Sude & Dvir-Gvirsman (2023) in Journal of Computer Mediated Communication
Two large panel surveys examined Americans’ perceptions of both the prevelance of uncivil behaviors on different platforms and key platform features.
Specifically, every social media platform that a participant reported using regularly was examined, meaning that perceptions that are due to the platform could be disentangled from perceptions due to the user.
First, perceptions were not stable – idiosyncratic user experiences made more of a difference than any “objective” differences between platforms.
Second, network association – which measures two key building blocks of community – was positively associated with perceptions of incivilility. This might suprise you. Platforms characterized by network association look more like offline communities: Your connections to other people are visible; your connections can easily meet and interact with one another. Your might expect that under these conditions, people would be on their best bahvior. Not so. In fact, people anticipate more uncivil behaviors on these platforms.
An additional major contribution of this work is that it offers an intuitive definition of incivility: Uncivil acts (are perceived to) expresss a commitment to conflict. Civil acts imply that sitting down to talk is still a worthwhile conflict management strategy.
https://doi.org/10.1093/jcmc/zmac035
Cite (APA style) as:
Sude. D. J., & Dvir Gvirsman, S. (2022). Different platforms, different uses: Testing the effect of platforms and individual differences on perception of incivility and self-reported uncivil behavior. Journal of Computer Mediated Communication. Advance online publication. https://doi.org/10.1093/jcmc/zmac035
Sude & Knobloch-Westerwick (forthcoming) in International Journal of Communication.
This study, examining impacts of voluntary exposure to political content harvested from a complete corpus of Reddit posts, demonstrated that the less sizeable we percieve that our partisan group is relative to the partisan out-group, the more warmth we report. Why? Integrating Noelle-Neumann’s Spiral of Silence Theory and Brewer’s (1999) extension of Social Identity Theory it is argued that feeling locked in intractable social conflict with an implaccable foe is in fact quite unpleasant. People are motivated to see the other side as reasonable. Unfortunately, media presentation and online encounters often frustrate this motivation, potentially leading to even more intense hostility.
Tags: Social identity, exemplification
Book chapter (open access): Sude & Knobloch-Westerwick (2022) in Strömbäck et al.’s volume on Knowledge Resistance in High Choice Information Environments. This chapter reviews theory and cutting edge researching demonstrating that social identities and social goals, as much as prior attitudes, shape the political content that we seek out and our reactions to it.
Westerwick, Sude, Brooks, Kaplan, & Knobloch-Westerwick (2022) in Mass Communication and Society.
Even non-political social identies can impact the time spent on political content. Participants whose gender identity was important to them spent more time reading work by same-gender authors, even if the content of this work challenged their political beliefs.
Further, participants who already felt that society respected their gender also showed this effect. As with many media effects studies, people may turn to media to maintain their existing positive attitudes. Representation matters. It communicates societal respect and acceptance. However, it likely takes more than a single website to convince people that society thinks that they matter.
Importantly, when it comes to political media, it’s not just our political party identities that matter.
Tag: Social identity
Budak, Garrett, & Sude (2021) in Communication Methods and Measures
Crowdcoding can be an effective way to generate the human-coded content that can be used for human-supervised machine learning. However, it can be very difficult to get people to reliably code “latent” content – content that requires human interpretation. In training MTurk Workers to code civil vs. uncivil content scraped from Reddit, we found that, rather than provide a detailed codedbook, the best way was simply to test them, give them feedback, and test them again. In general, training was, perhaps suprisingly, a lower cost and more effective method.
ICA 2021 Presentation: Sude, Westerwick, & Knobloch-Westerwick (Video)
Tag: Social identity
Sude, Pearson, & Knobloch-Westerwick (2021) in Computers in Human Behavior
The ability to vote (up or down) on an article shaped users’ own attitudes. Attitude-consistent voting dominated, independently of whether content was actually read. Thus, attitudes were reinforced. Self-expression interactivity may act as a megaphone, allowing people to focus solely on themselves and their ideas.
While people are thought to be motivated to maintain or defend their existing attitudes, Sude and collegues argue here that defending your views through a confirmation bias at least fosters greater sophistication, in the form of familiarity with arguments and evidence in favor of your position. Defending views through the mere act of up- or down-voting allows users to feel more extreme without mustering a better defense of their ideas.
For applications to user experience, and potential ways to harness self-expression for the public good, click.
For media coverage, see Press.
Tag: Self-expression
Westerwick, Sude, Robinson, and Knobloch-Westerwick (2020) (links to official citation, abstract, and download options) in Mass Communication and Society
People demonstrated distinct confirmation bias patterns when browsing content on a simulated social media platform versus a site with pieces by professional journalists. Note, this study employed an experimental design that randomly assigned identical content to the two different website sources. When posts were by peers, and people felt that the topic was personally important, they showed a particularly strong confirmation bias.
Potential explanations for impacts on the confirmation bias involve the greater credibility of professional sources {lower credibility of peer sources} as well as the potential social reward derived from reading work by like-minded peers on personally important topics.
Last, as in Sude, Knobloch-Westerwick, Robinson, and Westerwick (2019) described below, time spent reading content impacted perceptions of national public opinion in line with the content’s stance, and in turn impacted participants’ own attitudes. Once again, people “tuned” their attitudes to fit in with a national public, which is remarkable in and of itself!
Tag: Exemplification
Garrett, Sude, and Riva (2020) (links to official citation, abstract, and download options) in Political Communication
Social ostracism led people to reject political fact-check messages criticizing commonly held beliefs among their partisan-groups, even when they believed their responses would be private. Rooted in the social ostracism and motivated reasoning literatures, this study gives evidence that a desire for social-affiliation can lead to “toeing the party lie.” Ironically, fact-check messages may in fact promote conformity to erroneous but common beliefs within one’s political party, at least when participants are experiencing social threat.
Tag: Social identity
Sude, Knobloch-Westerwick, Robinson, and Westerwick (2019) (links to official citation, abstract, and download options) in Communication Monographs
People moderated their political attitudes in response to headlines alone; more so if they read the articles. This attitude change happened in part because the articles acted as a national public opinion cue and people “tuned” to perceived trends in national public opinion. In line with predictions derived from Spiral of Silence Theory, people appeared to want to hold less extreme views so that they could better navigate the social threat of political disagreement and achieve the rewards of social integration into the broader American public.
Tag: Exemplification
Knobloch-Westerwick, S., Westerwick, A., & Sude, D.J. (2019). Media choice and selective exposure. In M.B Oliver, A. Raney, & J. Bryant (Eds.), Media Effects (4th Ed.). London, UK: Routledge.
This chapter reviews the concept of selective exposure (selection of and engagement with media messages). It reviews different eras in selective exposure research and their associated methodological and theoretical approaches. It ends with cutting edge developments and future directions.
Tag: Selective Exposure
Broad Overview of Program of Research
Daniel Sude, Ph.D. 