How a peripheral ideology becomes mainstream: Strategic performance,audience reaction,and news media amplification in the case of QAnon Twitter accounts

Abstract

Social media platforms have been used by various actors to bypass traditional media gatekeepers to share messages, draw attention, and accumulate influence. We study how actors from peripheral groups gain influence on social media and how their social media behaviors evolve over time. Integrating online strategic performance and hybrid media literature, we hypothesize that peripheral groups perform group identities to spur social media audience reaction and news media amplification, to which they further adapt their performance. By analyzing 242 QAnon Twitter accounts using topic modeling and time series modeling, we find that their in-group solidarity and out-group animosity tweets boost retweets, but not followers; increased retweets and followers drive news media amplification largely undertaken by right-wing outlets and motivate future performance of group identity, particularly of out-group animosity. The implications of social media and news media for the growth of peripheral actors and ideologies are discussed.

Keywords

Identity performance news media amplification peripheral groups QAnon social identity social media

Social media platforms constitute a critical space for any actors, ranging from celebrities to politicians, social movement activists, fringe groups, and disinformation agents, to bypass traditional media gatekeepers to share messages, draw attention, and accumulate influence (e.g. Marwick and boyd, 2011; Tufekci, 2013). For example, far-right actors create memes and engage in trolling on social media to attract the attention of journalists and gain widespread public attention (Marwick and Lewis, 2017; Phillips, 2018). Hate groups proselytize their ideologies via Twitter and call for active participation through Facebook (Phadke and Mitra, 2020). The fact that a peripheral group can use social media as a springboard to gain broad societal attention reveals both exciting opportunities and dangerous undercurrents in an open media ecosystem without central gatekeepers (Blumler and Kavanagh, 1999; Chadwick, 2017).

In this article, we explore how actors from peripheral groups use social media to become mainstream—defined as increasing social media reaction and news media attention—while accounting for three interlinked components within the media ecosystem: social media actors, social media audience, and news media. In doing so, we draw on three strands of literature. First, according to the online self-presentation literature, people engage in strategic performance to seek attention and win trust from their networked audience (Djafarova and Trofimenko, 2019; Maddox, 2022; Marwick and Lewis, 2017), a tendency pronounced among fringe groups (Phadke and Mitra, 2020; Xu, 2020). Second, the hybrid media system literature underscores the deep linkage between social media and traditional media (Chadwick, 2017), showing how social media engagement can translate into news media coverage and result in broader amplification of strategic actors on social media (Wells et al., 2016; Zhang et al., 2021). Finally, social media actors can adapt their behaviors based on social media audience reaction and news media amplification, which function as concrete metrics and social feedback guiding strategic performance (boyd, 2015; Brady et al., 2021; Papacharissi, 2012). These mutually reinforcing processes can likely catapult obscure social media actors to popularity.

We situate this framework in the context of QAnon believers on Twitter, examining their identity performance, their retweets and followers, and news media amplification of them. Since it originated in 2017, QAnon has morphed from a wild peripheral ideology with only a few comments on 4chan/pol/ to a phenomenon that pervades most social media platforms and news media outlets, especially during Trump’s presidency (Anwar et al., 2021; Gallagher et al., 2020; Garry et al., 2021; Hannah, 2021). QAnon is undergirded by a general distrust in institutions, reflected in its central tenet that Donald Trump has been fighting a global cabal of elites that controls the government and media; it is enabled by the participatory opportunities offered by social media, where QAnon adherents assemble piecemeal information dropped by the anonymous Q and establish their “alternative facts” (Garry et al., 2021; Marwick and Partin, 2022; Zuckerman, 2019). How QAnon fits and thrives in the larger media ecosystem remains understudied. On one hand, QAnon believers seem insulated because the topics of their conversations on Telegram are more distinct to their group and less responsive to real-world events and news media coverage, compared with far-right communities (Zihiri et al., 2022). On the other hand, QAnon has experienced “normiefication,” traveling from fringe message boards to news media headlines and mainstream social media platforms like YouTube and Reddit (De Zeeuw et al., 2020). A platform central to American politics due to the active participation of American journalists, politicians, and activists (Jungherr, 2014; Jurkowitz and Gottfried, 2022), Twitter turns out to be a prominent network hub of QAnon believers (Anwar et al., 2021; Hanley et al., 2022; Jackson et al., 2021), making it an appropriate platform for the study of their growth.

Using the three-pronged framework of social media actors’ strategic performance, social media audience reaction, and news media amplification, we can gain a deeper understanding of how QAnon believers might have leveraged Twitter and news media to gain popularity. This inquiry contributes to broader conversations in the field of communication, including how social media and news media amplify marginal voices and actors (Freelon et al., 2018; Jackson and Foucault Welles, 2015; Zhang et al., 2021), what performances on social media spark audience attention and engagement (Brady et al., 2017; Fine and Hunt, 2021; Rathje et al., 2021), and how social media actors’ behaviors can be adaptive in order to seek attention, approval, and status (Bail, 2021; Brady et al., 2021; Walsh et al., 2020).

Strategic performance for audience reaction

Presenting selective aspects of oneself is a basic persona-sculpting strategy (Goffman, 1959). Such performativity of the self is a hallmark of modernity as people consciously and reflectively author their life projects (Giddens, 1991). Rather than simply providing a convenient stage for people to perform themselves for self-branding and promotion purposes, social media intensify this performative tendency by building popularity and status into platform architectures (Marwick, 2013; Van Dijck, 2013; Van Dijck & Poell, 2013). Metrics such as the number of shares and the number of followers are synonymous with clout; and such metricized popularity is coveted because it not only is a “status symbol” but also begets future popularity (Marwick, 2013: 77).

Performance of group identity

To stand out from the crowd and gain audience attention and social status, social media actors strive to project an attractive identity (Van Dijck, 2013). Some focus on establishing individual identities by demonstrating their tastes, accomplishments, skills, and expertise (Ellison et al., 2006; Maddox, 2022). Others may choose to project affiliative/group identities by situating themselves within social groups and emphasizing the sense of “us” (Tajfel et al., 1979). This is crucial for the mainstreaming of a peripheral group. A clear group identity affirms the group value and attracts like-minded others, thus creating a conducive space for the group to expand (Xu, 2020). The performance of group identities can also energize social media audiences (Bail, 2021; Rathje et al., 2021), which is essential for directing broad societal attention to the peripheral group (Xu, 2020). Peripheral right-wing groups exhibit a strong tendency to perform group identities: right-wing hate groups juxtapose their in-group superiority and grievances vis-à-vis out-group inferiority and immorality, a strategy to bond with their in-group members and educate the public about the out-groups (Phadke and Mitra, 2020). In addition, messages originating from right-wing networks on social media are highly centralized and coordinated (Haupt et al., 2021; Zhang et al., 2022a), suggesting possible group dynamics.

