News Use as Amplification: Norwegian National,Regional,and Hyperpartisan Media on Facebook

Abstract

This study details the influence of hyperpartisan media actors in comparison to regional and national news media competitors by gauging audience engagement in relation to news on Facebook in Norway. Adopting the perspective of news use as a way of understanding such engagement, the study finds that followers of hyperpartisan Facebook Pages are more active than those following mainstream media Pages. The study also looks closer into what kinds of news are engaged with to higher degrees than others, building on these results in suggesting opportunities for future research into news production and consumption on Facebook.

Keywords

interactivity Communication Technology (CTEC)Mass Communication and Society (MC & S)newspaper and online news social media Norway descriptive statistics basic inferential statistics one way ANOVA log data analysis journalism studies Internet and technology

Introduction

Mirroring the “utopian rhetoric that surrounds new media technologies” (Papacharissi, 2002, p. 9), early ideas regarding online journalism were unrealistic regarding the active role of audience members implied by digital technology. Besides changes internal to media organizations brought on by increased audience feedback (e.g., Braun & Gillespie, 2011), sometimes posing threats to established ways of working (e.g., Hermida & Thurman, 2008), empirical findings regarding audience activity in the digital era have largely found that the supposed “people formerly known as the audience” (Rosen, 2006) remained rather aloof in relation to the seemingly paradigmatic suggestions of being integrated into the news production process (e.g., Larsson, 2012).

Seemingly coupled with the increased importance of social media such as Facebook for news consumption, such audience disinterest more demanding modes of participation (as discussed by Chung, 2008) appears to have caught up with media professionals who create and curate online offerings. Analyzing survey data from 2012 to 2015 regarding views of audience involvement among Norwegian media professionals, Krumsvik (2017) found that while the understanding of audiences as co-creators was frequent in 2012, the view of audiences as distributors of finished news items had grown to clear prominence in 2015. Such a view of audiences corresponds to the notion of the “news user” (Picone, 2017) or the “active recipient” (Singer et al., 2011)—a conceptualization suggesting that audience engagement be allowed as “amplifying” (Zhang, Wells, Wang, & Rohe, 2018) already published content.

Relatedly, the increasing importance of platforms such as Facebook has brought with it novel difficulties for media professionals. Indeed, news provision on social media has become associated with issues of “fake news” and disinformation, sometimes disseminated from hyperpartisan outlets. This latter term is used here to denote media outlets that present themselves as truth-telling substitutes for established media actors, and who are often described by themselves or by their competitors as in conflict with specific political interests or with their mainstream media counterparts (Fletcher, Cornia, Graves, & Nielsen, 2018; Tandoc, Lim, & Ling, 2017). As such, the publicist efforts of hyperpartisan media clearly diverges from the principles often associated with traditional journalism (e.g., Tuchman, 1978).

The present study draws on the developments discussed above to empirically assess the spread of hyperpartisan news on Facebook as users share, react to, and comment on such news items. Detailing the aforementioned Norwegian context, the article analyzes data drawn from the Facebook pages of mainstream national and regional media outlets, on one hand, and data from a series of Facebook pages operated by hyperpartisan interests, on the other hand. As the bulk of research regarding these issues emanate from the United States, there is a clear need to assess the growth of hyperpartisan content and “the spread of false information” (Chadwick, Vaccari, & O’Loughlin, 2018, p. 3) in comparison to mainstream media content in other contexts (see also Fletcher et al., 2018).

Featuring a rich selection of national and regional newspapers (MediaNorway, 2018), the Norwegian media system is an interesting case to study in this regard—especially as it also features high levels of trust in the media and high levels of newspaper readership (Vaage, 2018). Beyond assessing the influx of hyperpartisan outlets into the Norwegian media system, the selected approach will allow for differentiation between national and regional media outlets (as suggested by Sjøvaag, Stavelin, Karlsson, & Kammer, 2018). Conceptualizing the audience as news users (Picone, 2017), the present study poses the following research question:

RQ1: How does Facebook news use vary across national, regional, and hyperpartisan media outlets?

