Editor's Recommendations:
The proliferation of fake political news on social media has become a major headache in Western democracies today. This paper explores the root causes of political fake news sharing through rigorous empirical analysis from the perspective of citizens' psychological motivation and using Twitter as an observation platform. The authors found that partisanship and disagreement caused by party polarization are the main psychological motivations for citizens to share political fake news on social media. Therefore, in order to curb the spread of "fake news", simply checking the facts of the news is not a cure, and in the final analysis, it is necessary to solve the growing problem of social and political polarization in the United States. The article has a clear awareness of the problem and appropriate research methods, which enriches the academic community's understanding of the phenomenon of political party polarization in the United States and its consequences.
Political party polarization is the main psychological motivation behind the sharing of political fake news on Twitter
Summary:
The rise of fake news is a major concern in contemporary Western democracies. However, research on the psychological motivations of citizens to spread political fake news on social media is surprisingly limited. Are citizens who share fake news ignorant and lazy? Are they motivated by evil and trying to undermine the status quo in society? Or do you want to attack party rivals in an increasingly polarized political environment? This article is the first to test these competing hypotheses. The study is based on a detailed infographic that includes psychographic data linked to behavioral sharing data from more than 2,300 U.S. Twitter users and sentiment analysis of more than 500,000 news headlines. The results of the study refute the ignorant view, provide some support for the sabotage view, and provide strong support for the party polarization view. As a result, those who hate political opponents are most likely to share political fake news and selectively share news content that helps disparage political opponents. Overall, our findings suggest that the sharing behavior of fake news is driven by the same psychological motivations as other forms of partisan behavior, including sharing party news published by traditionally credible news sources.
About the Author:
Mathias Osmundsen, 奥胡斯大学Alexander Bor, 奥胡斯大学Peter Bjerregaard Vahlstrup, 奥胡斯大学Anja Bechmann, 奥胡斯大学Michael Bang Petersen, 奥胡斯大学
Compilation Source:
Osmundsen M., Bor A., Vahlstrup PB., Bechmann A., Petersen MB. 2021. “Partisan Polarization Is the Primary Psychological Motivation behind Political Fake News Sharing on Twitter.” American Political Science Review, 115(3):999-1015.
从上到下:本文作者(Osmundsen M., Bor A., Vahlstrup PB., Bechmann A., Petersen MB.)
I. Formulation of the problem
With the advent of social media, the spread of "fake news" has become a major social concern. Fake news can be defined as "fake information that mimics news media content in form, but differs from the news media in terms of content organization process or production intent" (Lazer et al. 2018). In order to curb the wave of fake news on social media, the government, academia and the media have invested a lot of manpower, material and financial resources. Yet, despite the huge investment, we know very little about why people share fake news on social media. A viable solution to curb the spread of fake news should be based on research on the psychological motivations of citizens to share fake news. For example, fact-checking for news can only be effective if citizens really want to share truthful information, but need outside help to weed out the misinformation. However, current research has largely failed to adequately identify the reasons why ordinary citizens participate in the spread of fake news (Wagner and Boczkowski, 2019), and few studies have analyzed people's motivations for sharing news (Chadwick, Vaccari, and O'Loughlin, 2018). Based on the shortcomings of existing research, the research question raised in this paper is: What are the psychological motivations of citizens to share political fake news on social media?
2. Theoretical basis and literature review
(1) The psychological motivation behind news sharing on social media
Cognitive psychologists have shown that there are two fundamentally different motivational goals that shape the way individuals interact with information: accuracy-oriented motivation and goal-oriented motivation (the latter is sometimes referred to as directional motivation) (Kunda, 1990). Accuracy-oriented motivation pushes individuals to only be exposed to and receive information that is true and reliable, thus forming an accurate description of the world. Conversely, goal-oriented motivation prompts people to focus on useful information that has an established conclusion about a particular belief, group, or outcome.
For decades, political scientists and communication researchers have applied this dual-motivation framework to study how citizens interact with traditional media, such as newspapers and television news programs. However, in the age of the Internet, the strict editorial norms of news information in traditional media have lost their influence, and the news information disseminated on social media is significantly weaker than traditional media in terms of usefulness and authenticity. As a result, judgments about authenticity are perhaps more important than ever before.