Notably, QAnon is marked by its participatory fan culture centered around its star—Donald Trump (Anwar et al., 2021; Zuckerman, 2019), its performative efforts to establish group conventions and norms (Zihiri et al., 2022), and its deliberate efforts to hack algorithms to spread ideas across platforms (De Zeeuw et al., 2020; Forberg, 2022). Therefore, QAnon believers are likely to perform their group identity on social media. We focus on their performance of group identity on Twitter, operationalizing it as the performance of (a) in-group solidarity and (b) out-group animosity (Rathje et al., 2021).

According to social identity theory (Tajfel et al., 1979), individuals are more obligated to in-group members (based on kinship, race, nationality, political values, etc.) and expect in-group reciprocity on resources and interactions. Social media afford individuals the opportunity to directly network and communicate with in-group members and build group cohesion (Christensen and Jensen, 2018). For example, hip-hop celebrities project Black identity to build solidarity and mobilize collective action to combat racial injustice (Harlow and Benbrook, 2019). Similarly, the Alt-Right and Antifa use in-group hashtags in their Twitter profile descriptions and tweets for coalition building (Xu, 2020); and White supremacy groups post in-group status enhancement messages to burnish their group images (Phadke and Mitra, 2020).

In-group members also disparage the out-group to distinguish their in-group as a “better” one, which fulfills their need to belong and promotes in-group values (Cikara and Van Bavel, 2014). Hate groups use Twitter to educate their audience about problems with the out-groups, demanding restrictive policies targeting them (Phadke and Mitra, 2020). Notably, partisan posts featuring out-group animosity—negative expressions about counter-partisans—tend to receive more retweets on Twitter and reactions on Facebook than posts featuring in-group or moral-emotional language (Rathje et al., 2021). Also, right-wing political parties’ anti-immigrant messages on Facebook receive angry “reactions” through fanning out-group hate, which contributes to their follower increase on Facebook (Heiss and Matthes, 2020). A study analyzing popular QAnon-related comments on YouTube points in a similar direction: foreign affairs, particularly China-related topics, have remained quite robust over time and received more attention and “likes” (Miller, 2021).

Audience reaction

People on social media can react to others’ performance as their audience in many ways, from direct liking, sharing, and commenting to more complex reactions like getting an account “ratio-ed” (Bail, 2021). Such quantified engagement is valued and pursued by all kinds of actors, such as social media platforms, social media influencers, journalists, and politicians. We focus on two sought-after metrics of audience reaction on Twitter: the number of retweets and the number of followers. We argue that they are valid indicators of positive audience reaction to QAnon accounts’ identity performance on Twitter.

The retweet function allows an original tweet to spread quickly and extensively across Twitter networks, resulting in an exponential increase in message reach (Firdaus et al., 2018). As a key metric of message dissemination, it can indicate the content value of one’s tweets (Cha et al., 2010; Riquelme and González-Cantergiani, 2016). Retweets also shape the formation of new ties (Hutto et al., 2013), a way to grow followers (Zhang et al., 2021). Moreover, highly retweeted tweets can be considered engagement-worthy by algorithms and news-worthy in the eyes of journalists, contributing to algorithm and journalistic amplification (Wells et al., 2016; Zhang et al., 2018).

In a similar vein, followers not only function as direct audiences to facilitate message dissemination but also serve as a badge of popularity and status that further begets attention and power (Aral et al., 2009; Marwick, 2013; Wells et al., 2020). The follower count can provide a heuristic to determine the reach of content (Lee et al., 2018) and an indicator of the competence and trustworthiness of an account (O’Keefe, 2016). Brand followers on social media often engage in supportive behaviors: sharing or generating brand content and spreading the positive word-of-mouth about it (Kim, 2020). Similarly, politicians’ Twitter followers also amplify their messages, bestowing on them greater communicative power (e.g. Zhang et al., 2018).

News media amplification for expanded reach

In the hybrid media system where social media and news media depend on each other for audience attention, information can flow dynamically between social media and news media (Chadwick, 2017; Russell Neuman et al., 2014). Consequently, news media might amplify actors’ strategic performance on social media.

Existing research demonstrates that news media amplify the voices and enhance the influence of social media actors, including those from peripheral groups. The premise of this amplification is strong social media engagement. To move up the chain from social media to fringe media outlets and mainstream news media, movement activists (Freelon et al., 2018), clickbait media (Munger, 2020), extremists (Marwick and Lewis, 2017), and disinformation agents (Lukito et al., 2020) can stir up social media frenzies and collect proof of credibility and newsworthiness (e.g. likes and retweets) via coordinated messaging, trolling, providing controversial opinions, or supplying novel information. Donald Trump’s ability to make Twitter abuzz has increased news media attention to him (Wells et al., 2016). News media amplification can directly contribute to social media actors’ follower increase (Zhang et al., 2021) by exposing them to larger audiences. It can further help them gain broader public and elite attention by legitimizing their voices and stances (Freelon et al., 2018). Mainstream news media coverage of QAnon has increased substantially since QAnon adherents’ appearance at Trump’s Tampa Bay rally in 2018 (De Zeeuw et al., 2020). Although mainstream media have portrayed QAnon as “somewhat gullible or motivated by populist resentment,” their coverage has nonetheless introduced QAnon to large swaths of audiences and animated discussions about QAnon across social media platforms (De Zeeuw et al., 2020; Forberg, 2022).

Such news media amplification of social media actors and/or their content is related to one notable transition in journalistic practice. Increasingly, journalists incorporate in news stories social media content (Bossio, 2017) that supplies new information, represents “vox populi,” or has simply gone viral (Beckers and Harder, 2016; McGregor, 2019; Usher, 2017). This practice is propelled by the journalistic shift to including input from networked audiences (Loosen and Schmidt, 2012); it is also driven by a market orientation to give people what they want so as to capture their attention (Hanusch and Tandoc, 2019) and produce content cost-effectively under mounting economic pressure (Bakker, 2012).