By understanding news use such as sharing or commenting on Facebook posts as forms of amplification—actions that “increase measures of engagement surrounding a person, message, or idea” (Zhang et al., 2018, p. 6)—the study at hand provides needed insights into the popularity of hyperpartisan media in comparison to their mainstream counterparts. As 84% of Norwegians have a Facebook account, and as 86% of those account holders report to make use of Facebook on a daily basis (Ipsos/MMI, 2018), the focus on this platform will provide us with findings related to the dominating social media platform in the studied context.

Literature Review

“Active audience” research has demonstrated how people engage with news to construct everyday narratives to help them make sense of public events unfolding around them (e.g., Bird, 1998; Nightingale, 2011). As pointed out by Chadwick and co-authors (2018), social media must now be considered an essential part of this process. Such a move toward social media platforms as the basis for news engagement provides certain opportunities and constrain others for news providers and consumers alike (e.g., Nielsen & Ganter, 2018).

Rather than taking the view of news consumers, the present adopts a news user perspective to make sense of the opportunities offered by Facebook. Picone (2016, 2017) shows how the term is suitable to conceptualize how people engage with the content offered to them by media actors on platforms such as the one studied here. Granted, though some users are more advanced in their engagement patterns, earning conceptual labels such as “producers” (Bruns, 2010), the bulk of news users engage in what can be considered as less demanding ways as they “comment on articles [and] share stories with their social networks” (Picone, 2016, p. 128). A news user, then, is rarely involved in the production of news, but rather engages with already produced news items. As such, a shift toward a view of audiences as news users rather than as citizen journalists can be discerned (Krumsvik, 2017). The question, then, is what kinds of news use does Facebook allow for? Bruns (2018) suggest that while previous forms of digital technology were normalized to fit the established norms of news organizations, the increased dependency on social media of these organizations is ushering in a reverse tendency. In short, journalism appears to be adapted into the norms set by platforms. Specifically, Facebook and others are increasingly functioning as digital intermediaries between news organizations and news users, effectively deciding what type of maneuvers are allowed at the hands of news professionals—and thereby also their users (Nielsen & Ganter, 2018).

From an empirical viewpoint, news use processes on Facebook are defined here as the ways news users are allowed to engage with posts that leave publicly visible traces of amplification. As of this writing, such use could be fashioned in three different ways—reacting, commenting, and sharing.

First, reactions can be seen as a development of the virtual “thumbs-up” like button that has become a staple on social media platforms. Expanding the possibility to “like” a post into a series of other emotionally themed feedback options (e.g., “love,” “sad,” “angry,” or “wow”), this diversification of one of the most-used feedback options on the platform is arguably in line with news user’s preferences for less demanding varieties of engagement (e.g., Chung, 2008). Such preferences toward behaviors akin to “clicktivism” (e.g., Kalsnes, 2016) seem to fit well with the skepticism toward more extensive audience engagement often expressed by media professionals (Bergström & Wadbring, 2015) because this type of feedback does not typically demand professional attention.

The same cannot comfortably be said when referring to the next variety of news use—commenting. Although comment fields have been lauded for their ability to yield discussion and debate (Deuze, Bruns, & Neuberger, 2007), they have also been problematic for a series of reasons. Media professionals have pointed to “concerns about quality and editorial responsibility” (Ihlebaek & Krumsvik, 2015, p. 4) when discussing comments posted to forums hosted on news organization web pages. For comments posted to social media, however, problematic utterances might take a different role. For example, in the Braun and Gillespie’s (2011) study on the adaptation processes undertaken by news organizations to conform to platforms such as the one under scrutiny here, one interviewee suggested that while “Low-quality comments [on a traditional news site] are terrible—poison” (p. 395), such comments could be useful on social media platforms due to their amplifying qualities—potentially generating further engagement and click-throughs to the original story as posted on the news site. Although not necessarily representative, such a view of the audience can nevertheless be seen as less in line with ideals of fostering deliberation through participatory journalism (Domingo et al., 2008) and more in line with a strategic maximization of news amplification.

