Data-Shaming
Why sharing studies via social media can drive us further apart
Republican members of the US Congress share 9x more links to untrustworthy websites than Democrat members says a new study conducted by researchers at University of Bristol and Gratz University of Technology. The quality of information shared by Republicans has rapidly declined since 2016, while Democrat sharing of links to untrustworthy sites has remained stable.
As someone who laments the way low-quality information hinders the potential of social media to help us find new ways forward together, these findings seem extremely significant for a better understanding of how untrustworthy information gets disseminated. As someone who strives to use social media to unite rather than divide, I recognize that posting about these findings will alienate some within my digital networks.
Up until a few years ago, I was a registered Republican. If I had a seen a post about this study from a non-Republican in my network then, I likely would have dismissed it as biased, flawed, or flat-out wrong. It wouldn’t have resonated with my understanding of the world at the time — that Republicans care far more about the truth than Democrats, that they are maligned and mistreated by the biased mainstream media, and that the news sources deemed “trustworthy” by academics are only perceived as such because they reject and ridicule conservative thought.
Now, however, as an unaffiliated voter who has observed what I consider an alarming rise of mis- and dis-information being disseminated by Republican-leaning connections in my network, these study results resonate with my experiences. I want to share them because I believe them to be generally accurate. And I hope that if people review this research into the comparative sharing of untrustworthy information between the elected officials of both parties, we can have more meaningful conversations about where mis- and disinformation is really coming from.
But knowing how two versions of me respond to this data, means I know some might see it as a criticism of or attack on Republicans rather than as a neutral piece of information.
Data doesn’t exist in a vacuum. It’s always interpreted contextually. A big factor in how a person responds to data is whether they feel maligned, vindicated, or … as is often the case … unaffected by the picture it paints.
When we don’t feel like data affects us in any meaningful way, we ignore it. No matter how reliable or crucial the information may be, it has no impact on what or how we think. That’s why no amount of terrifying climate change data is convincing to those who think human activity plays little to no role in the daily weather disasters we are now experiencing. They can comfortably ignore any research regarding climate change because they believe it doesn’t impact them.
For example, a 2011 study cited in The Sum of Us: What Racism Costs Everyone and How We Can Prosper Together by Heather McGhee found that conservative white men are much more likely to be climate change deniers because of the story they’re receiving from other white males in the political media with whom they identify. They don’t see themselves as suffering the effects of climate change, McGhee says, so “their bias will be toward retaining a status quo that rewards them, even if it leads to suffering for others.” Even when, say, heatwaves are becoming so much more common, intense, longer lasting, and dangerous to human health that various entities are launching systems to name and rank them similar to ways hurricanes are categorized.
On the other hand, data that aligns with our perspective or makes us feel maligned, definitely grabs our attention.
When we feel vindicated by research findings, we pay attention because they “prove us right.” We’re more likely to share them with our social networks – especially if we think they might help the “other side” see the error of their ways.
But doing so is a fool’s errand. When data feels like an attack on our beliefs, we pay attention not for the purpose of expanding our understanding, but rather to search for errors and bias in the research to justify our rejection of the information. As much as we like to think of ourselves as rational creatures, our natural inclination is to find flaws in anything that challenges our understanding of the world.
Pick something you know with every fiber of your being to be true. Now imagine your reaction if someone posted a study proving the opposite. Would you be curious to learn more, or would you brush the study off as ridiculous, wrong, or even dangerous?
Look, if the study had found that women shared 9x more untrustworthy links than men, I’d immediately begin looking for flaws in the research design and evidence of misogynistic bias. If it said people of color share 9x more untrustworthy links than white people, I’d dismiss the findings outright as a product of systemic and structural racism in research institutions.
We have to be intellectually honest about how we’d perceive data that contradicted our understanding of the world in order to have empathy for people who dismiss data that contradicts theirs.
Data points like the ones at the start of this post can be perceived as an indirect form of shaming – and as Amanda Ripley says in High Conflict: Why We Get Trapped and How We Get Out, shaming usually backfires:
It almost always makes the opponent stronger. Especially if someone from another group does the shaming. It cements the division, bringing the other side closer together in fear or anger, emboldening them. It affirms the belief that they are on the side of good and their critics are on the side of evil.
Which means that, for some people, the more data disproves their beliefs, the more fiercely they cling to those beliefs and to others who share them. It becomes more important to define themselves against those with different views, “picking out certain points of contrast and exaggerating or idealizing them – eventually trying to act , in some respects, as little like [the other] as possible,” say the late David Graeber, professor of anthropology at the London School of Economics, and David Wengrow, professor of comparative archeology at University College London, in The Dawn of Everything: A New History of Humanity.