(2) The accuracy-oriented motivation and goal-oriented motivation for the sharing of true and false political news on social media
Accuracy-oriented motivation values the authenticity of political news information. People may be worried that accidentally sharing fake news will damage their reputation. In a way, this reputational concern influences accuracy-motivated social media behavior. It is not clear that sharing fake news is a rare phenomenon compared to sharing real news. At the same time, it is difficult to successfully distinguish between true and false news. First of all, fake news is specifically designed to look real; Second, fast-paced social media platforms deplete attention and cognitive resources (Pennycook and Rand, 2019b).
Goal-oriented motivation values the usefulness of information and has a key impact on the sharing of political news on social media, especially during the outbreak of political conflict. In recent years, political conflicts in the United States have been exacerbated by increasing political polarization (Mason, 2018). People tend to discredit political opponents or mobilize politically like-minded peers by sharing political news that is in the interests of their own party. In this regard, fake news reporting can be particularly useful in its ability to portray political opponents in an excessively negative manner without being constrained by reality (Acerbi, 2019). So, to some extent, goal-oriented motivation dominates social media users' interactions with news stories. People will share useful news that aligns with their political goals, and they won't even pay attention to the veracity of the information at all.
(3) The role of individual differences in news sharing
Accuracy-motivated ignorance theories argue that people end up sharing false information because they lack the cognitive reflective ability to distinguish between true and false information. Pennycook and Rand (2019b) provide a piece of evidence for the ignorance theory, where they found that people who performed better on cognitive reflection tests were better able to judge whether news headlines were true or false. Another piece of evidence comes from Nagler and Tucker (2019), who show that older adults are more likely to share fake news on Facebook. On this basis, we test the theory of ignorance by focusing on four individual-level predictors: age, cognitive reflection, realpolitik knowledge, and digital application literacy.
Goal-driven polarization theories focus on partisanship as a key driver of fake news sharing. In general, partisans are more willing to share information from ideologically close sources (barber et al. 2015) and are more skeptical of articles shared by political opponents (Lee, Kim, and Coe, 2018). Guess, Nagler, and Tucker (2019) found that partisanship also affects the sharing of fake news, as conservatives and Republicans were more likely than liberals and Democrats to share fake news on Facebook in 2016. To test the polarization theory, we examined whether Democratic supporters were willing to share political fake news from pro-Democratic sources, and whether the partisan asymmetry found by Guess, Nagler, and Tucker (2019) was widespread. At the same time, this study also combines the literature on political party emotional polarization to explore whether love within the party or hatred outside the party leads to party polarization.
Finally, we explore the goal-motivated destruction theory. In this regard, the media and psychology have been focusing on so-called "trolls" who take pleasure in misleading and harassing others online (Buckels, Trapnell, and Paulhus 2014). At the same time, political cynicism may also drive the sharing of fake news. For example, a recent study showed that Americans and Danes who are dissatisfied with political elites and the political system are more likely to spread conspiracy theories online (Petersen, Osmundsen, and Arceneaux 2018). Here, we carry out behavioral tests on this theory, including a measure of the psychology of provocative behavior and a measure of political cynicism.
(4) The role of content in news sharing
To elucidate the psychological motivations behind fake news sharing, we also need to understand what people share in the news. Specifically, this paper distinguishes between two aspects of news coverage: (1) whether the news report comes from a fake news source or a real news source (i.e., a generally credible source); and (2) whether the story is coming from a pro-Republican or pro-Democratic source.
The theory of ignorance argues that the dimension of authenticity is the most important, and that people with high cognitive abilities are less likely to share fake news and more likely to share real news, regardless of the political leaning of the news. The latter two theories make opposite predictions: they assume that as long as news is politically useful, people will not value its content as much as it is true or false. Specifically, the sabotage theory implies that individuals should selectively share as much fake news as possible in order to create confusion and uncertainty. Polarization theory argues that sharing fake news is a common partisan behavior. The media landscape in the United States is politically fragmented, with some news sources catering to Republican tastes while others provide content that appeals to Democrats' tastes (Groeling, 2013).
Data and methodology
(1) Data collection methods and the composition of empirical samples
We commissioned YouGov to recruit U.S. Twitter users to complete a 20-minute online survey between mid-December 2018 and mid-January 2019. The survey questions measure the indicators involved in ignorance theory, destruction theory, and polarization theory.
