmond, Virginia newspaper, and But when you added ‘1.4 million’, the returns were dramatically different – you got a link to Americans For Prosperity who funded the ad, and to the Republican Governors Association.

What this says is if you know your audience – and you know, for example, that conservatives are more concerned about fiscal responsibility – through search engine optimisation and tying that ‘1.4 million’ to your site, you can shift those search results. Ideological beliefs can be tied to very simple changes in Google search terms.

I observed the same thing with the NFL protests. If you type ‘NFL ratings up’ or ‘NFL ratings down’, the results will reaffirm your ideological position.

What worries me is people have become so fed up with news that they don’t trust it, so they are blindly turning to Google as a source of trusted information, without examining how Google is a corporate media company that also has a bottom line.

In a time when we’re being taught to be very critical of media, we’re not providing that same critical lens to sources like Google.

FEED: Are search algorithms leading us to more extreme content?

FT: I think there are a network of personalities in conservative media that dabble in content that is much more extreme. By having certain guests on your shows that, then, are guests on other – much more radical – shows, you are algorithmically connecting to those more radical points of view. On YouTube’s ‘Up Next’ you are more likely to appear beside them. That can help create these rabbit holes where viewers are being exposed to more radicalising thinking.

In Safiya Noble’s book, Algorithms of Oppression, there’s a great conversation about (convicted mass murderer) Dylan Roof Googling ‘black on white crime’. Google has now fixed it – if you Google ‘black on white crime’, you get links back explaining how it is not actually a real thing. But prior to Dylan Roof’s murdering nine African-Americans in a place of worship, when you Googled ‘black on white crime’, the top link was the Council of Conservative Citizens, which was actually a cloaked white supremacist website. So what you search for and when you search for it can have unintended consequences. Or intended consequences – I’m not sure what Dylan Roof was really looking for when he searched that in the first place.

I observed in my research people turning to Google as a way of fact-checking, or as a way even to challenge their existing beliefs, but what they typed in might end up reaffirming what they already believe, even if they went in trying to push their boundaries.

FEED: Is there something inherent in the technology we now use to get our news and information that has a radicalising effect?

FT: There is a way of thinking about news broadcasts of the past as if it was a golden era of news and information. But historically if you look at under-represented groups, they’ve never been represented in the news. When you had missing black women, that never made the nightly news. Many of the great media historians have described how the gatekeepers of news kept out a lot of news that mattered to under-represented people. Even in the era of print news, when we had these vibrant newspapers, there was still a pretty intense gatekeeping process in terms of what stories got covered.

I think there has always been ‘counter-news’, trying to galvanise or reach people who don’t feel represented by the mainstream narrative. And I think equally there are ways to try and manipulate those efforts into concerted disinformation. I’m sceptical that these are just brand-new problems. There have always been sections of the newspaper that people have pulled out and focused on and ignored the rest.

FEED: Are there healthier ways we could be consuming and delivering news? Ways we could be getting a broader and deeper perspective?

FT: I think the model the US media uses to operate is deeply problematic. When it’s based on how many eyeballs you can get, you get things like the election of Donald Trump, because that’s all they were covering all the time. People were reading it and they were getting the clicks. And that’s what their bottom line is based on.

People could counter that and say, ‘Well, during the Brexit vote, you had the BBC, so explain that’. But how we get information is very flawed.

One thing that does concern me is how polarised our news consumption is right now. That does seem unique to me.

I was really surprised, in doing this research, how many stories I read on conservative news sites that I never saw any mention of in mainstream media – and vice versa. Fox is mainstream media, but it is kind of a place on its own. When you watch some of these different outlets, it’s an entirely different conversation happening with a completely different sets of facts. It’s a completely different story. And, though their slant may be something I don’t agree with, I don’t think conservative new sites are selling untruthful stories.

What I wish would happen – what I think would be really incredible – is if we heard a little bit more from people we don’t agree with and had a way to get information to people we don’t agree with. It scares me how closed off we seem to be from any opinion that differs from our own.


This article first appeared in the June 2018 issue of FEED magazine