Finally, for shares, Chadwick and co-authors (2018) emphasize the importance of this mode of news use for the indirect reach of news—as users share the items posted, they spread them to a broader base of potential news users. Researchers have pointed out the complex motivations that appear to be involved in social media sharing. For example, as items shared on Facebook will typically end up featured on one’s own profile page, studies have found that users feel the need to carefully curate their pages, sometimes expressing skepticism toward excessive sharing. As suggested by García-Perdomo, Salaverría, Kilgo, and Harlow (2017) news sharing can cause “reputational damage” (p. 3). Conversely, as other studies have suggested that content characterized by controversial or emotional topics lead to higher degrees of sharing (e.g., Kalsnes & Larsson, 2018; Kümpel, Karnowski, & Keyling, 2015), the practice of sharing news is arguably a complex one.

As for the consequences of news use, we see these activities as related to what Zhang and co-authors (2018) describe as social media amplification. Drawing on examples from political communication, the authors describe amplification as “actions of individual users that intentionally or unintentionally increase measures of engagement surrounding a person, message, or idea” (Zhang et al., 2018, p. 6), suggesting that news use activities as defined above play highly important roles in how social media platforms present content to users, how this content is perceived by users—and thus in serving as a basis for decision-making in the newsroom. News use on Facebook is understood here as having amplifying effects. Indeed, whether perceived by audience members as demanding or not, all forms of news use will serve to trigger the algorithms utilized by Facebook, amplifying the content engaged with to varying degrees. While the algorithmic importance of news use options is largely black-boxed and prone to over-time change (e.g., Rieder, Abdulla, Poell, Woltering, & Zack, 2015), the importance of shares to determine popularity appears to have stabilized over time (Montells, 2017). Nevertheless, our current purposes will detail all three varieties of news use.

The present study will detail news use across three types of media outlets—national, regional, and hyperpartisan. As the bulk of research looking into news use has been performed in relation to national media (Sjøvaag et al., 2018, p. 2), and as different news outlets can be expected to be engaged with in different ways in relation to news use practices (e.g., Kalogeropoulos, Negredo, Picone, & Nielsen, 2017), employing this threefold approach is likely to tease out differences in terms of news user activity. Specifically, though mainstream national media might still be able to attract vast amounts of Facebook followers due to the path dependencies of their audiences (e.g., Webster & Ksiazek, 2012), the degree to which such attention translates into news use and thus amplification of posts remains largely unknown. Looking at the other two types of media outlets included in the study, the role of regional newspapers should not be underestimated in terms of news use. Indeed, as such outlets have functioned as “the main stage for local politics” (Skogerbø & Krumsvik, 2015, p. 353), and have served as a “backbone” or “social glue” for many Norwegian towns, cities, and villages (Sjøvaag et al., 2018, p. 17). Moreover, as regional media actors have been more willing to allow for news use on their own web presences (Almgren & Olsson, 2016), we expect that news use will be performed to higher degrees on the Facebook pages of regional media than on those operated by their national competitors. Our first hypothesis is formulated accordingly:

H1: News use on Facebook will be more pronounced in relation to regional media outlets when compared with national media outlets.

For hyperpartisan media, these outlets could be expected to stir up emotions and post on controversial topics—as mentioned previously, two characteristics that have been pointed to as crucial for sharing success. Presenting a survey of U.K. Twitter users, Chadwick and co-authors (2018) found that two thirds of respondents had shared news that was “not fully accurate, was exaggerated, or was made up” (p. 12). Similarly, in a comparison of news engagement in France and Italy, Fletcher et al. (2018) found that one outlet featuring false or indeed fake news vastly overreached most mainstream news brands by means of Facebook interactions (i.e., the sum of shares, reactions, and comments). Thus, we expect news use in relation to hyperpartisan media on Facebook to be more pronounced when compared with both national and regional media. Our second hypothesis reads as follows:

H2: News use on Facebook will be more pronounced in relation to hyperpartisan media outlets when compared with national and regional media outlets.