YouGov invited more than 27,000 U.S. Twitter users from their permanent web group to participate in our study, with a total of 8,741 participants accepting the invitation. The majority of these respondents (5,725) did not share their Twitter IDs at the end of the survey. Of the 2,976 respondents who shared their Twitter IDs, another 639 participants provided IDs that we could not access through Twitter's API. Therefore, our final sample size is N=2337. The demographic characteristics of participants in this study were similar to those of participants in other high-quality Twitter studies.
Based on Twitter data from 2,337 participants, as of mid-January 2019, we pulled a total of 2,709,052 tweets and retweets from them. The median number of (repeated) tweets from our participants was 488, and the vast majority of tweets (about 70%) were posted after the 2016 U.S. presidential election.
(2) The difference between fake news and real news
Because we can't tell if a specific story people share is true or false, our measure of fake news is at the publisher level, not the reporting level.
To determine the "sharing of fake news sources", we cross-checked the Twitter network links posted by all participants with a list of 608 fake news sources constructed by journalists and academics to determine whether the news was true or false. The creators of the list coded the most popular fake news sources based on partisan leanings, with 12 of them being pro-Democratic and 30 being pro-Republican fake news sources. To judge "sharing from real news sources," we cross-referenced participants' tweets with a list of real news publishers obtained from the AllSides organization. AllSides aims to help citizens get the information published in online media right by providing a "balanced news and civic discourse".
To further test the relevance of content to sharing behavior, we also constructed two unique datasets of news headlines. We focus on headlines because headlines are more accessible than the full text, and previous research has shown that just seeing headlines can have a big impact on the decision to share news on social media (Effron and Raj, 2019). We performed two types of automated analysis of these headlines and applied the sentimentR package to measure the sentiment reflected by the headlines.
(3) Survey methodology: predictors of fake news sharing
All participants answered the same survey questions that were designed to test the potential predictors of fake news sharing from the three theories discussed earlier: ignorance theory, destruction theory, and polarization theory.
In the theory of ignorance, we use four methods to test it. The first is a cognitive reflex test, which aims to measure people's tendency to overturn false intuitive responses after reflection. This is followed by the age of the participants. Finally, there is the realpolitik knowledge test and the digital media literacy measurement. In the sabotage theory, we refer to the "cyberbullying" literature and build a scale with four provocation indicators (Buckels, Trapnell, and Paulhus, 2014), such as asking participants if they agree that "I like to provoke people in forums or website comment sections". We measured political cynicism through eight survey tools (Dekker and Meijerink, 2012), asking participants whether they agreed that "American politics is pathological." In polarization theory, we include two measures of political polarization: partisanship and feelings toward Republicans and Democrats.
Fourth, empirical results
(1) Will people share news from fake news sources?
Through a descriptive analysis of the data, we found that news sharing was only a small fraction of the Twitter activity of our sample population. Of the 2.7 million tweets in our data, only 85,344 (≈3%) contained links to national news sites, whether fake or real. Of the Twitter news links, 3,269 (≈4%) were from websites with a history of publishing fake news. The vast majority of these stories (2,563) came from pro-Republican fake news publishers. In contrast, real news coverage from pro-Democratic sources (60%) is more than double that from pro-Republican sources (23%). Taken together, these two observations suggest that the degree of exposure to fake news depends largely on the ideological leanings of one's Twitter network. In pro-Democratic networks, only a small percentage of news may come from fake news sources, while pro-Republican networks may show a higher ratio of fake news to real news.
While these totals suggest that fake news sharing may be widespread in some social networks, they also mask a huge heterogeneity in sharing behavior. The panel on the left of Figure 1 below clearly shows that the people who share fake news sources are mainly concentrated in a small group of people: 1% of the group members are responsible for sharing about 75% of the fake news publisher links, and 11% are responsible for sharing all fake news sources. At the same time, the top 1% of real news "super sharers" are only responsible for sharing about 30% of the real news sources.
Figure 1 Empirical cumulative distribution function showing the distribution of true and fake news sharing
(2) What is the motivation for sharing fake news and real news?
