Many Americans believe that they are regularly being exposed to misinformation online, and most are confident in their ability to distinguish real news from fake news.
However, Americans' confidence in their ability to spot fake news isn't always grounded in reality. A YouGov survey of 1,516 U.S. adult citizens tested Americans' susceptibility to falsehoods in news, and found that on average, they failed to correctly classify one-third of headlines as either real or fake. And certain segments of the population are even more likely to be fooled by misinformation. In particular, the poll finds that there are certain subgroups who are more likely than others to fail to distinguish fake headlines from real ones: Younger adults perform worse than older adults, Republicans perform worse than Democrats, and people who get their news from social media perform worse than people who do not.
How susceptible Americans are to misinformation overall
To determine how susceptible Americans are to misinformation, YouGov showed U.S. adult citizens a set of 20 headlines — 10 of which were real and 10 of which were fake headlines — in a randomized order. The survey participants were asked whether they believed each headline was real or fake. This survey design comes from the Misinformation Susceptibility Test, a framework developed by researchers at the University of Cambridge.
The fake news headlines, which were created with the assistance of artificial intelligence, were designed by the Cambridge researchers to encompass a wide range of misinformation properties — such as a person's tendency to accept statements that are meaningless or false and their belief in conspiracy theories. To select the real news headlines, the researchers used the Media Bias Fact Check database to identify news sources marked as the least biased while scoring very high on factual reporting. For both the real and fake sets of headlines, the researchers used several methods to narrow more than 400 potential headlines down to 20. That process yielded five real headlines from the Pew Research Center, three from Gallup, and one each from the Associated Press and Reuters.
Overall, Americans correctly classify a headline about two-thirds of the time. The average number of correctly classified headlines is 13 out of 20, with similar accuracy overall for the real and fake headlines; the average for each set is about 7 out of 10. (This article refers to the weighted mean number of correctly identified headlines as the average; the mean and median scores are the same for Americans overall.)
The headline that Americans are most likely to correctly identify as real is from a Pew Research story: "Republicans Divided in Views of Trump’s Conduct, Democrats Are Broadly Critical." Four in five Americans (80%) correctly identify that as a real headline, with Democrats (78%) and Republicans (83%) being similarly likely to answer correctly. Majorities of adults under 45 (70%) and Americans 45 and older (88%) identify it as real.
The real headline that Americans are most likely to identify incorrectly — by labeling it as fake — was another Pew Research headline: "Attitudes Toward EU Are Largely Positive, Both Within Europe and Outside It." While 61% of Americans correctly identify this headline as real, 39% incorrectly say it is false. Democrats (70% correct) are more likely than Republicans (54% correct) to correctly categorize this headline. Adults under 45 (59%) and Americans 45 and older (62%) are similarly likely to identify it as real.
When it comes to the fake, AI-generated headlines that were designed to feature common misinformation techniques, Americans are most likely to incorrectly classify two headlines that focus on distrust in government: "Government Officials Have Manipulated Stock Prices to Hide Scandals" and "The Government Is Manipulating the Public's Perception of Genetic Engineering in Order to Make People More Accepting of Such Techniques."
Slightly fewer than half (46%) of Americans incorrectly say "Government Officials Have Manipulated Stock Prices to Hide Scandals" is a real headline, with Republicans (53%) being more likely than Democrats (40%) to label it incorrectly. Republicans (60%) are even more likely to say the fake headline about the government manipulating people into being more accepting of genetic engineering is real, while only 36% of Democrats call it real.
Younger adults are more likely than older Americans to believe the fake headline about government officials manipulating stock prices (51% vs. 41%) is real, but the groups were similarly likely to incorrectly say that the headline about the government manipulating people into being more accepting of genetic engineering is real (46% vs. 45%).
Younger Americans are less skilled than older adults at identifying real from fake news
Younger Americans perform worse than older Americans on the misinformation test. They score an average of 12 out of 20 correctly, compared to 15 for older adults. The mean and median number of correctly identified headlines are the same for younger and older adults. For the purposes of analysis, YouGov split the number of correct responses into roughly four even groups. People in the highest scoring category were those who got 17 to 20 correct, people in the "upper middle" group got 14 to 16 correct, people in "lower middle" got 11 to 13 correct, and "bottom scoring" got 10 or fewer correct.
When looking at these groupings by age, we found that Americans who are 45 and older perform better than adults under 45 did. For instance, 37% of adults under 45 are in the bottom scoring group, compared to 14% of people who are 45 and older. Meanwhile, just 15% of adults under 45 were in the highest scoring group, compared to 35% of older adults.
Younger Americans perform worse than older Americans in their ability to identify real headlines (scoring an average of 6 out of 10, compared to 8 for people who are 45 and older) as well as slightly worse in their ability to identify fake headlines (scoring an average of 6 out of 10, compared to 7 for people who are 45 and older). This indicates that younger Americans may be more likely than older adults to suspect misinformation where there is none and may trust news headlines less overall. For instance, a relatively straightforward real headline, "Hyatt Will Remove Small Bottles from Hotel Bathrooms," was marked as real by just 55% of 18- to 44-year-olds, compared to 72% of older Americans.