From audience reaction and media amplification back to strategic performance

After discussing how actors from peripheral groups can gain social media engagement as well as news media amplification by strategically staging their social media performance, we now turn to the possibility that social media performance might in turn be influenced by audience reaction and media amplification. While certain goals drive strategic performances, social media actors are generally attentive to their target audience (boyd and Heer, 2006) and may learn from audience reaction and adapt their self-presentation accordingly (Ellison et al., 2006). On social media, as Bail (2021) eloquently puts it, it is “all too human” for us to “perform different identities, observe how other people react, and update our presentation of self to make us feel like we belong” (p. 53). Since positive feedback is associated with perceived popularity and self-esteem (Burrow and Rainone, 2017; Meeus et al., 2019), the strategic performance that draws positive audience reaction (i.e. more retweets and followers) can be repeatedly enacted. Furthermore, social reinforcement and norm learning can motivate additional performance: expression of moral outrage by politically engaged users on Twitter will intensify if rewarded with more retweets and likes (Brady et al., 2021). In addition, when social media actors are aware of their posts being quoted in news articles, such media amplification and, by extension, legitimization might further embolden their performance on social media.

With the discussion above, we propose a set of hypotheses about the relationships between strategic performance, audience reaction, and news media amplification in the context of QAnon believers on Twitter (Figure 1). We first hypothesize that by posting messages that (a) affirm and valorize in-group membership and (b) launch out-group criticism and attack, QAnon accounts can stimulate positive audience reaction, that is, more retweets and followers.

H1: The number of QAnon accounts’ tweets featuring in-group solidarity positively predicts the number of (a) retweets and the number of (b) followers they get.

H2: The number of QAnon accounts’ tweets featuring out-group animosity positively predicts the number of (a) retweets and the number of (b) followers they get.

We also hypothesize that QAnon accounts’ performance of group identity and the audience reaction it sparks can both trigger news media amplification.

H3: The number of tweets featuring (a) in-group solidarity and the number of tweets featuring (b) out-group animosity positively predict the amount of news media amplification QAnon accounts receive.

H4: The number of (a) retweets and the number of (b) followers that QAnon accounts get positively predict the amount of news media amplification they receive.

Finally, we hypothesize that both audience reaction and news media amplification strengthen QAnon accounts’ performance of group identity.

H5: The number of (a) retweets and the number of (b) followers that QAnon accounts get positively predict the number of tweets featuring in-group solidarity.

H6: The number of (a) retweets and the number of (b) followers that QAnon accounts get positively predict the number of tweets featuring out-group animosity.

H7: The amount of news media amplification QAnon accounts receive positively predicts the number of tweets featuring in-group solidarity.

H8: The amount of news media amplification QAnon accounts receive positively predicts the number of tweets featuring out-group animosity.

Method

Data

We selected QAnon Twitter accounts from an existing project that first conducted a network sampling of influential accounts that talked about politics in 2018 and 2019, respectively, and then collected the tweets of the sampled accounts on a daily basis (Zhang et al., 2022b). This user-centered data collection approach is suitable to chart individual QAnon accounts’ growth. Among the sampled accounts, we searched in their profile descriptions for QAnon-related keywords: “q,” “qanon,” “deep state,” “deepstate,” “pizzagate,” “pedogate,” “wwg1wga,” “wwgwga,” “obamagate,” “obama gate,” and “sethrich,” which yielded 263 such accounts in the 2018 sample and 342 accounts in the 2019 sample.

Three additional steps were taken to ensure data quality. First, two authors perused each account’s profile description to eliminate ostensibly irrelevant accounts (down to 412 unique accounts from both samples). Second, to remove accounts with data gaps too large for the subsequent analysis, we filled the missing values in each account’s follower counts with linear interpolation and ARIMA modeling. We also removed accounts whose follower counts could not be reasonably forecast possibly due to data sparsity (the forecast follower counts for those accounts were negative, down to 242 accounts). Third, among the remaining accounts, we checked the potential coordination in the form of mutual retweeting and found little evidence for it.¹ We selected two periods where tweets for each sample were continuously collected: between 1 November 2018 and 1 October 2019 for the 2018 sample (3,937,505 tweets with 626,983 non-retweets and 3,310,522 retweets); and between 1 February 2020 and 31 December 2020 for the 2019 sample (3,325,777 tweets with 467,497 non-retweets and 2,858,280 retweets).² Overall, this set of 242 QAnon accounts was fairly active (producing 82 tweets per day on average) and some achieved a significant measure of success (one account accumulating 190,080 followers and one tweet receiving 21,855 retweets). These features make them appropriate cases for this study.

To investigate media amplification, we collected data about these QAnon accounts’ media uptake from Media Cloud, an open-source platform that ingests news stories from around the world. To comprehensively capture all media uptake of the QAnon accounts’ tweets, we compiled a total of 1503 unique media sources with pre-coded partisan leanings based on four sources: Media Cloud’s own collection (Faris et al., 2017), Buzzfeed’s hyperpartisan³ and fake news⁴ collections, and the US far-right conspiracy collections from Media Cloud (Collection #214598068) (see Supplemental Appendix I for all the media outlets). Then each account’s screen name (@screen_name) was used as a keyword for story search in Media Cloud among the media outlets. These accounts received sparse media uptake. Only 31 accounts’ tweets appeared in 110 stories during the two time periods. Among them, 100 stories came from hyper-conservative outlets, 2 from conservative outlets, 2 from moderate outlets, and 6 from liberal/hyper-liberal outlets, demonstrating that the media uptake of QAnon accounts occurred mostly in the hyper-conservative media sphere.

Measures

Number of in-group solidarity tweets and number of out-group animosity tweets

To map out the identity performance strategies, we conducted topic modeling of the unique tweets with latent Dirichlet allocation (LDA) (Maier et al., 2018) and Gibbs sampling (Griffiths and Steyvers, 2004). Log-likelihood and estimated perplexity (Barberá et al., 2019) suggested that the optimal number of topics should be 140. Two authors independently interpreted the topics and assigned them to one of the three categories—in-group solidarity, out-group animosity, and other—before discussing and finalizing the category assignment. In-group solidarity tweets tagged and networked with in-group members, expressed support for Trump, and interpreted or created QAnon messages. Out-group animosity tweets criticized various targets like the Washington establishment, liberals, Democrats, Hollywood, and the mainstream media. Each tweet was assigned the topic number with the highest topic probability, a process that yielded 1,452,238 (20% of all tweets) in-group solidarity tweets and 1,127,715 (16% of all tweets) out-group animosity tweets. For each account, the daily in-group solidarity and out-group animosity tweet counts were tallied. See Supplemental Appendix II for the topic modeling process and results.

Number of retweets

Given the uneven number of tweets the QAnon accounts posted, we computed the mean number of retweets that a QAnon account received within a day by averaging the retweet counts of its original tweets (i.e. non-retweets; tweets created by oneself). The 242 QAnon accounts on average received about 12 retweets per original tweet they posted.

Number of followers

At the time of daily data collection, an account’s follower count was queried in real time. When an account had multiple follower counts within a day, we took the average. The 242 QAnon accounts experienced uneven follower change, with some enjoying steady follower growth and some suffering gradual or occasionally sharp follower loss. Among them, about one-third (N = 84) experienced consistent follower growth (see Supplemental Appendix III for how consistent follower growth was measured).

