Does the Style of Misinformation Condition Its Effect? An Experiment in Brazil

• In Event: The Online and Offline Effect of Misinformation

Abstract

Much empirical evidence suggests that misinformation can change beliefs (Ecker et al., 2022). These belief changes have in turn been hypothesized to affect an array of consequential downstream attitudes, such as beliefs about outgroups or preferences for extreme politics (Tucker et al., 2018). Yet evidence of this causal connection is scarce. While there are many studies addressing the drivers of belief in misinformation (Spohr, 2017; Pennycook et al., 2018; Lazer et al., 2018; Tandoc Jr., 2019; Ecker et al., 2022), scholars have not entirely identified when and why misinformation does affect downstream attitudes and behaviors, and specifically, when and why it is likely to increase polarization between groups. When and why is exposure to outgroup-related misinformation believed? When and why does it in turn lead to intergroup polarization?
We posit that the style of misinformation conditions its effect on intergroup attitudes, and hence that only specific subtypes of misinformation should have downstream effects. Namely, distinguishing between neutral-style misinformation and emotional misinformation, we argue that neutral-style misinformation content should increase belief, since it is designed to be credible, while emotional-style misinformation will increase polarization despite being less likely to be believed.
To test this hypothesis, we field in April 2024 a survey experiment in Brazil. Brazil is arguably a good place to run such an experiment because the level of both misinformation circulating and polarization between petistas and antipetistas is quite high. The experiment prompts participants with a battery of social media pieces in which a (false) rumor about Lula da Silva is presented in either emotional terms (e.g. messages include capital letters and/or an derogatory visual depictions of Lula da Silva) or in a more neutral and journalistic manner (our control condition). We additionally compare these various experimental conditions to a placebo control.
Participants are then asked a series of questions in order to assess the extent to which they believe the claims and also their level of polarization. The former is just measured with a slider where they need to select on a scale from 1 to 4 to what extent they believe the content. The latter is measured in two ways. First, ideological polarization is measured by means of usual spatial distance to Brazilian political parties. Second, affective polarization is assessed using a novel quasi-behavioral measurement asking participants if they are willing to collaborate with someone from the outgroup to make a task. If they accept they are rewarded, but they can decline.
The experiment is to be carried out thrughout this year, hence results are not available yet but we hope to have the data around September. This research holds profound implications for understanding why individuals gravitate towards extreme options in their political choices. Even when individuals may not fully believe the misinformation they encounter, our research suggests that emotions play a significant role in shaping their political behaviors. Leaders and groups advocating extreme positions may strategically leverage emotional misinformation to sway public opinion, pointing to the intricate interplay between belief, emotion, and polarization in contemporary discourse. As we await the unveiling of our results, the potential impact of this evidence cannot be overstated in disentangling the complexities of political decision-making in an era characterized by information overload and manipulation.

Authors

• Rodrigo Fernández Caba

• Simon Chauchard, University Carlos 3 of Madrid

• Fernando Barros de Mello, Carlos III-Juan March Institute of Social Sciences