This study examined how interactions with ChatGPT about flu vaccination and climate change influenced users’ beliefs and attitudes.

Following Entman’s observation that policy frames define social problems, diagnose causes and suggest remedies, we examined the strategies that 12 U.S. governors (from states matched according to population size and density, demographic composition, per capita incomes, geographic proximity, and COVID-19 incidence) used to frame COVID-19 policy agendas. After scraping the governors’ statements about COVID-19 from press releases issued from January 2020 to May 2023 (N = 14,629), we leveraged ChatGPT (GPT) to identify and assess the intensity of public health, economic stability, and civic vitality frames. Subsequent analysis explored differences in the framing strategies according to the governors’ political party and gender. In the process, this study underscores the importance of AI prompt engineering to realize GPT’s transformative potential to facilitate communication research by efficiently identifying and assessing the content of policy frames.

This study investigates the role of visual framing in shaping public perceptions of the Black Lives Matter (BLM) protests through an online experiment, informed by news value and protest paradigm theory. It specifically examines the effects of conflict versus solidarity visual frames on audience reactions when paired with textual framing of protest goals. Results indicate that solidarity visuals significantly decrease negative perceptions about protesters compared to conflict visuals, increase positive affect, and boost engagement intentions. Furthermore, the interaction between visual and textual framing demonstrates that "defund" and "reform" messages paired with solidarity visuals effectively reduce negative perceptions. These findings underscore the impact of visual elements in news coverage, highlighting the need for ethical considerations in visual journalism and suggesting that such framing can significantly influence audience attitudes and engagement with social issues.

To examine the effects of numerical evidence and message framing in communicating vaccine efficacy information about infectious diseases, an online experiment presented to U.S. adults different versions of a vaccination promotional message that vary by numerical vaccine efficacy evidence: (low efficacy rate: 60% vs. high efficacy rate: 95%), outcome framing (preventing disease-related infection vs. preventing disease-related severe illness), and gain vs. loss framing, using a factorial between-subjects design. While there was no significant interaction between numerical vaccine efficacy evidence and message framing, findings showed that a higher vaccine efficacy rate increased positive beliefs about vaccination and outcome framing emphasizing infection prevention increased message processing fluency. Given that infectious diseases pose higher risks for severe illness among older adults, follow-up analyses by age showed that only younger adults were sensitive to message framing where outcome framing emphasizing infection prevention increased processing fluency.

Alternative messages that present logically equivalent information, often referred to as equivalence frames, have been shown to influence readers' opinions on public issues. While equivalence framing has been studied in the context of issue advocacy, exhibiting pervasive effects across domains of decision-making, little attention has been paid to whether the general public is able to choose these equivalence frames based on the goal of persuasion. Given that framing effects have important implications on democratic decision-making, this paper reports on experiments that manipulate the strategic goal of policy advocacy (i.e., supporting alternative policy proposals) and ask respondents to select between equivalence frames to enhance the persuasive power of the advocacy toward the specified goal. Findings across three issue topics suggest that for the general adult population, only a small proportion of people were able to select equivalence frames based on the goal of persuasion with most people failing to do so. Also, a follow-up study with a university student sample showed that familiarity with one equivalence frame over the other was a more consistent predictor of equivalence frame use than the goal of advocacy in communicating policy issues.

Metaphorical language describing the COVID-19 pandemic as a war has been pervasive in public discourse (e.g. "the pandemic is a war," "the virus is an enemy," and "the vaccine is a weapon"). This study employs an online survey experiment (N = 551 U.S. adults) to examine the impact of war metaphors compared to non-metaphorical literal frames and fire metaphors (e.g. "the pandemic is a wildfire"). War metaphors exhibited little advantage over literal frames across a variety of desirable outcomes (i.e. the adoption of pro-health behaviors against COVID-19, perceived solidarity and collective responsibility to curb the pandemic, and intentions to discuss and share the health news with others). However, this study revealed some benefits of war metaphors over fire metaphors. Compared with fire metaphors, health news featuring war metaphors increased both positive emotions and perceived threats of COVID-19, which in turn promoted pro-health behaviors against COVID-19 and perceived solidarity to cope with the public health crisis. Moreover, positive emotions in response to war metaphors also indirectly encouraged the retransmission of science-based COVID-19 health news. This study thus showcased the benefits and limitations of war metaphors and revealed the mediating roles of perceived threats and positive emotions in explaining war metaphorical framing effects. Implications of using war and fire metaphors for communicating about public health crises are also discussed.