We considered that younger Americans could have been more likely to skip questions in the survey — which would result in a question being automatically scored as incorrect — but there was no significant difference in the skip rate among 18- to 44-year-olds and people 45 and older.
There is another reason why the presumably digital-savvy younger Americans may have performed worse than older adults on this misinformation test. Young adults self-report spending more time online than people who are 45 and older do, something that is closely tied to people's susceptibility to news misinformation. Just 23% of 18- to 44-year-olds report spending 2 or less hours of recreational time online each day, compared to 44% of people who are 45 and older.
People who spend two or fewer hours of recreational time online each day are twice as likely as people who spend 9 or more hours online per day to be in the highest-scoring category (30% vs. 15%). Two-thirds of people who spend the most time online each day (67%) land in the lower-middle or lowest-scoring categories, compared to 41% of people who spend the least amount of time online.
In addition to their greater amount of recreational time spent online, young adults also are much more likely than older adults to report using social media as a source for news. Our research indicates that people who use social media as a news source do not perform as well on the misinformation test.
The fake headlines that younger adults are more likely than older Americans to be fooled by include "Government Officials Have Manipulated Stock Prices to Hide Scandals" (51% of 18- to 44-year-olds incorrectly label this as real, compared to 41% of people 45 and older) and "The Corporate Media Is Controlled by the Military-Industrial Complex: The Major Oil Companies Own the Media and Control Their Agenda" (42% of 18- to 44-year-olds incorrectly label this as real, compared to 20% of people 45 and older).
Republicans are more susceptible than Democrats to misinformation
There is also a significant relationship between partisanship and susceptibility to misinformation: Republicans perform worse than Democrats and Independents on the misinformation test. Democrats and Independents performed similarly to one another overall. One-third of Democrats and 28% of Independents are in the highest-scoring group, compared to 14% of Republicans.
On average, Republicans score lower than Democrats and Independents on their ability to spot the fake headlines. There are a few fake news headlines that a majority of Republicans are most likely to be tricked by: "The Government Is Manipulating the Public's Perception of Genetic Engineering in Order to Make People More Accepting of Such Techniques" (60% of Republicans incorrectly label it as real), "New Study: Left-Wingers Are More Likely to Lie to Get a Higher Salary" (56% Republicans incorrectly label it as real), and "Left-Wing Extremism Causes 'More Damage' to World Than Terrorism, Says UN Report" (54% of Republicans). Fewer than 40% of Democrats incorrectly label each of these three headlines as real.
It appears that Democrats and Republicans are more likely to label as real the headlines that reinforce positive sentiment about their own party or negative views about the opposing party — regardless of whether the headlines really are real or fake.
For instance, Republicans (82%) are much more likely than Democrats (67%) to correctly label this Pew Research Center headline as correct: "International Relations Experts and US Public Agree: America Is Less Respected Globally," potentially because it appears to reinforce a negative view about a Democratic president. The headline is from 2018, when Donald Trump was president, but respondents were not given the dates of the headlines and responded with Joe Biden in the White House.
Democrats (80%), on the other hand, are much more likely than Republicans (68%) to correctly label as real the Pew Research headline, "Democrats More Supportive than Republicans of Federal Spending for Scientific Research." Democrats (70%) also are more likely than Republicans (54%) to correctly classify as real the Pew Research headline, "Attitudes Toward EU Are Largely Positive, Both Within Europe and Outside It."
How media consumption influences susceptibility to misinformation
For this analysis, YouGov also explored the connection between people's sources of news and how well they performed on the misinformation test. YouGov showed respondents 56 news sources in four categories: broadcast, print, digital, and social media. They could tell us which items, if any, they had personally used as a source for news in the last month. A majority of Americans (85%) use more than one source from among the 56 sources we asked about, meaning that there is significant overlap in many of the audiences. There were 56 news sources asked about, but four were excluded from reporting because the sample size was too small.
We analyzed the top and bottom performers among media outlets' audiences using four metrics: by median score, by mean score, or by the percentage of consumers who were in the highest and lowest scoring group. By each of the four metrics, people who say they consume news from the Associated Press are at the top for all categories of telling real from fake headlines, either alone or in a tie. AP news consumers get the highest average for correctly identifying headlines (16 out of 20), are tied for the highest median correct score (16.5 out of 20), have the lowest score for average incorrect answers (3.99), and are tied for the lowest median incorrect score (2.5). Most people who say they use the AP as a news source (55%) land in the highest-scoring category.
People who get their news from National Public Radio, Axios, HuffPost, and PBS are more likely than consumers of most other outlets to be in the highest-scoring group. Consumers of MSNBC, The Hill, Reuters, Politico, and the Washington Post also do comparatively well on the test.