Figure 2 gives the mean marginal effect of the logistic regression model, which estimates the correlation between our primary predictors and the probability of sharing at least one news story from fake news or real news sources. We estimated the models separately for each of the primary predictors, each of which controlled for gender, income, education, ethnicity, and political interest.
The panel at the top of Figure 2 examines the theory of ignorance. We found in the top left corner that older panelists were more likely to share content from pro-Republican and pro-Democratic fake news sources. From the youngest (18-22 years old) to the oldest (83 years old) age group, the probability of sharing at least one story from a pro-Republican fake news source increased by 15 percentage points, and the probability of sharing a pro-Democratic fake news source increased by 8 percentage points. However, older people seem to share all kinds of news casually on social media, and there is no preference for fake news sharing. In addition, participants with the lowest digital literacy are the least likely to share content from fake news sources; Lower realpolitik knowledge also corresponds to less political fake news sharing. Together, these findings do not support the claim that people fall victim to fake news because they can't tell the real from the fake. Conversely, fake news sharers are more politically savvy and often use the internet for political activities.
We found little evidence that sharing fake news is an act of "trolling" to vent social grievances (Buckels, Trapnell, and Paulhus, 2014). There is plenty of evidence that fake news sharing is more of a reflection of disruptive sentiment in the political sphere. From the least politically cynical participants to the most cynical ones, the likelihood of sharing pro-Democrat and pro-Republican fake news sources increased by about 6-10 percentage points. However, political cynics are also more likely to share content from authentic news sources, regardless of the political bias of the news source.
The panel below Figure 2 examines the polarization theory, which argues that partisan motivation is a central factor in understanding the sharing of fake news. We found that the likelihood of sharing fake news sources increased significantly under a distinct party identity. Nowhere is this correlation strongest in Republican partisanship. Next, we find support for the predictor of "negative partisanship" (Abramowitz and Webster, 2018). The negative extra-party influence coefficient is roughly twice that of a positive intra-party influence coefficient, suggesting that hostility to political opponents rather than positive sentiment towards the party drives the sharing of fake news sources. Once again, we observe partisan asymmetry: this association is strongest among individuals who don't like Democrats and share content from pro-Republican fake news sources.
As shown in the lower right corner of Figure 2, the partisan motivations behind the sharing of fake news and real news are highly overlapping. One difference compared to sharing fake news sources is that our estimates gain greater precision because more participants share real news sources. Another difference has to do with partisan asymmetry: when it comes to truthful news sharing, we find no asymmetry between Democrats and Republicans; Both men are equally partisan, and negative feelings towards their political opponents are also strong motivations.
Overall, these results fit well with the polarization theory's predictions: partisans share news that is biased toward their own party, mainly because they are hostile to people outside the party. In addition, the ignorance theory test further suggests that these sharers are more concerned about the political usefulness of the news than the quality of the information.
Figure 2 Predictor diagram of news sharing
(3) The role of news content: Fake news is at the extreme of the political party news sequence
1. Source sharing patterns on different types of party news
If partisan motivations dictate the sharing of fake news and real news, social media users should share news sources with similar political leanings, especially partisan news that is at the extreme end of the party news sequence. To test this, Figure 3 plots a heatmap of news-sharing relevance based on seven different newsfeed types. From the chart, it can be found that the people who share the most articles from real news sources with ideological bias are also the most likely to share fake news sources with similar political biases. Conversely, non-diagonal correlations suggest a reluctance to share sources from politically hostile parties. Moreover, the graph also reveals partisan asymmetry: sharing real pro-Republican news is highly correlated with sharing pro-Republican fake news, and sharing pro-Democratic fake news is much less correlated with real news. Conversely, participants who share pro-Democratic news often share centrist news as well. So, while social media users generally favor partisan news sources, the influence is most pronounced in the Republican camp.
Figure 3 Heatmap of shared news relevance for seven newsfeed types
2. The predictive power of partisanship in the political party news sequence
If there are systemic differences in the extent to which news sources cater to partisan motivations in their content, the link between partisanship and news-sharing actions should be stronger for news sources that are at the extreme end of the source sequence. Figure 4 verifies this judgment. The graph on the left estimates the logistic regression coefficients for the models that examine the association between seven partisanship and shared news content from seven types of news sources, while the graph on the right gives the raw count of shared links accumulated by participant affiliation.