Amount of media uptake

Although news media sources were grouped into five categories based on their partisan leanings, we collapsed them into a single media uptake count due to data sparsity. A qualitative analysis revealed that QAnon tweets appeared most frequently in conservative and hyper-conservative outlets as information sources, evidence, or public opinion,⁵ though very few articles in hyper-liberal outlets quoted QAnon tweets to analyze this phenomenon.⁶

Time series modeling

The data collection process above yielded daily counts of the following variables for each QAnon account: in-group solidarity tweets, out-group animosity tweets, retweets, followers, and news media uptake.⁷ We applied log transformation to the number of in-group solidarity tweets, the number of out-group animosity tweets, and the number of retweets to address the heavy-tailed distribution of these variables (Keene, 1995); we also applied square root transformation to the number of followers for variance stabilization (Bartlett, 1947) (Supplemental Appendix IV and Supplemental Appendix V).

We analyzed the time series data using the techniques from Barberá et al. (2019) in their analysis of the temporal relationship between politicians, publics, and news media regarding their attention to different topics of public interest on Twitter. Specifically, we performed the augmented Dickey–Fuller test on the time series of the transformed endogenous variables and found that they were all stationary; then we applied vector autoregression (VAR) modeling, where each dependent variable is a function of its own previous lags plus the lags of the other variables, to estimate the relationships between the five endogenous variables while accounting for the exogenous variables (241 dummy variables for the 242 QAnon accounts and another dummy variable indicating the two samples). Finally, we applied the impulse response functions (IRFs) on the fitted VAR model results to estimate the significance, duration, and magnitude of the effect of a one-unit increase in one variable on another variable.

Results

Following Barberá et al.’s (2019) approach, we focus on two different types of IRFs for a 15-day period based on VAR modeling. In Figure 2, each panel shows how much cumulative percentage-point increase in the panel title is predicted to occur after a one-time (in gray) and a permanent (in black) 10 percentage-point increase in the variables in the y-axis 15 days ago.

We first present findings on whether the performance of group identity predicts audience reaction. In-group solidarity tweets predict retweets positively (supporting H1a) and followers negatively (not supporting H1b). Retweets (log-transformed) will increase by 0.73% (95% confidence interval [CI] = [0.59%, 0.82%]) after a one-time 10 percentage-point increase in in-group solidarity tweets (log-transformed), and by 1.86% (95% CI = [1.52%, 2.09%]) following a permanent 10 percentage-point such increase. However, followers (square root transformed) are predicted to decrease by 0.31% (95% CI = [−0.40%, −0.21%]) after a one-time 10 percentage-point increase in in-group solidarity tweets and by 0.80% (95% CI = [−1.03%, −0.56%]) after a permanent 10 percentage-point such increase.

Out-group animosity tweets have a similar predicted effect on retweets: retweets are predicted to increase by 0.70% (95% CI = [0.56%, 0.78%]) 15 days after a one-time 10 percentage-point increase in out-group animosity tweets; by 2.24% (95% CI = [1.81%, 2.52%]) after a permanent 10 percentage-point increase in such tweets (supporting H2a). Neither a one-time increase nor a permanent increase in out-group animosity tweets predicts followers (not supporting H2b). These results suggest that the performance of group identity can directly and positively influence retweets, though it has little or a small negative relationship with followers. It is worth noting that followers are positively predicted by retweets. The number of followers will increase by 0.13% (95% CI = [0.08%, 0.18%]) and 0.47% (95% CI = [0.29%, 0.64%]) after a one-time and a permanent 10 percentage-point increase in retweets.

Then we examine whether in-group solidarity tweets (H3a), out-group animosity tweets (H3b), retweets (H4a), and followers (H4b) positively predict news media amplification. Our results show that the performance of group identity does not predict media uptake, rejecting H3a and H3b. However, media uptake increases by 0.13% (95% CI = [0.05%, 0.20%]) and by 0.42% (95% CI = [0.16%, 0.66%]) following a one-time and a permanent 10 percentage-point increase in retweets. Similarly, after a one-time and a permanent 10-point increase in followers, media uptake also increases by 2.45% (95% CI = [1.68%, 2.80%]) and 4.81% (95% CI = [3.30%, 5.49%]), respectively. Therefore, H4a and H4b are supported.

Finally, we turn to whether audience reaction and news media amplification may in turn influence identity performance. Our results show that in-group solidarity tweets are predicted to increase by 2.18% (95% CI = [2.03%, 2.18%]) after a one-time 10 percentage-point increase in retweets; by 7.00% (95% CI = [6.52%, 7.00%]) after a permanent 10-point increase in retweets, supporting H5a. However, in-group solidarity tweets are negatively predicted by followers (one-time effect: −2.62%, 95%CI = [−3.15%, −1.77%], permanent effect: −5.04%, 95%CI = [−6.11%, −3.39%]), thus not supporting H5b.

Out-group animosity tweets are predicted by both retweets and followers, lending support for H6a and H6b. A one-time 10 percentage-point increase in retweets and followers predicts a 1.09% (95% CI = [0.97%, 1.14%]) and 4.64% (95% CI = [4.46%, 4.76%]) increase in out-group animosity tweets after 15 days, respectively. A permanent 10 percentage-point increase in retweets and followers predicts a 3.49% (95% CI = [3.13%, 3.63%]) and a 9.10% (95% CI = [8.71%, 9.37%]) increase in out-group animosity tweets, respectively. The results suggest that the performance of out-group animosity is more adaptive than the performance of in-group solidarity, since the former responds to both retweet and follower changes.

Media uptake has a neglectable or insignificant effect on in-group solidarity tweets (95% CI = [0.00%, 0.01%] and [0.00%, 0.05%]) and out-group animosity tweets (95% CI = [0.00%, 0.01%] and [0.00%, 0.05%]), lending little support for H7 and H8. See Supplemental Appendix VI for the IRFs data table. Figure 3 provides a summary of the supported hypotheses.

Discussion

To explore how obscure groups in society can become mainstream, we propose a framework consisting of three interlinked components: strategic performance by social media actors, social media audience reaction, and news media amplification. Our empirical analysis demonstrates that a peripheral group can expand and adapt over time by playing the game of social media and tapping into the interconnected media system. Specifically, the performance of group identity helps QAnon accounts attract audience engagement in the form of retweets. Also, the media amplification of QAnon accounts results not directly from their group identity performance, but from positive audience reaction to their performance on Twitter. Furthermore, audience reaction on Twitter has fueled their group identity performance, particularly their performance of out-group animosity.

