Effect of corrections

Poynter Results

  • Correcting historical misperceptions works — but it's not magic

    This study seeks to explain whether or not corrective information affects the views Jewish Israelis hold about the conflict with Palestine. Researchers randomized an experiment in which an online sample of 2,170 Jewish Israelis ages 18 or older either received solely an extremist message, which denied Israeli wrongdoing in the 1948 Palestinian exodus, or that message plus corrective information about the conflict. They also randomized participants’ feelings of high or low control. While the proportion of Jewish Israelis who denied wrongdoing in the conflict with Palestine increased by 8 percent from the baseline to the low control, uncorrected condition, the prevalence of denialism decreased by between 5 and 11 percent for the inverse conditions. The findings suggest that when people are induced to feel a lack of control, they’re more vulnerable to a denialist message — but corrective information is still quite effective.

    Study Title
    Fighting the Past: Perceptions of Control, Historical Misperceptions, and Corrective Information in the Israeli-Palestinian Conflict
    Study Publication Date
    Study Authors
    Brendan Nyhan, Thomas Zeitzoff
    Journal
    Political Psychology
    Peer Reviewed
    Yes
  • Giving corrective economic information works, but it doesn't change views

    The study supplied people with past economic data, then asked them what they think of the current state of the U.K. economy. The researchers found that while partisanship was a key part of how people viewed the economy in the U.K., most people’s economic perceptions were rooted in real economic indicators, like job growth and unemployment. And — most importantly — people who held inaccurate views of the economy generally changed them when presented with corrective information. The essential results are similar to those posited by other work — corrections work, but only to a certain extent.

    Study Title
    Facing up to the facts: What causes economic perceptions?
    Study Publication Date
    Study Authors
    Catherine E. De Vries, Sara B. Hobolt, James Tilley
    Journal
    Electoral Studies
    Peer Reviewed
    Yes
  • Labeling some fake stories on social media increases the believability of untagged fake stories

    This study examines the effects of adding disputed labels to fake news stories on social media outlets like Facebook, in line with the real partnership the social network launched in December 2016. While researchers found that adding warnings to fake content decreased those posts perceived accuracy, they also found that the presence of fake news tags alone also increased the perceived accuracy of untagged fake stories. This "implied truth" effect was stronger among subgroups who were more likely to believe online information (such as young adults and Trump supporters). Participants saw an equal mix of right and left-wing headlines, both fake and real, and answered questions about their validity and shareability.

    Study Title
    The Implied Truth Effect: Attaching Warnings to a Subset of Fake News Stories Increases Perceived Accuracy of Stories Without Warnings
    Study Publication Date
    Study Authors
    Gordon Pennycook, David G. Rand
    Journal
    SSRN
    Peer Reviewed
    No
  • Low critical thinking may determine whether you believe in fake news

    Respondents were showed "Facebook-like" posts carrying real or fake news. Across three different study designs, respondents with higher results on a Cognitive Reflection Test were found to be less likely to incorrectly rate as accurate a fake news headline. Analytic thinking was associated with more accurate spotting of fake and real news independent of respondents' political ideology. This would suggest that building critical thinking skills could be an effective instrument against fake news.

    Study Title
    Who Falls for Fake News? The Roles of Analytic Thinking, Motivated Reasoning, Political Ideology, and Bullshit Receptivity
    Study Publication Date
    Study Authors
    Gordon Pennycook, David G. Rand
    Journal
    SSRN
    Peer Reviewed
    No
  • The most effective way to fact-check is to create counter-messages

    Researchers examined a final selection of 20 experiments from 1994 to 2015 that address fake social and political news accounts in order to determine the most effective ways to combat beliefs based on misinformation. The headline finding is that correcting misinformation is possible, but it's often not as strong as the misinformation itself. The analysis has several take-aways for fact-checkers, most notably the importance of creating counter-messages and alternative narratives if they want to change their audiences’ minds and getting on to the correction as quickly as possible.

    Study Title
    Debunking: A Meta-Analysis of the Psychological Efficacy of Messages Countering Misinformation
    Study Publication Date
    Study Authors
    Man-pui Sally Chan, Christopher R. Jones, Kathleen Hall Jamieson, Dolores Albarracín
    Journal
    Psychological Science
    Peer Reviewed
    Yes
  • Twitter users are more likely to accept correction from people they know

    The study looked at corrections made on Twitter between January 2012 and April 2014 to see how fact-checking is received by people with different social relationships. Researchers ultimately isolated 229 “triplets” where the person sharing a falsehood responds to a correction by a second tweeter. Corrections made by “friends” resulted in the person sharing a falsehood accepting the fact 73 percent of the time. Corrections made by strangers were accepted only 39 percent of the time. Put simply: When we’re wrong on Twitter, we’re more likely to own up to it if someone we know corrected us.

