The following is an essay written by an SSRC Social Data Dissertation Fellow, based on her project. The program was generously funded by the Omidiyar Network.
Social media misinformation is likely to persist as an enduring challenge for democratic societies. Although research has provided comforting evidence that online misinformation, usually shared through social media, likely occupies only a small portion of users’ information diets, social media misinformation have been a prominent topic in news and in public opinion in recent years. Most Americans view social media as having a negative influence on US society, citing misinformation as the primary reason. In fact, public opinion polls show that misinformation has become a more salient concern for Americans than crime and terrorism.
Given that many consider misinformation to be a serious issue, a savvy social media user might ask, “How do I become a good skeptic?” The answer to this question is more complicated than it may initially seem. Skepticism, a subject feeling of distrust, has been previously studied by researchers in the context of news media skepticism. News media skepticism can be a roadblock toward fostering an informed citizenry, as it usually discredits the journalism practices that produce credible information. On the other hand, skepticism may become necessary and beneficial in contexts where there is a valid concern about potential misinformation, such as on social media. However, social media skepticism still may not always lead to pro-democratic outcomes. While being skeptical towards social media misinformation could be a sign of having digital media literacy, people can vastly overestimate the prevalence of social media misinformation and be skeptical of information that is in fact true. Skepticism may also be cast through an identity lens and used for political ends, such as blaming social media misinformation on the opposing group, which can deepen the political divide over truth and falsehoods.
Does skepticism towards social media misinformation mean different things to different people? Does skepticism help people become good citizens, or can it actually be harmful? To unpack the different manifestations of social media skepticism, I distinguish between “accuracy-motivated” skepticism and “identity-motivated” skepticism toward social media misinformation according to their different motivations. Accuracy-motivated skepticism is a type of skepticism driven by people’s concern about how social media misinformation affects one’s knowledge and decision-making. Identity-motivated skepticism, on the other hand, is a type of skepticism driven by a concern about how social media misinformation is related to competing ideologies and political (dis)advantages.
My findings show that “accuracy-motivated” skepticism can lead to several positive outcomes: when people’s skepticism towards social media misinformation is driven by accuracy motivations, they are more likely to seek news on social media and find out more information about an issue. Accuracy-motivated skepticism can also attenuate the partisan divide in perceiving social media misinformation as mostly favoring the other side. In contrast, “identity-motivated” skepticism can be counterproductive: identity-motivated skepticism turns people off from using social media for news, potentially hindering political learning. Identity-motivated skepticism also exacerbates the partisan divide in interpreting social media misinformation, making people more likely to attribute misinformation to their political opponent. Further, identity-motivated skepticism towards misinformation can also influence how people view elections, especially for the losing side. When people’s skepticism towards social media misinformation is driven by identity reasons, they are likely to attribute losing the election to the influence of misinformation and use misinformation to justify the view that the disliked election result was illegitimate. Simply being skeptical is not enough. Rather, accuracy motivation is a necessary component of a healthy type of skepticism crucial for becoming a responsible social media user and informed citizen.
What does it mean to be a skeptical social media user?
Skepticism towards social media misinformation can take various forms. I call one type of skepticism “accuracy-motivated” skepticism. This type of skepticism is driven by people’s concern about how social media misinformation affects one’s knowledge and decision-making. For example, people may worry that they will believe stories on social media that later turn out to be false, or that misinformation on social media will negatively affect their own decisions, such as what they buy or who they vote for. They are skeptical of information on social media partly because they recognize that everyone, including themselves, carry biases that may lead them to be tricked by false stories.
On the other hand, there is a second type of skepticism that I call “identity-motivated” skepticism. Identity-motivated skepticism is driven by a concern about how social media misinformation is related to competing ideologies and political (dis)advantages. Given that the public discourse about social media misinformation is tightly associated with the discourse about political campaigns and elections, people may not trust what others share on social media, worrying that others share false stories that favor their own group. Such skepticism may be driven by the belief that on social media, there are too many lies by partisan elites or stories fabricated to make their candidate or side look bad. Some may even feel that information on social media often does not represent “their own truth.”
Accuracy-motivated and identity-motivated skepticism are two distinct concepts that are only weakly correlated to each other. That is, a person can hold both types of skepticism at the same time and holding a type of skepticism at a high level does not necessarily mean holding the other type of skepticism at either a high or a low level. In my survey, I asked people to answer three questions about accuracy-motivated skepticism and five questions about identity-motivated skepticism on a 7-point scale (1 = strongly disagree, 7 = strongly agree).
Taking a quick look at the distribution of accuracy-motivated skepticism: on average, about a third of the participants reported a low level of accuracy-motivated skepticism (those who answered strongly disagree, disagree, or somewhat disagree); a third reported a high level of accuracy-motivated skepticism (those who answered somewhat agree, agree, or strongly agree), and a third scored in the middle.
Age makes a difference in the extent that people hold accuracy-motivated skepticism. Younger adults were more likely to hold accuracy-motivated skepticism than older adults. 34.6% respondents aged 18-29 reported a high level of accuracy-motivated skepticism, compared to 32.2% respondents aged 30-49, 27.0% respondents aged 50-64, and 24% respondents aged 65 and above.
Partisanship, on the other hand, does not strongly influence the extent of accuracy-motivated skepticism people hold, although Independents were more likely to score in the middle than Republicans and Democrats.
The picture looks very different when it comes to identity-motivated skepticism. Using the same definitions of low, medium, and high levels of skepticism as above, only 0.9% respondents reported a low level of identity-motivated skepticism and 27.4% reported a medium level. Majority of the respondents (71.7%) scored high on the scale, holding skepticism toward social media driven by political considerations.
