Algorithmic personalization of source cues in the filter bubble: Self-esteem and self-construal impact information exposure

How do users pick out online information sources? Building on a self-regulation perspective to media use, this study investigates routes to self-enhancement (i.e. state self-esteem [SSE] increase) through selective exposure to sources of political online information. Personal-self importance (PSI) and social-self importance (SSI) were conceptualized as moderators of self-enhancement. An experiment mimicked the filter bubble, as participants ( n = 88) browsed only attitude-aligned political content. The experiment varied source cues, with two (of eight) bylines displaying individual participants’ name initials as author initials. The selective exposure time participants spent on messages from same-initials authors was logged to capture egotism (based on the well-established name-letter effect). Pre-exposure SSE influenced self-enhancement, contingent upon both PSI and SSI. Perceived source similarity affected post-exposure SSE, contingent upon the same moderators. The findings show that algorithms can personalize source cues to attract users and impact self-esteem.

Debunking Misinformation About Genetically Modified Food Safety on Social Media: Can Heuristic Cues Mitigate Biased Assimilation?

Focusing on debunking misinformation about genetically modified (GM) food safety in a social media context, this study examines whether source cues and social endorsement cues interact with individuals’ preexisting beliefs about GM food safety in influencing misinformation correction effectiveness. Using an experimental design, this study finds that providing corrective messages can effectively counteract the influence of misinformation, especially when the message is from an expert source and receives high social endorsements. Participants evaluate misinformation and corrective messages in a biased way that confirms their preexisting beliefs about GM food safety. However, their initial misperceptions can be reduced when receiving corrective messages.

The Effects of Source Cues and Issue Frames During COVID-19

The health and economic outcomes of the COVID-19 pandemic will in part be determined by how effectively experts can communicate information to the public and the degree to which people follow expert recommendation. Using a survey experiment conducted in May 2020 with almost 5,000 respondents, this paper examines the effect of source cues and message frames on perceptions of information credibility in the context of COVID-19. Each health recommendation was framed by expert or nonexpert sources, was fact- or experience-based, and suggested potential gain or loss to test if either the source cue or framing of issues affected responses to the pandemic. We find no evidence that either source cue or message framing influence people’s responses – instead, respondents’ ideological predispositions, media consumption, and age explain much of the variation in survey responses, suggesting that public health messaging may face challenges from growing ideological cleavages in American politics.

Distrust, danger, and confidence: A content analysis of the Australian Vaccination-Risks Network Blog

The Australian Vaccination-risks Network is Australia’s most active counter-vaccine lobby group. This study employs a content analysis of the organization’s 2012–2019 blog posts, while further considering Australian-specific vaccine contexts. The goal is to identify the persuasion attributes of these counter-vaccine articles, and the ways that the group’s media employs persuasive cues when communicating to Australian publics. The project gauges the occurrence rates of message variables associated with the Elaboration Likelihood Model of persuasion, including those labeled as the Scarcity Principle, Arousal of Fear, Asking Questions, Source Cues, the Contrast Principle and Negativity Effect, as well as Statistics and Technical Jargon. Three overarching themes collectively exhibited by these message variables are further identified and described as Distrust, Danger, and Confidence. In view of these findings, the study then considers how persuasive cue expression in Australian Vaccination-risks Network blog posts corresponds with Australian vaccine hesitancies and the country’s No Jab No Pay/Play policies.

Framing Automatic Voter Registration: Partisanship and Public Understanding of Automatic Voter Registration

Automatic voter registration (AVR) is a recent innovation in voter registration in the United States, passed by 18 states plus DC in the last 4 years. AVR has generally escaped partisan polarization about election reform, having passed in Republican and Democratic controlled states. Using a survey experiment in the 2018 Cooperative Congressional Election Study, we investigate effects of source cues about support of AVR from different party elites and from election administrators on the public’s expectations about AVR’s impact on turnout, voter fraud, fairness, and election problems. Our experimental results show an asymmetric partisan effect. When AVR is endorsed by Democratic leaders, Republicans (and independents) expect AVR to reduce the fairness and legitimacy of elections, while Democrats are generally resistant to partisan cues.

Sources on social media: Information context collapse and volume of content as predictors of source blindness

Although social media has become a primary news platform, the effects of social media features on users’ information processing remains under-explored. This study explores how social media design features affect use of sources. A 2 × 2 between-subjects experiment examined effects of “information context collapse” (ICC)—where different content types are presented in the same form and location—and volume of content (VoC). These features were hypothesized to predict inattentive (System 1) processing, which predicts “source blindness”—where users fail to process source cues during news use. A mock social media site was created with participants queried about posts shown on the site. Results find that while VoC has no effect, ICC significantly predicts source blindness mediated by System 1 processing. This suggests collapsed information environments lead to inattentive processing of source information, increasing potential negative outcomes of social media news use. Implications of these findings are discussed.

Picture Power? The Contribution of Visuals and Text to Partisan Selective Exposure

Today’s high-choice media environment allows citizens to select news in line with their political preferences and avoid content counter to their priors. So far, however, selective exposure research has exclusively studied news selection based on textual cues, ignoring the recent proliferation of visual media. This study aimed to identify the contribution of visuals alongside text in selective exposure to pro-attitudinal, counter-attitudinal and balanced content. Using two experiments, we created a social media-style newsfeed with news items comprising matching and non-matching images and headlines about the contested issues of immigration and gun control in the U.S. By comparing selection behavior of participants with opposing prior attitudes on these topics, we pulled apart the contribution of images and headlines to selective exposure. Findings show that headlines play a far greater role in guiding selection, with the influence of images being minimal. The additional influence of partisan source cues is also considered.