Method

National and regional newspapers were identified using the register of Norwegian newspapers with online presences featured in the MediaNorway database (MediaNorway, 2018). This list of 119 regional (86% of the total sample studied) and 14 national newspapers (10%) was combined with six hyperpartisan actors (4%). Actors within this latter category are not featured in the aforementioned data base given that they do not adhere to the publishing guidelines of the Norwegian Press Association. With this in mind, the six included hyperpartisan media outlets were selected based on their prominence in public debate about the quality of news at the time of data collection (e.g., Lindebø, 2018).

Data regarding news use in relation to the Facebook pages operated by these media outlets were collected by means of the Netvizz application, which utilizes the Facebook Application Programming Interface (API) to archive public page content and metadata such as metrics detailing amount of news use (Rieder, 2013). Such data were archived in relation to the 100 latest posts for each of the identified 139 media outlets. Data collection was undertaken on May 1 and May 2, 2018. As no time period was specified for data, the number of posts gathered for each page was within the limits of what Netvizz can accommodate (Rieder et al., 2015). Given the different level of activity undertaken by the studied pages, the time periods covered by these latest 100 posts for each page varied, for some pages reaching back to 2017. Because we are interested in detailing the news use activity undertaken per follower of each page, the procedure described here was complemented by visiting each of the pages and registering the number of Facebook followers for each page.

Analysis was undertaken in two ways. First, taking into account that the number of followers per page varied substantially (M = 29,045.32, SD = 65,186.51, Mdn = 11,739), initial analysis focused on the amount of news use activity undertaken per follower for each of the studied pages. This entailed dividing the amount of news use undertaken with the number of followers for each page. For presentational purposes, the figures featuring these analyses are focused on the top pages in this regard. Although such a focus on the most interacted with outlets could be questioned (e.g., Lorentzen & Nolin, 2017), the selected approach was deemed suitable as it allowed for a clearer emphasis on the news outlets most engaged with in the studied ways (following suggestions by Mahrt & Scharkow, 2013). Moreover, a focus on top outlets allows us to more efficiently assess the popularity of hyperpartisan news outlets in the Norwegian context (as suggested by Fletcher et al., 2018). To move beyond these top Facebook pages and to investigate our hypotheses, Figures 2 to 4 were complemented with a series of statistical analyses. Analysis of variance (ANOVA) tests featuring effect sizes (eta squared, η²) were undertaken for a number of purposes. First, to test for the average number of followers per page across the studied types of media outlets, the ANOVAs were complemented with post hoc testing using Tukey’s honestly significant difference (HSD; as suggested by Pallant, 2001) to see if any significant differences can be uncovered. Second, for our hypotheses regarding the news use data, planned comparisons corresponding to the hypotheses were entered into the ANOVA models by means of the contrasts option in SPSS (as suggested by Pallant, 2001). As the data utilized here were positively skewed, the variables were logarithm transformed to facilitate analysis (as recommended by Field, 2005).

Although these first analyses provided us with the results necessary to evaluate our hypotheses, they did not provide much detail as to the types of news items that succeed in terms of news use. The second part of the results section thus details the most amplified news items across the three categories of media studied here.

Results

To provide an overarching view of the popularity of the studied Facebook pages, Figure 2 details the top pages in terms of number of followers.

Given their role in the everyday lives of many Norwegians, the clear dominance of Facebook presences operated by national news providers among the most liked pages comes as no surprise. What is perhaps a bit more surprising is the influx of certain hyperpartisan actors in this regard as shown in Figure 1. When compared with the followers of regional media outlets (M = 15,912.54, SD = 17,155.86, Mdn = 10,110), hyperpartisan actors appear to do slightly better in terms of how many follow their pages (M = 16,097.67, SD = 12,651.03, Mdn = 12,510)—but not to the extent that pages operated by national news actors enjoy (M = 146,222.93, SD = 161,033.26, Mdn = 84,513). An ANOVA revealed a significant difference between the three types F(2, 136) = 39.01, p < .0001, η² = .365. Following the guidelines suggested by Cohen (1988), the η² could be considered as showing a medium size effect of media type on the amount of followers. Post hoc testing using Tukey’s HSD revealing the mean difference between the national and regional media to be significant (p < .0001) as well as between national and hyperpartisan media (p < .0001)—but not between regional and hyperpartisan media (p = .146). As such, though national media clearly dominate in terms of followers when compared with their regional and hyperpartisan competitors, these initial results suggest that while the latter group did not succeed in securing a significantly larger average followership than their regional competitors, they nevertheless appear to be making an impact among the most followed pages as shown in Figure 2.