First of all, QAnon accounts employ a range of tactics to perform in-group solidarity and program their network, such as networking with in-group members, highlighting shared political values, and expressing support for political leaders. Consistent with social identity theory (Tajfel et al., 1979), this finding shows that QAnon accounts promote shared values and beliefs to bond with in-group members, speaking of how social media provide a hotbed for members of a group, marginal or not, to network with each other, seek gratification, and build group cohesion (Benkler, 2006; Castells, 2007; Törnberg, 2022).

Attacking out-groups is another strategy in QAnon accounts’ social media performance playbook. And out-group animosity tweets are effective, arguably more so in driving retweets than in-group solidarity tweets, evidenced by their larger permanent effect on retweets. According to social identity theory, disliking and discriminating against common enemies can foster a sense of belonging and catalyze group formation. By attacking out-groups (e.g., Democrats and the mainstream media), QAnon accounts might appeal to and resonate with those not initially belonging to the QAnon community, such as mainstream conservatives. This suggests the power that peripheral groups can gain by setting up boogeymen and provoking emotions and engagement. In this process, the participatory culture of online conspiracy groups can produce detrimental social outcomes, such as physical violence and spread of misinformation (Ovide, 2021).

Although the performance of group identity does not predict the increase in followers directly, it might do so indirectly through driving up retweets, which can positively predict follower increase. This finding is consistent with existing evidence showing the role of retweets in expanding one’s social network (Wells et al., 2020; Zhang et al., 2021), suggesting that for an obscure account to grow its follower network, collecting social proof, like gaining more retweets, is important.

Second, the key for QAnon accounts to gain media amplification is not performing group identity, but their audience’s enthusiastic reaction to the performance. In other words, news media amplification of a given social media user might be contingent on other social media users’ amplification of the user. This pattern indicates the importance of social media public opinion in journalists’ decisions to incorporate certain social media content. In particular, since massive followers and retweets can signal the credibility of a source as well as the wide reach of its messages on social media (Lee et al., 2018), quoting its “hot takes” in news reports can potentially maximize news appeal. This finding dovetails with prior research showing how journalists construct news stories by incorporating social media content and how audience reaction to social media content can influence their news judgment (Mourão and Harlow, 2020). In the hybrid media system driven by the attention economy (Chadwick, 2017; Webster, 2014), news producers strive to adapt to new challenges in order to produce engaging content in a cost-effective way. Embedding engaging content from influential social media accounts exactly serves this purpose.

Importantly, the vast majority of stories (91%) quoting QAnon tweets appear in conservative and hyper-conservative news outlets. The few stories in moderate, liberal and hyper-liberal news outlets are mainly news analyses of the QAnon conspiracy. With indiscriminate media amplification resulting in the mainstreaming of anti-Muslim organizations and sentiment (Bail, 2012), the spread of disinformation (Lukito et al., 2020), and the growth of extremist groups (Phillips, 2018), scholars call for news media to adapt to new reality. Donovan and boyd (2021) recommend that journalists practice strategic silence before they know how to strategically amplify content in a socially responsible way. Our finding seems to suggest that mainstream and moderate news media have learned their lesson: they have exercised their amplification power with greater caution to guard against exploitation by potential bad actors (Abubakar, 2020). The institution-wide reckoning might have produced a welcomed short-term change in journalistic practice.

Nevertheless, the finding that QAnon accounts’ influence beyond Twitter is relatively bounded within an echo chamber of hyper-conservative media outlets does not detract from their success. QAnon accounts have banded together and created an active online community on Twitter, directly paving the way for like-minded news outlets’ amplification of them. This way, a peripheral group can expand its influence beyond social media, gain recognition from media that they trust, and become more outspoken about its views. This finding aligns with research showing that the right-wing media ecosystem is overall cloistered and insulated from the larger media ecology (Benkler et al., 2018; Faris et al., 2017). In this regard, the power of the hyper-conservative wing of this ecosystem might not come from scale but from depth: the uptake of QAnon accounts’ tweets in hyper-conservative media can lead to the further radicalization and audacity of QAnon believers.

Finally, we show that though media amplification has little impact on identity performance, audience reaction can greatly influence it. QAnon accounts are not only strategic in their online performance, but also attentive to the audience reaction they get. This finding corresponds to existing research on how positive social feedback can foster a given behavioral pattern (Brady et al., 2021) and how people selectively self-disclose to maximize social approval and minimize criticism (Walsh et al., 2020). Audience reaction seems to affect the performance of out-group animosity more than the performance of in-group solidarity, with the former sensitive to both retweet and follower increase. Arguably, by retweeting and following QAnon accounts, social media audience can add fuel to the QAnon outrage machine.

This article fills some important gaps in current research. While some studies investigate the behaviors of QAnon actors on social media (e.g. Moran et al., 2021; Papasavva et al., 2021), little is known regarding how some QAnon accounts have become influential on social media and whether their popularity is aided by their social media performance and/or news media amplification. Focusing on what contributes to QAnon’s increased influence, this research sheds light on how the media system—including social media as well as news media—can facilitate the growth of conspiratorial actors and fringe ideologies. Also, though abundant research shows what message characteristics generate user engagement in the form of retweets or reactions, such as moral-emotional language, negativity, and identity-related terms (Brady et al., 2017; Fine and Hunt, 2021; Rathje et al., 2021), we demonstrate that group identity performance can also drive follower growth and media amplification. Finally, by taking into account the possibility of actors adapting to audience reaction, we reveal the role that social media audiences play in influencing the messaging strategies of social media actors.

However, this study has several limitations. The current analysis cannot rule out alternative explanations of the relationships between our key variables. For example, the overall level of activity rather than the strategic performance might explain the increase in retweets, such that the more tweets QAnon accounts post, the more engagement they get, regardless of the tweet content. We conducted an analysis with everything else remaining the same except for replacing in-group solidarity tweets and out-group animosity tweets with the total number of tweets. Results show that though the total number of tweets can positively predict the number of retweets (one-time effect: 0.76%, 95% CI = [0.59%, 0.87%]; permanent effect: 1.69%, 95% CI = [1.31%, 1.92%]), the permanent effect is much smaller than that of the strategic performance of in-group solidarity and out-group animosity. This provides further evidence for how strategic performance of group identity is a key predictor of audience engagement, rather than indiscriminate hyperactivity. Besides, we conducted another robustness check by replacing the average number of retweets with the total number of retweets, which yielded identical results (Supplemental Appendix VII). Although our study is limited in drawing causal inferences, these analyses provide additional support for our findings.