Overall, Figure 4 provides more evidence of partisan factors in favor of news-sharing: both panels show that Republicans are more likely than Democrats to share ultra-pro-Republican news sources and less likely to share extremely pro-Democratic news sources, and vice versa. The panel on the right further shows that Republicans and Democrats never share fake news sources that are inconsistent with their own political positions.
Finally, Figure 4 also highlights the partisan asymmetry between pro-Democratic and pro-Republican fake news sharing. Democratic respondents are more likely than Republicans to share real news about centrists, pro-Democrats, and pro-Democrats, as well as pro-Democratic fake news. At the same time, partisan asymmetry is broadly similar in pro-Republican news sources, but more pronounced in pro-GOP fake news.
Fig. 4 Relationship between news sharing and party affiliation on seven types of news sources
3. The negative impact of different news content on the news sequence of political parties
Finally, we turn to sentiment analysis of news headlines shared by panelists on Twitter, as well as the average headlines on the first page of news sources that panelists most often share. The theory of polarization implies that news shared from Democratic news sources should attack Republican elites. Coverage by Republican news sources should have a negative impact on the Democratic elite. To test this, Figure 5 shows the negative impact of news headlines from different sources in the left panel, provided that the headlines of that source refer to Republican or Democratic politicians. The panel on the right shows the number of story headlines in each source type, based on mentions. The empirical results show that pro-Democratic news sources have a negative attitude towards the Republican Party, while pro-Republican news sources have a negative attitude towards the Democratic Party. For partisans, the most politically useful news sources when promoting news sources on social media are news that has a negative attitude outside the party and a positive attitude towards the party. As a result, Democrats are more likely to share centrist news sources, while Republicans prefer fake news sources. For the purposes of Figure 5, first, as with the headlines of genuine pro-Democratic sources, centrist headlines have a significantly greater negative impact on Republicans than on Democrats. Second, genuine pro-Republican headlines have a similar negative impact on Republican and Democratic elites. Third, and most importantly, only pro-Republican fake news sources consistently publish negative portrayals of Democrats. This explains why Republicans prefer political fake news over Democrats. We believe that in order to get a stable source of news that matches their political tastes, Republicans must turn to more extreme news sources, including those known for publishing disinformation. Democrats, by contrast, will have to look for genuine news sources that are centrist and pro-Democrat to achieve their partisan goals.
Fig. 5 Negative impact scale of news story headlines from seven different types of news sources
V. Discussion and conclusions
We tested two psychological motivations that may influence sharing political fake news: accuracy-oriented motivation and goal-oriented motivation. The accuracy motivation means that people are more willing to share the truth, but sometimes they can't recognize fake news due to lack of concentration or personal numerical qualities. Like others, we found that older people are more likely than younger people to share fake news, an effect that is often blamed on individual ignorance. However, after examining the theoretical measures of ignorance (cognitive reflection, factual political knowledge, and political digital literacy), we did not find that fake news sharing reflected a lack of ability to distinguish between true and false information.
In contrast, we found strong support for goal-oriented motivational engagement. While destructive targets play a role in the sharing of fake news, it is primarily partisan goals that influence the sharing of fake news (and the sharing of real news). Moreover, people share fake news more out of hostility to their political opponents than positive evaluations of their own parties.
This suggests that fake news is not categorically different from other sources of political information from a partisan motivation perspective. As an analysis of news content shows, partisans' decisions to share fake news and real news sources depend on how politically useful the news is to disparage people outside the party. There are differences in the degree to which different news sources meet partisan goals. Therefore, under partisan thinking, most people do not prioritize whether the news is true or false when deciding to share news on fake news sites. At the same time, the results show the dual effects of accuracy-oriented and goal-oriented motivation. First, fake news sharing is a relatively rare phenomenon that only prevails among a small percentage of the population. Second, Republicans are more likely than Democrats to share fake news.
Overall, these conclusions have important implications for how to successfully curb the spread of "fake news". If the main concern is whether a news story can attack a political opponent, then fact-checking of news will not actually stop the proliferation of fake news. Our findings suggest that the sharing of political fake news is related to the growing phenomenon of political party polarization in the United States, and it is difficult to improve the information dissemination environment in which political fake news is rampant without addressing the larger problem of political polarization.
本文转自 | Political理论志