A new study suggests that the habitual tagging of tweets containing falsehoods about vote tampering in the 2020 US presidential election with a "disputed" label might not only have no effect but could also cement misinformation among Trump voters.
Researchers John Blanchard and Catherine Norris released findings on how those very same labels affected voters' beliefs.
The Study: How 'Disputed' Tags Backfired
In December 2020, researchers surveyed 1,072 Americans who were presented with four tweets from Donald Trump making election fraud claims.
One group had the tweets labeled as "disputed," while the control group read the same tweets without any label. The experiment should measure how those kinds of tags influence perceptions of truthfulness and their belief in general levels of fraud during the election.
"These 'disputed' tags are meant to alert a reader to false/misinformation, so it's shocking to find that they may have the opposite effect," Norris said.
Read Also: Why Did DeportElonMusk Trend on X? Musk Deletes Harris-Biden Assassination Joke
Trump Voters Strengthen Beliefs Despite Fact-Checks
According to The Guardian, the experiment had a bit of a counterintuitive finding: Trump voters who went into the experiment doubting there was widespread fraud in the election became more likely to believe false claims with the "disputed" tags.
Surprisingly, this study provided evidence contradicting a previous consensus stating that fact-checking labels are nothing more than a futile effort in dealing with spreading misinformation for those in the politically knowledgeable group.
Blanchard noticed that instead of having no effect, the misinformation seemed to be counterproductive, reinforcing misinformation for this group.
Impact on Voting Groups
It is even more fascinating that voters for Biden were not affected much. Still, third-party voters or those who did not vote had a weak negative correlation with the belief in the existence of false claims based on reading the tagged tweets. This questions the efficacy of misinformation labels, especially in polarized political environments.
Why the Labels Backfired
Blanchard and Norris predicted that correcting labels would have less impact on politically well-informed voters, an idea for which much prior work had already been prepared because politically engaged people often ignore correction attempts in favor of their beliefs. Yet they found themselves surprised at how powerfully correcting labels entrenched misinformation among the high-knowledge voters.
The Harvard Kennedy School's Misinformation Review finding suggests that political distrust of the platforms- the suspicion that Twitter, in this case, was against conservatives- may have played a significant role in this reaction.
In fact, at the time of the study, conservatives were more antagonistic towards the perception of Twitter as an affront to their freedom through "disputed" tags, which may have elicited a doubling down on false claims.
Does Fact-Checking Effectuate Change?
This study contributes to the growing debate over how much suitable fact-checking labels do on social media. Among other companies, Twitter (now X) has tried a range of things: there have been "disputed" tags and newer Community Notes, which enable users to provide peer-reviewed context for posts. Still, as Blanchard and Norris' work shows, these may not consistently achieve what is hoped.
The Bigger Picture: Implications for Misinformation Research
The findings are significant for broader efforts to combat misinformation in today's highly polarized political landscape. Social media, news, and fact-checking organizations ought to reconsider their approach to determine whether it is helping or contributing to the spread of misinformation.
Although the study does not provide strong evidence of a broad backfire effect, it does indicate that terms like "disputed" are less persuasive for motivated and engaged Trump supporters.
As the political environment evolves and the X platform and other companies continue to make different decisions on content moderation, much more critical work must be done to understand how to most effectively challenge falsehoods online fully.
An Ongoing Battle Against Disinformation
So, with such claims of election fraud and deep divisions over the land still lingering, the fight against misinformation has a long way to go. Such studies have revealed how labeling and fact-checking patterns might shape different voter segments.
Along with social media, which keeps growing, an urgent need arises to realize the right balance between freedom of expression and curbing misinformation.
Related Article: Senate Committee Passes Bills to Combat AI Misinformation Ahead of US Elections