Figure 1.

Top 40 most liked pages.

Figure 2.

Reactions activity undertaken per follower.

Although reactions constitute a series of different types of similarly executed, push-button functionalities (e.g., the original “like” button, “Angry,” and “Haha”), Figure 2 presents them in a combined fashion. Although this essentially means conflating some of the unique differences between the different types, our interest here is not necessarily in delving deeper into such discrepancies, but rather to map out the degree to which these functionalities are used in relation to different types of media outlet, contributing to the amplification of their posted content. As shown in Figure 2, Facebook reactions appear to be rather heavily skewed toward the hyperpartisan variety of pages. In fact, while the bulk of the very top pages here belong to this particular category, the very top page in this regard—Rapport-X (with an amount of reactions per follower of 9.2)—enjoys such vast amounts of news use in this regard that we might suspect some sort of automated or bot-related activity at play here. Interestingly, no national news providers emerge among the pages enjoying the most active users with regard to reactions—which is also reflected in their low standing when comparing average amount of reactions per page follower across all categories (national media outlets: M = 0.11, SD = 0.72, Mdn = 0.11; regional: M = 0.28, SD = 0.22, Mdn = 0.20; hyperpartisan: M = 2.32, SD = 3.34, Mdn = 1.17). An ANOVA test revealed a significant effect of media outlet type on the Reactions activity undertaken per follower F(2, 136) = 19.1, p < .0001, η² = .283. Utilizing the aforementioned interpretation guidelines for effect sizes (Cohen, 1988), the reported η² indicates a small to medium effect of media type in relation to this variety of news use. In relation to H1, planned comparisons showed that for the Reactions variety of news use, the means for national and regional media outlets were significantly different from each other (t = 3.43, value of contrast = 0.17, p < .0001). For H2, planned comparisons revealed the mean reported for hyperpartisan media to be significantly higher than the mean reported for their national and regional competitors (t = 4.39, value of contrast = 2.13, p < .0001).

A similar and perhaps even clearer picture of hyperpartisan news user dominance is shown in Figure 3, which details the amount of commenting activity per user across media outlet type. Although the scale employed is boosted to fit the somewhat suspiciously large amount activity undertaken in relation to the aforementioned Rapport-X page, the remainder of commenting represented in Figure 3 is performed at a much smaller rate. Such a finding mirrors the suggestions discussed previously that commenting would be a more difficult type of activity to get news users to actually take part in—especially when compared with reacting, which, as shown in Figure 2, was indeed undertaken to comparably higher degrees. These differences are visible also when comparing the means of commenting activity per follower of hyperpartisan (M = 2.85, SD = 6.61, Mdn = 0.15), national (M = 0.03, SD = 0.02, Mdn = 0.03), and regional (M = 0.04, SD = 0.03, Mdn = 0.03) pages, again demonstrating the prowess of hyperpartisan pages when it comes to sustaining news use and thereby the amplification of their stories. Much like for reactions, an ANOVA performed in the manner described earlier revealed a significant effect of media outlet type on commenting activity undertaken per follower F(2, 136) = 20.23, p < .0001, η² = .172. Following Cohen’s suggestions on the interpretation of η², the presented effect size indicates a small effect of media type on commenting activity as studied here. Furthermore, for H1, planned comparisons analysis showed the mean difference between national and regional media outlets to be nonsignificant (t = 1.18, value of contrast = 0.01, p = .253)—a differing result when compared with that previously reported for the reactions variety of news use, and a result that suggests weakened support for our first hypothesis. Conversely for H2, the mean difference reported for hyperpartisan media outlets in comparison with the other two categories was found to be significant (t = 3.01, value of contrast = 2.13, p = .028), thus further corroborating our second hypothesis.