Our measurement of media amplification focuses on the media uptake of individual QAnon accounts and their tweets, rather than media coverage of QAnon in general. Since we are interested in individual trajectories, this measurement is appropriate. However, it is reasonable to argue that just by giving exposure to the QAnon movement, news media can lend credibility to individual QAnon accounts.

In addition, though our automated text analysis presents clear patterns in QAnon tweets, it cannot capture nuances. Peripheral groups might employ strategies other than group identity performance, such as fear appeal and moral judgments (Phadke and Mitra, 2020). Future research can examine their tweets qualitatively. An in-depth interview with QAnon social media users might offer rich insights into their motivations, goals, and behaviors on social media.

Finally, this is a case study in a specific context. To increase the external validity of our conclusions, future quantitative research should study peripheral online communities on other social media platforms and in other countries. In particular, though Twitter functions as a key platform for journalists to gather information (Jurkowitz and Gottfried, 2022), social media platforms have drastically different dynamics and cannot be reduced to a monolith, requiring attention to cross-platform differences. Future research should chart the growth of peripheral ideologies in other platforms, especially the alt-tech ones like Gab, Parler, and Rumble.

Conclusion

Exploring the mechanisms of the mainstreaming of peripheral groups in the current media ecology, our three-pronged framework attends to social media logics and hybrid media dynamics. It can be applied to different social media platforms, where users perform their identities, accumulate shares/retweets/reposts and subscribers/followers, and garner news media attention. Situated in the context of QAnon accounts on Twitter, our article demonstrates how peripheral groups like QAnon can use Twitter to perform their group identity and spark audience reaction, which can in turn reinforce their group identity performance and drive news media amplification. These dynamics speak of not only how social media actors can tap into the affordances of social media (Benkler, 2006; boyd, 2015) to form communities and project their voices (Freelon et al., 2018; Phillips, 2018), but also how news media can further amplify and legitimize their voices. Both paths can empower and embolden a peripheral group, providing a potent formula for political radicalization and mobilization. It is worth noting that Twitter audience, likely QAnon believers and sympathizers in this case, play a critical role in all these processes: by retweeting QAnon tweets and following QAnon accounts, they inflame QAnon’s expression of out-group animosity and prompt partisan media’s amplification of QAnon tweets. Taken together, these results portray a media ecosystem conducive to the overall amplification of peripheral groups, regardless of whether they are good or bad for society.

Supplemental Material

sj-docx-1-nms-10.1177_14614448221137324 – Supplemental material for How a peripheral ideology becomes mainstream: Strategic performance, audience reaction, and news media amplification in the case of QAnon Twitter accounts

Supplemental material, sj-docx-1-nms-10.1177_14614448221137324 for How a peripheral ideology becomes mainstream: Strategic performance, audience reaction, and news media amplification in the case of QAnon Twitter accounts by Yini Zhang, Zhiying Yue, Xiyu Yang, Fan Chen and Nojin Kwak in New Media & Society

Footnotes

Data availability

The social media and news media datasets are available upon request. The R code for data analysis is available at

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Yini Zhang

Zhiying Yue

Supplemental material

Supplemental material for this article is available online.

Notes

Author biographies

Yini Zhang (PhD University of Wisconsin–Madison) is an Assistant Professor in the Department of Communication at the University at Buffalo, State University of New York. She studies social media and political communication, using computational methods.

Zhiying Yue (PhD University at Buffalo) is a postdoctoral Research Fellow at the Digital Wellness Lab, Boston Children’s Hospital, and Harvard Medical School. Her research focuses on the use and effect of social media on individual and community well-being.

Xiyu Yang (MA University of Wisconsin–Madison) is a data science master’s student at Harvard John A. Paulson School of Engineering and Applied Sciences. Her current research interests include social network analysis, causal inference, and applications of statistics in social sciences.

Fan Chen (PhD University of Wisconsin–Madison) is a Data Scientist at Google. He specializes in developing statistical tools for social media and genomics data analysis.

Nojin Kwak (PhD University of Wisconsin–Madison) is Vice Provost for International Education and Professor in the Department of Communication at the University at Buffalo, State University of New York. His research includes the role of social and digital media in citizenship and politics.

References

Abubakar

(2020) News values and the ethical dilemmas of covering violent extremism. Journalism & Mass Communication Quarterly 97(1): 278–298.

Anwar

Ilyas

Yaqub

, et al. (2021) Analyzing qanon on twitter in context of us elections 2020: analysis of user messages and profiles using VADER and BERT topic modeling. In: Proceedings of the 22nd annual international conference on digital government research, Omaha, NE, 9–11 June, pp. 82–88. New York: ACM.

Aral

Muchnik

Sundararajan

(2009) Distinguishing influence-based contagion from homophily-driven diffusion in dynamic networks. Proceedings of the National Academy of Sciences of the United States of America 106(51): 21544–21549.

Bail

(2012) The fringe effect: Civil society organizations and the evolution of media discourse about Islam since the September 11th attacks. American Sociological Review 77(6): 855–879.

Bail

(2021) Breaking the Social Media Prism. Princeton, NJ: Princeton University Press.

Bakker

(2012) Aggregation, content farms and Huffinization: the rise of low-pay and no-pay journalism. Journalism Practice 6(5–6): 627–637.

Barberá

Casas

Nagler

, et al. (2019) Who leads? Who follows? Measuring issue attention and agenda setting by legislators and the mass public using social media data. American Political Science Review 113(4): 883–901.

Bartlett

(1947) The use of transformations. Biometrics 3(1): 39–52.

Beckers

Harder

(2016) “Twitter just exploded”: social media as alternative vox pop. Digital Journalism 4(7): 910–920.

10.

Benkler

(2006) The Wealth of Networks: How Social Production Transforms Markets and Freedom. New Haven and London: Yale University Press. 528 pp. $40.00 (papercloth). Social Science Computer Review 26(2): 259–261.

11.

Benkler

Faris

Roberts

(2018) Network propaganda: Manipulation, disinformation, and radicalization in American politics. Oxford University Press.

12.

Blumler

Kavanagh

(1999) The third age of political communication: influences and features. Political Communication 16(3): 209–230.

13.

Bossio

(2017) News in social media environments: journalism in a “post-truth” world. Journalism and Social Media 2017: 133–152.

14.

boyd

(2015) Social media: A phenomenon to be analyzed. Social Media+ Society 1(1): 5580148.

15.

boyd

Heer

(2006) Profiles as conversation: networked identity performance on Friendster. In: Proceedings of the 39th annual Hawaii international conference on system sciences HICSS’06 3906(HICSS’06), Kauai, HI, 4–7 January. New York: IEEE.

16.

Brady

McLoughlin

Doan

, et al. (2021) How social learning amplifies moral outrage expression in online social networks. Science Advances 7(33): eabe5641.