But it’s just data! Data is neutral! How can people define themselves against data? How can sharing data be perceived as shaming? Well, no one wants to be associated with untrustworthiness. And the study associates a specific group of people — Republicans — with 9x (!) more untrustworthy sharing than Democrats, the group of people Republicans define themselves against.
It’s only natural for people who identify as Republicans to perceive the data as an unfounded attack. It has to be mistaken, because, if it’s accurate, it means Republicans are the ones disseminating “fake news” at alarmingly high rates … not Democrats.
Unsurprisingly, conservatives have long had issues with NewsGuard, the database the Gratz and Bristol researchers used to assign trustworthiness values to the links being shared by US, UK, and German elected officials. Conservatives say Newsguard, a self-proclaimed “independent, apoliticial trust rating for online news sources,” rates left-leaning media sources as more trustworthy than right-leaning ones.
There’s a study to prove it. The Media Research Center, which, in its own words is “America’s leading media watchdog in documenting, exposing and neutralizing liberal media bias,” released a study in December 2021 saying:
The average NewsGuard score for the “left” and “lean left” outlets — which included leftist outlets like Jacobin and The Nation — was 93/100. While the average rating for “right” and “lean right” outlets — which included Fox News, Washington Times and New York Post — was a low 66/100.
Fox News, however, had been consistently rated by NewsGuard as “trustworthy” until last week when it was downgraded to “proceed with caution.” NewsGuard now says Fox News “fails to adhere to several basic journalistic standards” and “has published numerous false and misleading claims, including about politics and COVID-19.” The move was met with swift criticism by conservatives pointing to NewsGuard’s left-leaning bias.
Nandini Jammi, co-founder of adtech watchdog group Check My Ads Institute, agrees that NewsGuard is biased, but for a completely different reason. She says in a LinkedIn post that NewsGuard “omitted Fox News’ ongoing role in promoting Stop The Steal and The Big Lie during the 2020 election cycle, from their evaluation of Fox News” and was only able to maintain a “trustworthy” rating for FoxNews because of those omissions.
Yet another study, though, says NewsGuard doesn’t impact news consumption behavior of most people anyway. Researchers from the Center for Social Media and Politics at New York University and the Department of Politics at Princeton University found that a NewsGuard credibility rating “does not measurably improve news diet quality or reduce misperceptions, on average, among the general population.” In discussing their findings, the researchers noted that this is “a more partisan age in which attitudes toward news sources are strongly correlated with partisanship.”
It’s another example of how data doesn’t change minds. Third-party credibility ratings don’t impact people’s perception of a news source’s trustworthiness because they’ve already decided a source is trustworthy based on other people they trust (members of their party) saying a source is trustworthy.
And around and around it goes.
Our social media posts about data aren’t going to change the minds of our connections who hold different views. Those posts might actually push our connections further away from us and from the ideas we want them to consider.
Does that mean we should avoid posting about data at all? I don’t think so. But crafting relationship-building posts around data requires a bit more effort than simply sharing the link and a snippet from the article or study. Here are some suggestions (thanks to digital ad and advocacy expert Regan Opel, founder of Anonymous Gazelle, for brainstorming with me):
Explain why we decided to share the data, focusing on why we believe the information is important to building a world where everyone thrives, and asking for people to share their thoughts about whether the data helps achieve that goal.
Use the data as a conversation-starter, making it clear that all voices are welcome as long as they remain respectful of others. Try a caption like, “I found this research interesting. Given that we all have different experiences and perspectives, I’m curious about how this data makes you feel and why. Please share your stories and be considerate of the stories of others.”
Incorporate data points into posts about our lives, contextualizing data through human experience.
Maintaining contact and open dialogue matters. There is a positive flip side to the human tendency to define ourselves against one another when cultures collide. As Graeber and Wengrow point out, “it is only to be expected that people on both sides of the divide will learn from one another and adopt each other’s ideas, habits, and technologies.”
If we share more of our stories, experiences, values, and visions than facts and data devoid of meaningful context, we’re far more likely to learn from each other.
What stories and experiences do you have about sharing data in digital spaces? Asking for myself because I DID post the statistic from the opening paragraph on Twitter, then hesitated before posting it on Facebook and wrote this instead. I eagerly await your thoughts — respectful disagreement welcome!
ALSO BY LAUREN M. HUG …
Superb thought-provoking writing, as ever. Thank you Lauren.