The bottom performers among news sources, in terms of the share of people who get their news from the source and place in the highest-scoring category, are the social media platforms of WhatsApp, Truth Social, Snapchat, and TikTok. Consumers of a few right-leaning outlets, including OAN and Newsmax, also are among the lowest percentage who place in the highest-scoring category.
If instead news sources are ranked on the basis of the highest percentage of their news consumers who land in the lowest scoring-category, the worst performers generally are social media platforms. That includes Snapchat (53% of people who get news from it land in the lowest-scoring category), Truth Social (45%), WhatsApp (44%), TikTok (41%), and Instagram (38%). People who say they use no broadcast news outlet as a news source also perform poorly (39% in the lowest-scoring category).
Americans' views on their own ability to discern misinformation
A majority of Americans (53%) say they see what they think is false or misleading information online every day. Seven in 10 adults (71%) believe they are being exposed to disinformation at least once per week. Older Americans are much more likely to believe they are being shown online falsehoods every day. Two-thirds of people who are 65 and older (66%) say they see misleading information on a daily basis, compared to 37% of adults under 30.
Three-quarters of Americans (75%) are very confident (23%) or somewhat confident (52%) in their ability to tell real news from fake news. There are similarly high levels of confidence among older and younger people in their skills distinguishing real news from fake news, but Americans who are 45 and older tend to have a bit more confidence (80%) when compared to younger adults (69%). Democrats (79%) and Republicans (78%) are similarly confident that they would be able to tell real news from fake news.
People who are very or somewhat confident that they can tell real news from fake news score higher on the misinformation test than people who say they are not very confident or not confident at all. Three in 10 people who are confident land in the highest-scoring category, compared to 12% of people who are less confident.
— Carl Bialik and Taylor Orth contributed to this article. Rakoen Maertens, Joe Williams, and Ian Davis contributed to analysis. Kathy Frankovic and Douglas Rivers contributed to editing.
See the results from this poll and email uspress@yougov.com for additional information
Related:
- Trust in Media 2023: What news outlets do Americans trust most for information?
- Trust in Media 2022: Where Americans get their news and who they trust for information
- How often and where Americans get information on the weather
Methodology
Overall: The poll was conducted among 1,516 U.S. adult citizens from April 3 - 9, 2023. Respondents were selected from YouGov’s opt-in panel using sample matching. A random sample (stratified by gender, age, race, education, geographic region, and voter registration) was selected from the 2019 American Community Survey. The sample was weighted according to gender, age, race, education, 2020 election turnout and presidential vote, baseline party identification, and current voter registration status. Demographic weighting targets come from the 2019 American Community Survey. Baseline party identification is the respondent’s most recent answer given prior to June 1, 2022, and is weighted to the estimated distribution at that time (34% Democratic, 31% Republican). The margin of error for the overall sample is approximately 3%.
News Outlets: YouGov asked respondents whether they had consumed news from 56 possible news sources in the last month. These were split into four categories (broadcast, digital, print, and social media) and asked as a multi-select question. For each category, respondents could say "None of these." The outlets asked about were: ABC, Al Jazeera, Axios, BBC, Bloomberg, Breitbart News, Business Insider, BuzzFeed News, C-SPAN, CBS, CNN, Daily Kos, Facebook, Forbes, Fox News, HuffPost, Infowars, Instagram, LinkedIn, MSNBC, National Public Radio (NPR), National Review, NBC, New York Post, Newsmax, Newsweek, One America News (OAN), Parler, Politico, Public Broadcasting Service (PBS), Reddit, Reuters, Slate, Snapchat, The Associated Press, The Atlantic, The Daily Beast, The Daily Caller, The Economist, The Federalist, The Guardian, The Hill, The New York Times, The New Yorker, The Wall Street Journal, The Washington Examiner, The Washington Post, The Weather Channel, TikTok, TIME Magazine, Truth Social, Twitter, USA Today, WhatsApp, Yahoo News,YouTube. From this list, four outlets (Parler, Infowars, Daily Kos, and Slate) had too small of a sample and were cut from analysis.
Editor's Notes:
- YouGov worked with the creators of the Misinformation Susceptibility Test to slightly modify 3 of the 20 headlines. Both YouGov and the MIST developers are confident that the modified headlines retain the same psychometric properties.
- In this study, the real headlines came from outlets that were determined to be the least biased and most factual. There was one headline included from The Associated Press ("Hyatt Will Remove Small Bottles from Hotel Bathrooms") and one from Reuters ("Morocco’s King Appoints Committee Chief to Fight Poverty and Inequality"). The AP and Reuters were the only outlets included in the poll whose headlines were included in the MIST test.
- A programming error caused The Economist to be included in the survey twice, once alongside print news sources and once alongside digital news sources. Respondents saw The Economist listed on two separate pages. The results from the digital sample were removed.
Image: Adobe Stock (awrey)