Importantly, partisans held identity-motivated skepticism to different degrees. 79.7% Republicans held a high level of identity-motivated skepticism, compared to 68.0% Democrats and 60.9% Independents, coinciding with research that shows partisan asymmetry in the tendency to engage in biased reasoning.
Moreover, older adults were more likely to hold identity-motivated skepticism than younger adults. 77.5% respondents aged 65 and above reported a high level of identity-motivated skepticism, while the number decreased to 53.8% among respondents aged 18-29.
Skeptics on social media
The two types of skepticism shape how people perceive and engage with social media. While accuracy-motivated skepticism towards social media can encourage people to seek more news on social media, directional-motivated skepticism can turn people off from using social media for news. 19.8% of those with high accuracy-motivated skepticism reported using social media for news fairly or very often in the last few weeks, compared to 13.9% of those with low or medium accuracy-motivated skepticism combined.
In contrast, 13.2% of those with high identity-motivated skepticism reported using social media for news fairly often or very often in the last few weeks, while 21.6% of those with low or medium identity-motivated skepticism have used social media for news fairly often or very often. This suggests that skepticism towards misinformation may lead people to adjust their behaviors on social media, and exactly how they shift their behaviors is dependent on their motivations. A skeptical user who is motivated by accuracy will seek more news to cross-check their information, which may lead to virtuous cycles of knowledge gain. However, when a user is skeptical about social media misinformation mainly for identity reasons, their skepticism might lead them to avoid getting news from social media, potentially resulting in less political learning and more closemindedness.
Further, the types of skepticism may also shape how people interpret bias in the social media misinformation they encounter. On average, Republicans saw misinformation on social media as mostly favoring Democrats, while Democrats saw it as mostly favoring Republicans. However, having a high level of accuracy-motivated skepticism attenuated this tendency to point fingers, moving both sides towards the middle ground. This effect was more salient for Democrats: having a high level of accuracy-motivated skepticism helped Democrats more than it helped Republicans in decreasing the tendency to see the misinformation as favoring the other side.
Identity-motivated skepticism had the opposite effect. Having a high level of identity-motivated skepticism exacerbated the tendency for partisans to perceive social media misinformation as favoring the other side, moving them towards the more extreme ends of partisan reasoning. This was especially salient for Republicans, who showed a prominent tendency to interpret the social media misinformation as favoring Democrats when having a high level of identity-motivated skepticism.
Identity-motivated skepticism and election legitimacy
As misinformation has become an important part of how citizens make sense of politics, skepticism towards misinformation, when driven by identity motivations, may be used as a way for citizens to explain unfavorable political outcomes. In my survey conducted in December 2020, respondents answered 1) how much they thought social media misinformation influenced the result of the 2020 election and 2) whether they thought the 2020 election was legitimate. While accuracy-motivated skepticism towards social media was not related to opinions about the 2020 election, identity-motivated skepticism towards social media was significantly associated with how people form perceptions about the 2020 election.
Particularly for Republicans, a high level of identity-motivated skepticism increased the extent to which respondents’ thought social media misinformation influenced the result of the 2020 US presidential election. Republicans were more likely than Democrats to say that social misinformation influenced the 2020 election on average. Moreover, Republicans with a high level of identity-motivated skepticism were particularly likely to say so. In contrast, Democrats showed ambivalence to this question by and large.
Such relationships between identity-motivated skepticism and the perception of misinformation swaying election results might also contribute to how people think about the legitimacy of the 2020 US presidential election overall. Democrats, regardless of their levels of identity-motivated skepticism, strongly believed that the 2020 election was legitimate, including that it was decided in a fair way, that Joe Biden was the rightful winner, and that Joe Biden’s presidency was legitimate. However, when asked the same questions, Republicans cast doubt on the legitimacy of the 2020 election; in particular, Republicans with a high level of identity-motivated skepticism were particularly likely to see the 2020 election as illegitimate.
Taken together, these findings show that depending on the underlying motivation, skepticism towards social media misinformation can produce distinct outcomes in terms of how people engage with and perceive social media, and how people view elections. Being a good skeptic means that one’s skepticism is driven by the motivation to make accurate decisions and accompanied with the recognition of one’s own limitations. People with high levels of accuracy-motivated skepticism frequently seek news on social media and are less likely to assume that social media misinformation favors the opposing side. On the other hand, when identity becomes the anchor for skepticism towards social media misinformation, people could turn away from news on social media and more deeply engage in the weaponization of misinformation for political ends.
Understanding these different types of skepticism is important for how media practitioners and educators can better design implement digital literacy interventions that combat misinformation. Simply advocating for “being skeptical” is usually not enough since various forms of skepticism can cast contrasting effects on individuals’ use of social media and understanding of politics. Instead, it is crucial to design targeted intervention and education focused on accuracy-based norms, values, and skills, and avoid perpetuating identity biases in understanding truth and falsehoods.
How I conducted the research
I constructed and validated the scales to measure social media skepticism by conducting two pilot surveys using Amazon Mechanical Turk (987 respondents in total). Funded by the SSRC Social Data Dissertation Fellowship, I collected a two-wave panel survey through Qualtrics before and after the 2020 U.S. presidential election. I used quota sampling so that the features of gender, age, race, education, and income of the sample broadly matched the US census data. 1709 Respondents completed the first survey in the two weeks leading to the election day and were recontacted for a second survey two weeks after the election day; 826 respondents completed both waves. I analyzed the results using the data reported by the 826 respondents who completed both waves. In particular, I used respondents’ answers to social media skepticism and demographics in the first wave and their answers to social media news use, perception of misinformation on social media, and election legitimacy in the second wave to explore how different groups of people hold types of skepticism, and how types of skepticism shape people’s beliefs, attitudes, and behaviors.