Figure 3.

Commenting activity undertaken per follower.

If liking and reacting are more easily undertaken than commenting, the practice of sharing news items on Facebook was previously suggested as involving an even more complex series of motivators. This hesitance toward sharing is mirrored in Figure 4, which details the comparably minuscule amount of sharing activity undertaken per follower across the top pages in this regard. Although the scale used here is indeed much smaller than for the previous graphs, the pattern with regard to who gains amplification by means of this scarcely employed variety of news use remains similar. Indeed, the very top emerges again to be dominated by hyperpartisan news outlets, followed by pages operated by primarily regional and then national news providers. Taking all pages into account, the results are reminiscent of those reported for commenting and reacting as hyperpartisan media outlets emerge to succeed in amassing the most amplification per follower in this regard as well (M = 0.41, SD = 0.67, Mdn = 0.14) followed by their regional (M = 0.02, SD = 0.02, Mdn = 0.01) and national (M = 0.01, SD = 0.01, Mdn = 0.01) competitors. ANOVA testing showed a significant effect of media outlet type on the sharing variety of news use undertaken per follower F(2, 136) = 30.22, p < .0001, η² = .274. Similar to the analysis performed for the reactions variety of news use, the η² suggests a small to medium effect of media type on the studied sharing activity. Planned comparisons analysis in relation to H1 revealed that for sharing, the means for national and regional media outlets were not significantly different from each other (t = 1.95, value of contrast = 0.01, p = .070). With regard to H2, however, the mean reported for sharing per follower of hyperpartisan media Facebook pages emerged as significantly higher when compared with the corresponding statistic reported for national or regional media outlets (t = 1.95, value of contrast = 0.01, p = .046).

Figure 4.

Sharing activity undertaken per follower.

We now turn to assess the sum of reported results in relation to our hypotheses. H1 suggested that news use would be more pronounced in relation to regional media outlets when compared with national media outlets. Although regional media outlets emerged with higher means than their national competitors across all three categories of news use, only the planned comparison analysis performed for the reactions variety emerged as significant. Thus, we find only limited support for H1, leading us to reject it. For H2, it was suggested that news use would be more pronounced in relation to hyperpartisan media than in relation to national and regional media. As shown by the means, ANOVAs and planned comparisons featured previously, such differences were indeed reported and found to be statistically significant for all categories of news use. We can thus confirm our second hypothesis.

Although the analyses presented above has provided us with important insights into the apparent and sometimes suspicious dominance of hyperpartisan media outlets on Facebook, these figures say rather little about what types of news items as posted by what types of news actors succeed in reaching comparably higher levels of amplification through the news use practices here. Figures 5 to 7 will provide insights into such issues, presenting activity undertaken in relation to national, regional, and hyperpartisan outlets in separate figures.

Figure 5.

Most popular news for hyperpartisan news outlet pages.

Figure 6.

Most popular news for regional news outlet pages.

Figure 7.

Most popular news for national news outlet pages. ABBA = a Swedish pop group; NOK = Norwegian krone; VG = Verdens Gang, Norwegian tabloid; WWII = World War II.

Figures 5 to 7 all feature a series of nodes, and each one of these nodes represent a specific post made by a specific actor to their respective Facebook pages. Node placement and size in these figures is dependent on the amount of news use it had amassed at the time of data collection. Specifically, placement on the horizontal axis details the amount of comments received, whereas orientation on the vertical axis suggests the amount shares performed in relation to each post. Node size, then, is dependent on the degree to which the corresponding post had been reacted to as described above—the bigger the node, the more reactions. In the figures, some nodes have been identified that had succeeded in gaining especially large amounts of traction in relation to these metrics. These nodes are provided with explanations and summaries of what the corresponding post was about.

Starting with the hyperpartisan media outlets in our sample, it appears as clear from the captions provided to the outlier popular posts in Figure 5 that asking for news user activity will result in such activity. A series of identified posts are formulated as questions directed to the reader, asking for their feedback, or as urges to share posts that show support for certain causes. Thematically, we see issues pertaining to immigration and Islam throughout the figure—clearly topics that are very much in line with the right-wing agenda adhered to by many of these outlets (Fletcher et al., 2018) and that one would thus suspect to resonate well with their follower base.