17.

Brady

Wills

Jost

, et al. (2017) Emotion shapes the diffusion of moralized content in social networks. Proceedings of the National Academy of Sciences of the United States of America 114(28): 7313–7318.

18.

Burrow

Rainone

(2017) How many likes did I get? Purpose moderates links between positive social media feedback and self-esteem. Journal of Experimental Social Psychology 69: 232–236.

19.

Castells

(2007) Communication, power and counter-power in the network society. International Journal of Communication 1(1): 238–266.

20.

Cha

Haddadi

Benevenuto

, et al. (2010) Measuring user influence in Twitter: the million follower fallacy. In: Proceedings of the international AAAI conference on Web and social media (vol. 4: no. 1). Available at: https://ojs.aaai.org/index.php/ICWSM/article/view/14033#:~:text=In%20this%20paper%2C%20using%20a, We%20make%20several%20interesting%20observations.

21.

Chadwick

(2017) The Hybrid Media System: Politics and Power. Oxford: Oxford University Press.

22.

Christensen

Jensen

(2018) The JohnLock conspiracy, fandom eschatology, and longing to belong. Transformative Works and Cultures 27: 2.

23.

Cikara

Van Bavel

(2014) The neuroscience of intergroup relations: an integrative review. Perspectives on Psychological Science 9(3): 245–274.

24.

De Zeeuw

Hagen

Peeters

, et al. (2020) Tracing normiefication: a cross-platform analysis of the QAnon conspiracy theory. First Monday. Available at: https://firstmonday.org/ojs/index.php/fm/article/view/10643

25.

Djafarova

Trofimenko

(2019) “Instafamous”: credibility and self-presentation of micro-celebrities on social media. Information, Communication & Society 22(10): 1432–1446.

26.

Donovan

Boyd

(2021) Stop the presses? Moving from strategic silence to strategic amplification in a networked media ecosystem. American Behavioral Scientist 65(2): 333–350.

27.

Ellison

Heino

Gibbs

(2006) Managing impressions online: self-presentation processes in the online dating environment. Journal of Computer-mediated Communication 11(2): 415–441.

28.

Faris

Roberts

Etling

, et al. (2017) Partisanship, propaganda, and disinformation: online media and the 2016 U.S. presidential election. Berkman Klein Center Research Publication. Available at: https://dash.harvard.edu/handle/1/33759251

29.

Fine

Hunt

(2021) Negativity and elite message diffusion on social media. Political Behavior. Epub ahead of print 11 August. DOI: 10.1007/s11109-021-09740-8.

30.

Firdaus

Ding

Sadeghian

(2018) Retweet: a popular information diffusion mechanism —a survey paper. Online Social Networks and Media 6: 26–40.

31.

Forberg

(2022) From the fringe to the fore: an algorithmic ethnography of the far-right conspiracy theory group QAnon. Journal of Contemporary Ethnography 51(3): 291–317.

32.

Freelon

McIlwain

Clark

(2018) Quantifying the power and consequences of social media protest. New Media & Society 20(3): 990–1011.

33.

Gallagher

Davey

Hart

(2020) The genesis of a conspiracy theory: key trends in QAnon activity since 2017. ISD reports. Available at: https://www.isdglobal.org/wp-content/uploads/2020/07/The-Genesis-of-a-Conspiracy-Theory.pdf

34.

Garry

Walther

Rukaya

, et al. (2021) QAnon conspiracy theory: examining its evolution and mechanisms of radicalization. Journal for Deradicalization 26: 152–216.

35.

Giddens

(1991) Modernity and Self-Identity. Cambridge: Polity Press.

36.

Goffman

(1959) The Presentation of Self in Everyday Life. Garden City, NY: Doubleday.

37.

Griffiths

Steyvers

(2004) Finding scientific topics. Proceedings of the National Academy of Sciences of the United States of America 101(suppl_1): 5228–5235.

38.

Hanley

Kumar

Durumeric

(2022) No calm in the storm: investigating QAnon website relationships. In: Proceedings of the international AAAI conference on Web and social media, pp. 299–310. Available at: https://arxiv.org/abs/2106.15715

39.

Hannah

(2021) QAnon and the information dark age. First Monday 26(2): 10868.

40.

Hanusch

Tandoc

Jr (2019) Comments, analytics, and social media: the impact of audience feedback on journalists’ market orientation. Journalism 20(6): 695–713.

41.

Harlow

Benbrook

(2019) How# Blacklivesmatter: exploring the role of hip-hop celebrities in constructing racial identity on Black Twitter. Information, Communication & Society 22(3): 352–368.

42.

Haupt

Jinich-Diamant

, et al. (2021) Characterizing twitter user topics and communication network dynamics of the “Liberate” movement during COVID-19 using unsupervised machine learning and social network analysis. Online Social Networks and Media 21: 100114.

43.

Heiss

Matthes

(2020) Stuck in a nativist spiral: content, selection, and effects of right-wing populists’ communication on Facebook. Political Communication 37(3): 303–328.

44.

Hutto

Yardi

Gilbert

(2013) A longitudinal study of follow predictors on Twitter. In: Proceedings of the SIGCHI conference on human factors in computing systems, Paris, 27 April–2 May, pp. 821–830. New York: ACM.

45.

Jackson

Gorman

Nakatsuka

(2021) QAnon on Twitter: An Overview. Washington, DC: Institute for Data, Democracy & Politics, George Washington University.

46.

Jackson

Foucault Welles

(2015) Hijacking# myNYPD: social media dissent and networked counterpublics. Journal of Communication 65(6): 932–952.

47.

Jungherr

(2014) Twitter in politics: a comprehensive literature review. Available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2402443

48.

Jurkowitz

Gottfried

(2022) Twitter is the go-to social media site for US journalists, but not for the public. Pew Research Center. Available at: https://www.pewresearch.org/fact-tank/2022/06/27/twitter-is-the-go-to-social-media-site-for-u-s-journalists-but-not-for-the-public/

49.

Keene

(1995) The log transformation is special. Statistics in Medicine 14(8): 811–819.

50.

Kim

(2020) The value of followers on social media. IEEE Engineering Management Review 48(2): 173–183.

51.

Lee

Kim

Coe

(2018) When social media become hostile media: an experimental examination of news sharing, partisanship, and follower count. Mass Communication and Society 21(4): 450–472.

52.

Loosen

Schmidt

J-H

(2012) (Re-) discovering the audience: the relationship between journalism and audience in networked digital media. Information, Communication & Society 15(6): 867–887.

53.

Lukito

Suk

Zhang

, et al. (2020). The wolves in sheep’s clothing: How Russia’s Internet Research Agency tweets appeared in US news as vox populi. The International Journal of Press/Politics 25(2): 196–216.