Taiwan’s legalization of same-sex marriage in 2019 made it the first nation in Asia to grant marital rights to gay and lesbian couples. In the years leading up to legalization, the Marriage Equality Coalition Taiwan (pro-same-sex marriage) and the Coalition for the Happiness of Our Next Generation (anti-same-sex marriage) mobilized large-scale social movements on social media between 2016 and 2017 to influence the legislative process. The network structure and affordances of digital platforms have facilitated communication and mobilization for social movements. However, new technology alone does not guarantee participation, and cultural aspects of mobilization on digital platforms are an important area of study. This paper examines the framing strategies these two organizations used on Facebook pages and the political and cultural contexts that facilitated or constrained frame alignment. A mixed-method framing analysis combining quantitative and qualitative methods of their Facebook posts revealed that the supporting group framed same-sex marriage as an issue of human rights and as a democratic development linked to Taiwan’s goal of national independence, whereas the opposing group framed it as a destruction of traditional culture concerning family values and social order. Our analysis identified the distinct features of framing strategy in Taiwan’s marriage equality movement and countermovement, including the appeal to nationalism and the downplaying of religion, that were affected by Taiwan’s specific political and cultural contexts.

Response efficacy information indicating the effectiveness of a recommended behavior in risk reduction is an important component of health communication. For example, many messages regarding COVID-19 vaccines featured numerical vaccine efficacy rates in preventing infections, hospitalizations, and deaths. While the relationship between disease risk perceptions and fear has been well established, we know less about the psychological factors involved in communicating vaccine efficacy information, such as response efficacy perceptions and hope. This study examines the effects of numerical vaccine efficacy information and message framing on vaccination intentions and their relationship to perceived response efficacy and hope, using a fictitious infectious disease similar to COVID-19. Findings suggest that communicating a high efficacy rate of the vaccine in preventing severe illness increased perceived response efficacy, which in turn boosted vaccination intention directly and indirectly through increasing hope. Also, fear about the virus was positively associated with hope about the vaccine. Implications of using response efficacy information and hope appeals in health communication and vaccination promotion are discussed.

The murder of George Floyd by a Minneapolis police officer on 25 May 2020, sparked widespread protests led by the Black Lives Matter movement throughout the summer of 2020. Subsequent news coverage of these protests prominently featured acts of civil disobedience even though almost all protests were peaceful. In turn, protest “violence” was picked up by conservative political elites as evidence to promote legislation to control protests and keep communities safe. Since summer 2020, eight states have passed such legislation with additional bills pending in 21 states, raising concerns that the legislation suppresses political expression. This paper brings together literature on free expression, the protest paradigm, and news framing to provide the basis for a quantitative and qualitative analysis of 379 news stories and editorials covering Florida’s HB1 protest legislation. Results reveal that the most frequent news frame was fighting crime, with relatively less attention to free expression, political strategy, and race frames. In addition, very little attention was paid to the legislation’s potential chilling effects suppressing constitutionally protected speech and assembly. These results indicate news media were deficient in providing the public with a sufficient assessment of the implications of protest legislation.

In a series of essays, scholars respond to Perloff and Shen's article, "The Third-Person Effect 40 Years After Davison Penned It." They offer further thoughts on how to measure the phenomenon, where future research is headed, and even whether the effect is real.

With increasing evidence on deepening cleavages along geographic lines, we argue that the local political climate plays an important role in political decision-making and engagement. In this study, we aim to understand the role of political contexts in shaping different forms of political participation, whether centered in the local community or in digital spaces. We specifically consider two important contextual factors that potentially relate to participation: the partisan composition of the neighborhood environment and the nature of political representation at the state government level. We introduce two sets of competing arguments: Mobilization and Resignation vs. Activation and Complacency to explain different participatory mechanisms. Using both national survey data collected during the 2016 U.S. election period and zip code and state-level contextual data, we employ three-level multilevel modeling to tease out how multiple factors operating at different levels are related to online or public forms of participation. In general, our findings reveal that individuals living in a state with political underrepresentation are more likely to engage in public forms of actions. Additionally, we examine subgroup analyses to show how contextual relationships with participation are different according to political orientations, such as party identification and political interest.

We examine how individuals' interactions with the shifting contemporary communication ecology-either by seeking information selectively from partisan sources or immersing themselves in a broad range of partisan communications - relate to shifting levels of social trust and online engagement. Using national panel surveys of young adults (i.e., millennials age 18-34) collected over the 2016 U.S. presidential election, we find that individuals' partisan communication flows-calculated by algorithmically combining patterns of news consumption, social media use, and political talk-explain: (a) polarized shifts in levels of trust towards people of other nationalities, religions, races, and ethnicities and (b) increases in levels of online political engagement. By elaborating the relationship between citizens' communication patterns and their levels of trust and participation, this research forces a reconsideration of theoretical traditions in the field of communication, especially those linking mass and interpersonal processes in the study of social capital.

This review introduces a conceptual framework with three elements to highlight the richness of the framing effects literature, while providing structure to address its fragmented nature. Our first element identifies and discusses the Enduring Issues that confront framing effects researchers. Second, we introduce the Semantic Architecture Model (SAM), which builds on the premise that meaning can be framed at different textual units within a text, which can form the basis of frame manipulations in framing effects experiments. Third, we provide an Inventory of Framing Effects Research Components used in framing effects research illustrated with salient examples from the framing effects literature. By offering this conceptual framework, we make the case for revitalizing framing effects research.