Taking the differences of scale into account, when comparing with the activity undertaken in relation to hyperpartisan pages in Figures 5 and 6, details news use in relation to regional news outlet Facebook pages. Although the tendency is not as clear as for the hyperpartisan variety, a few of the news items that emerged in Figure 6 as especially popular nevertheless have the same tendency to ask for news users to spring into action—whether through opportunities to win patio heaters, dinners or festival tickets, or by means of asking for page visitor thoughts on dog owners who do not clean up after themselves. As such, though the themes dealt with are obviously different when comparing hyperpartisan with regional outlets, the techniques employed to gain attention—and thus amplification—nevertheless emerge as similar.

Moving on to identify the items posted that succeeded to raise comparably higher amounts of news use for national news outlets, the posts detailed in Figure 7 are to a lesser extent reminiscent of the tendency identified for the previous two varieties of news outlets. Although a few of these types of call-to-action type news items are indeed visible (e.g., inviting visitors to compete for movie premiere tickets, or for them to provide their opinions on the suggested ban in Islamic calls to worship), most of the identified popular posts as comparably more popular are of what we might refer to as a more traditional journalistic variety, reporting on current affairs, celebrity deaths, and featuring op-ed pieces from societal actors.

Taken together, then, it would appear when comparing all three types of outlets that an activating mode of authoring posts—encouraging for news use to take place in some way, shape or form—is likely to result in heightened levels of such activity. Not very surprising, perhaps—but this and other tendencies uncovered above could be expected to have implications for journalistic and news user practice, implications that will be discussed in the final and concluding section of the article.

Discussion

Comparing news use across national, regional, and hyperpartisan news outlets on Facebook, this study has found that though the former of these three actors are clearly in the lead when it comes to the number of Facebook followers, hyperpartisan actors emerge as more successful in terms of the degree to which their followers engage as news users, amplifying their content. Although some foul play in terms of automated intervention was suspected—consider the surge of activity undertaken in relation to the hyperpartisan Rapport-X page—statistical analyses taking such outliers into account nevertheless show us the dominance of hyperpartisan media in terms of engaging and activating their follower base—resulting in the amplification of the perspectives purveyed by these pages. As such, the tendencies visible in other parts of the world for hyperpartisan actors such as these to be on the rise are clearly visible also in Norway. Although the data employed here do not allow for any deeper insights into whether part of the traffic studied was automated or not, the results presented do raise a few other issues which will be discussed below.

Beyond the overall result that hyperpartisan actors emerged as more successful in terms of news use among their Facebook followers, a key to further understanding these developments could be to take a closer look at the types of posts provided by these actors as well as their regional and national competitors that seemed to have reached especially high levels of amplification. Plenty of these most successful posts have what could be described as a mobilizing tone in common. As employed within the study of political communication, a mobilizing tone or indeed strategy often involves political actors using their platforms to try and generate feedback and activity from their potential voters rather than simply informing about political prioritizations or policy issues (e.g., Foot & Schneider, 2006; Graham, Broersma, Hazelhoff, & van’t Haar, 2013; Lilleker et al., 2011). Such prowess for mobilizing content at the hands of primarily hyperpartisan actors could be seen as taken right out of the proverbial playbook suggested by network logic (Klinger & Svensson, 2015) or indeed social media logic (van Dijck & Poell, 2013). Following these concepts, the topic to be covered needs to have a certain degree of “shareability” (García-Perdomo et al., 2017; Trilling, Tolochko, & Burscher, 2017) in addition to it being provided for the news user in a way that makes news use and thus amplification easy to undertake—framed as a question, a call to action or indeed a competition. As such, these ways of phrasing and presenting news could be the way forward for an industry struggling to make sense of how to handle their social media intermediaries—with certain repercussions.