    Study Title
    Political Fact-Checking on Twitter: When Do Corrections Have an Effect?
    Study Publication Date
    Study Authors
    Drew B. Margolin, Aniko Hannak, Ingmar Weber
    Journal
    Political Communication
    Peer Reviewed
    Yes
  • Voters gradually change their opinions when presented the facts

    In this study, respondents were given a factual question like "From 2009, when President Obama took office, to 2012, median household income adjusted for inflation in the United States fell by more than 4 percent" and asked to rate it as "True" or "False." Over the course of four subsequent rounds, they were given signals that the information was indeed accurate or not and told that these signals were right 75% of the time. The results indicate that respondents updated their beliefs towards the correct answer regardless of their partisan preference. The study's elaborate design makes it hard for fact-checkers to draw real life lessons. However, it does seem to offer additional evidence that fact-checking doesn't fall on deaf ears.

    Study Title
    Learning Together Slowly: Bayesian Learning about Political Facts
    Study Publication Date
    Study Authors
    Seth J. Hill
    Journal
    The Journal of Politics
    Peer Reviewed
    Yes
  • Fact-checking Marine Le Pen corrected misperceptions but didn't affect voting preferences

    Researchers surveyed French individuals online in four regions where the far-right Front National party (FN) had done best in the 2016 regional elections. Respondents were put into one of four groups; the first received false claims on immigration made by Marine Le Pen, the FN's presidential candidate and the second obtained statistics on the same issues. The other two groups were given both or neither, respectively. Across all groups, the researchers tested respondents' understanding of the facts, their support for Le Pen on immigration and their voting intentions. Overall, knowledge of the facts was negatively affected when respondents only read Le Pen's claims but improved when they were offered the facts alone or both the facts and Le Pen's claims. More surprisingly, the intention to vote for Le Pen improved not just among respondents subjected to her claims but also among respondents who were offered the facts alone.

    Study Title
    Facts, Alternative Facts, and Fact Checking in Times of Post-Truth Politics
    Study Publication Date
    Study Authors
    Oscar Barrera, Sergei Guriev, Emeric Henry, Ekaterina Zhuravskaya
    Journal
    SSRN
    Peer Reviewed
    Yes
  • Social media comments are just as effective at correcting health misinformation as algorithms

    This study measures the extent to which algorithms and comments on Facebook that link to fact checks can effectively correct users' misconceptions about health news. Researchers tested this by exposing 613 survey participants to simulated news feeds with three condition. Participants were shown misinformation about the Zika virus and different corrective news stories either surfaced by algorithm or posted by another Facebook user. The experimental results found that algorithmic and social distribution of fact checks were equally effective in limiting participants' misperceptions — even for people who are more inclined to believe conspiracy theories. Researchers conclude that this is likely because breaking health news events often deal with new phenomena, which allows for great receptivity to comments and the possibility of opinion change among news consumers early on.

    Study Title
    See Something, Say Something: Correction of Global Health Misinformation on Social Media
    Study Publication Date
    Study Authors
    Leticia Bode, Emily K. Vraga
    Journal
    Health Communication
    Peer Reviewed
    Yes
  • On social media, users believe corrections if they include sources

    This study attempts to determine the most effective way to correct misinformation on social media by testing both the content of corrections and how they're presented. In a survey with 613 valid responses, of which 271 were analyzed, participants saw either a simulated Facebook or Twitter feed and were assigned to one of three conditions with varying levels of misinformation and corrections, both with and without sources. Based on the experimental results, researchers found that, when everyday users share corrections on social media, linking to credible sources increases the probability that other users will believe the corrections. In the control condition, in which participants weren't shown corrections with sources, misperceptions were largely unaltered. On Facebook, linked sources in comments on articles led to increased perceptions of credibility, while the same effect was absent in Twitter replies.

    Study Title
    I do not believe you: how providing a source corrects health misperceptions across social media platforms
    Study Publication Date
    Study Authors
    Emily K. Vraga, Leticia Bode
    Journal
    Information, Communication & Society
    Peer Reviewed
    Yes
Email IconGroup 3Facebook IconLinkedIn IconsearchGroupTwitter IconGroup 2YouTube Icon