54.

McGregor

(2019) Social media as public opinion: how journalists use social media to represent public opinion. Journalism 20(8): 1070–1086.

55.

Maddox

(2022) Micro-celebrities of information: mapping calibrated expertise and knowledge influencers among social media veterinarians. Information, Communication & Society. Epub ahead of print 13 August. DOI: 10.1080/1369118X.2022.2109980.

56.

Maier

Waldherr

Miltner

, et al. (2018) Applying LDA topic modeling in communication research: toward a valid and reliable methodology. Communication Methods and Measures 12(2–3): 93–118.

57.

Marwick

(2013) Status update: Celebrity, publicity, and branding in the social media age. Yale University Press.

58.

Marwick

boyd

(2011) I tweet honestly, I tweet passionately: Twitter users, context collapse, and the imagined audience. New Media & Society 13(1): 114–133.

59.

Marwick

Lewis

(2017) Media Manipulation and Disinformation Online. New York: Data & Society Research Institute.

60.

Marwick

Partin

(2022) Constructing alternative facts: populist expertise and the QAnon conspiracy. New Media & Society. Epub ahead of print 1 May. DOI: 10.1177/146144482210902.

61.

Meeus

Beullens

Eggermont

(2019) Like me (please?): connecting online self-presentation to pre-and early adolescents’ self-esteem. New Media & Society 21(11-12): 2386–2403.

62.

Miller

(2021) Characterizing QAnon: analysis of YouTube comments presents new conclusions about a popular conservative conspiracy. First Monday 26(2): 10168.

63.

Moran

Prochaska

Schlegel

, et al. (2021) Misinformation or activism: mapping networked moral panic through an analysis of # savethechildren. AOIR Selected Papers of Internet Research. Available at: https://journals.uic.edu/ojs/index.php/spir/article/view/12212

64.

Mourão

Harlow

(2020) Awareness, reporting, and branding: exploring influences on Brazilian journalists’ social media use across platforms. Journal of Broadcasting & Electronic Media 64(2): 215–235.

65.

Munger

(2020) All the news that’s fit to click: the economics of clickbait media. Political Communication 37(3): 376–397.

66.

O’Keefe

(2015) Persuasion: Theory and Research. Thousand Oaks, CA: SAGE.

67.

Ovide

(2021) Trump: Twitter—QAnon: who’s to blame? New York Times, 7 January. https://www.nytimes.com/2021/01/07/technology/trump-twitter-qanon.html (accessed 9 September 2022).

68.

Papacharissi

(2012) Without you, I’m nothing: performances of the self on Twitter. International Journal of Communication 6: 1989–2006.

69.

Papasavva

Blackburn

Stringhini

, et al. (2021) “Is it a qoincidence?”: an exploratory study of QAnon on Voat. Proceedings of the Web Conference 2021: 460–471.

70.

Phadke

Mitra

(2020) Many faced hate: a cross platform study of content framing and information sharing by online hate groups. In: Proceedings of the 2020 CHI conference on human factors in computing systems, Honolulu, HI, 25–30 April, pp. 1–13. New York: ACM.

71.

Phillips

(2018) The Oxygen of Amplification: Better Practices for Reporting on Extremists, Antagonists, and Manipulators Online. New York: Data & Society Research Institute.

72.

Rathje

Van Bavel

Van Der Linden

(2021) Out-group animosity drives engagement on social media. Proceedings of the National Academy of Sciences 118(26): e2024292118.

73.

Riquelme

González-Cantergiani

(2016) Measuring user influence on Twitter: a survey. Information Processing & Management 52(5): 949–975.

74.

Russell Neuman

Guggenheim

Mo Jang

, et al. (2014) The dynamics of public attention: agenda-setting theory meets big data. Journal of Communication 64(2): 193–214.

75.

Tajfel

Turner

Austin

, et al. (1979) An integrative theory of intergroup conflict. In: Hatch

Schultz

(eds) Organizational Identity: A Reader. Oxford: Oxford University Press, pp. 56–65.

76.

Törnberg

(2022) How digital media drive affective polarization through partisan sorting. Proceedings of the National Academy of Sciences of the United States of America 119(42): e2207159119.

77.

Tufekci

(2013) “Not this one” social movements, the attention economy, and microcelebrity networked activism. American Behavioral Scientist 57(7): 848–870.

78.

Usher

(2017) Venture-backed news startups and the field of journalism: challenges, changes, and consistencies. Digital Journalism 5(9): 1116–1133.

79.

Van Dijck

(2013) The Culture of Connectivity: A Critical History of Social Media. Oxford: Oxford University Press.

80.

Van Dijck

Poell

(2013) Understanding social media logic. Media and Communication 1(1): 2–14.

81.

Walsh

Forest

Orehek

(2020) Self-disclosure on social media: the role of perceived network responsiveness. Computers in Human Behavior 104: 106162.

82.

Webster

(2014) Theories of the Information Society. New York: Routledge.

83.

Wells

Shah

Pevehouse

, et al. (2016) How Trump drove coverage to the nomination: hybrid media campaigning. Political Communication 33(4): 669–676.

84.

Wells

Zhang

Lukito

, et al. (2020) Modeling the formation of attentive publics in social media: the case of Donald Trump. Mass Communication and Society 23(2): pp181–205.

85.

(2020) Mapping connective actions in the global alt-right and Antifa counterpublics. International Journal of Communication 14: 22.

86.

Zhang

Chen

Lukito

(2022a) Network amplification of politicized information and misinformation about COVID-19 by conservative media and partisan influencers on Twitter. Political Communication. Epub ahead of print 18 August. DOI: 10.1080/10584609.2022.2113844.

87.

Zhang

Chen

Rohe

(2022b). Social media public opinion as flocks in a murmuration: Conceptualizing and measuring opinion expression on social media. Journal of Computer-Mediated Communication 27(1): zmab021.

88.

Zhang

Lukito

M-H

, et al. (2021) Assembling the networks and audiences of disinformation: how successful Russian IRA Twitter accounts built their followings, 2015–2017. Journal of Communication 71(2): 305–331.

89.

Zhang

Wells

Wang

, et al. (2018) Attention and amplification in the hybrid media system: the composition and activity of Donald Trump’s Twitter following during the 2016 presidential election. New Media & Society 20(9): 3161–3182.

90.

Zihiri

Lima

Han

, et al. (2022) QAnon shifts into the mainstream, remains a far-right ally. Heliyon 8(2): e08764.

91.

Zuckerman

(2019) QAnon and the emergence of the unreal. Journal of Design and Science 6(6): 1–14.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

7.50 MB