The digital media environment has transformed the ways information about “collective preference” is communicated. Using 2 survey experiments, this study examines how embedded context may condition the processing and influence of an opinion poll in a multicue, source-confusion environment. Our results suggest that, in general, opinion polls are evaluated more negatively when the results are embedded in a politician’s tweet. Consistent with motivated reasoning, congruent polls that support one’s side tend to be perceived as more credible, which in turn leads to a more polarized issue position via poll-aligned opinion climate perception. This self-serving perception may be heightened by politician repurposing of polling outcomes, especially in the lack of pollster brand names. Importantly, there is partisan asymmetry in how contextual information may alter the processing of polling results. Above and beyond an average effect, politician uptake of polling data undermines a poll’s perceived credibility to a greater extent among Democrats than Republicans.

We examine how individuals’ interactions with the shifting contemporary communication ecology—either by seeking information selectively from partisan sources or immersing themselves in a broad range of partisan communications — relate to shifting levels of social trust and online engagement. Using national panel surveys of young adults (i.e., millennials age 18–34) collected over the 2016 U.S. presidential election, we find that individuals’ partisan communication flows—calculated by algorithmically combining patterns of news consumption, social media use, and political talk—explain: (a) polarized shifts in levels of trust towards people of other nationalities, religions, races, and ethnicities and (b) increases in levels of online political engagement. By elaborating the relationship between citizens’ communication patterns and their levels of trust and participation, this research forces a reconsideration of theoretical traditions in the field of communication, especially those linking mass and interpersonal processes in the study of social capital.

Many major news websites have recently opted to remove comment sections that appear beneath their online news articles. However, researchers know very little about how news audiences feel about the silencing of this interactive feature. Our study analyzes data from adult Internet users in the United States in an online survey to provide empirical evidence regarding motivations underlying different engagement in news comment systems and attitudes of news readers toward comment system removal. Overall, findings suggest that compared to non-users, people who read or post comments are more likely to oppose removal. Moreover, comments removal attitude is dependent on motives of using news comment sections. Information-seeking motives are negatively related to the support for comments removal among lurkers, whereas affective socialization motives are positive predictors of comment system removal among commenters.

While music as an artistic form is well studied, the individuals behind the art receive relatively less attention. In this article, we provide evidence of celebrity advocacy with a systematic examination of musicians’ political engagement on Twitter. This study estimates the extent to which musicians use Twitter for political purposes, with particular attention to whether such engagement varies across music genres. Through a computational-assisted analysis of 2,286,434 tweets, we group 881 musicians into three categories of political engagement on Twitter: not engaged (comprising the majority of artists), circumstantial engagement, and active political engagement. We examine the latter categories in detail with two qualitative case studies. The findings indicate that musicians from different genres have distinct patterns of political engagement. The Christian music genre shows the most engagement as a whole, especially in philanthropy. On the contrary, the most active accounts are rock and hip-hop artists, some of whom discuss political issues and call for mobilization. We conclude with suggestions for future research.

Statistical information permeates media messages, but little is known about how the use of different presentation formats influences message processing. Thus, we explore numerical framing effects by examining how presentation formats interact with gain/loss frames to alter message processing and issue perceptions. We found that logically equivalent information embedded in gain/loss frames generated different levels of comprehension when it was presented in a frequency format. The gap, however, disappeared when it was displayed in a percentage format. Different comprehension levels then shaped negative emotions differently, which in turn affected perceived issue seriousness. Mediational analyses tentatively suggest that numerical framing occurred through cognitive and emotional responses. The implications are discussed.

Message frames have been found to influence relevant issue attitudes by influencing the weight of issue considerations emphasized in the message. As such message frames often originate from advocacy interest groups, this study investigates differences in the framing effects of advocacy groups, depending on whether the message fits readers' expectations for the communicators' issue position (expected advocacy) or not (unexpected advocacy). Across two issue topics, findings suggest that unexpected advocacy significantly influenced readers' perceived belief importance, which in turn influenced issue attitudes, whereas the same mediated path of framing effects was not supported in the case of expected advocacy.

This review synthesizes the existing literature on cognitive media effects, including agenda setting, framing, and priming, in order to identify their similarities, differences, and inherent commonalities. Based on this review, we argue that the theory and research on each of these cognitive effects share a common view that media affect audience members by influencing the relative importance of considerations used to make subsequent judgments (including their answers to post-exposure survey questions). In reviewing this literature, we note that one important factor is often ignored, the extent to which a consideration featured in the message is deemed usable for a given subsequent judgment, a factor called judged usability, which may be an important mediator of cognitive media effects like agenda setting, framing, and priming. Emphasizing judged usability leads to the revelation that media coverage may not just elevate a particular consideration, but may also actively suppress a consideration, rendering it less usable for subsequent judgments. Thus, it opens a new avenue for cognitive effects research. In the interest of integrating these strands of cognitive effects research, we propose the Judged Usability Model as a revision of past cognitive models.