Indeed, while it is perhaps not so surprising to find these rhetorical forms emanating from the types of hyperpartisan actors studied here, the comparable popularity of posts put together in such ways invites further reflection on the ever-tightening competition for news user attention. As both industry and academic perspectives have pointed to the increasing importance of news use and amplification on social media such as the processes studied here for the future of news provision, the insights into the preferences of the “networked audience” (Marwick & boyd, 2011) that have been presented are important if we want to trace future tendencies of what news stories get selected for placement on Facebook—and how those stories are framed and presented to the news users to maximize amplification. This is not to say that we should expect legacy media to start behaving and publishing in the same vein that their hyperpartisan competitors do any time soon. Rather, the suggestion is made here that these developments and the results presented here make it necessary for us to be acutely aware of the influences that structural aspects and competition for attention have had on news providers in the past—and that are thus likely to yield influence again. Granted, though scholars have suggested that the influences of such “quantified audiences” (Anderson, 2011) are not as steadfast as we might assume—at least not when it comes the home pages of news organizations (as suggested by Zamith, 2016)—other researchers have nevertheless identified how patterns of audience engagement or indeed news use such as the variety studied here has yielded influence over editorial decisions of various kinds (e.g., Lee, Lewis, & Powers, 2012; Tandoc, 2014; Vu, 2014; Welbers, van Atteveldt, Kleinnijenhuis, Ruigrok, & Schaper, 2016). Thus, it would seem as if the entry of social media into the publishing fray has contributed to a strengthening of the audience in this relationship, further affecting the balance upkept by reporters in giving people not only the news they need—but also the news they want (Gans, 1980). To this classic dualism, we might now add a third criteria, as we can expect editorial staff to also increasingly take into account the expected degree to which a news item will be used in the ways studied here—does the item have the potential to gain amplification? Somewhat reminiscent of the market-based model of news production traced by McManus (1995) since at least the 1980s, we might expect the ability to trace news user preferences on social media such as has been performed here to further strengthen the influence of metrics over editorial decisions.

Although some of these suggestions might seem quite a distance away from the media environment of today, it is important to remember that journalistic formats and work practices have indeed been influenced by novel actors and techniques previously. Indeed, tabloidization is sometimes suggested as having changed both the format and the content of journalism (Bird, 1998; Esser, 1999), and the first online varieties of journalism saw media personnel changing and adapting their texts to fit the new platform (Powers, 2011). As social media such as the platform studied here are likely to become more and more important for news distribution purposes, future research might find it fruitful to see if the perspectives and tendencies identified here are indeed tangible in an over-time research setting. Such diachronic or even longitudinal research designs could help us gain useful knowledge into the prioritizations of news providers—be they national, regional, or hyperpartisan—as they attempt to navigate the ever changing online landscape increasingly formatted by digital intermediaries such as Facebook.

Moreover, researchers might also find it useful to draw on the perspectives raised here and study the amplifying activities of news users in different ways. Indeed, the material presented in the article at hand raises a series of questions worthy of further inquiry. For example, to what extent do these users understand the amplifying role they have to play as they engage with certain news posts while leaving other posts behind? As we know from previous research that certain types of content tend to be shared and commented to comparably higher degrees than other types (e.g., Kalsnes & Larsson, 2018; Kümpel et al., 2015), researchers interested in the issues studied here might find it useful to engage with Facebook posts in a more large-scale, quantitative content analysis fashion. The graphs presented in Figures 5 to 7 do indeed give us useful insights into the most amplified or indeed viral (e.g., Nahon & Hemsley, 2013) posts, but studying the remainder of posts in the suggested fashion might provide further advantageous understandings. Relatedly we might ask if these amplifying activities are undertaken in a more haphazard fashion, or if we can identify more organized varieties of amplification through news use, perhaps akin to what Bennett and Segerberg (2012) have referred to as connective action? Moreover, given the apparent influx of automated activity in the material studied here, to what extent are news users—or indeed the media outlets studied here—aware of such activities? Looking beyond Facebook, future research should look into these issues as they are structured and shaped on other platforms. Twitter, for example, is often viewed as a highly politicized platform in the Norwegian context, and could thus make for an interesting opportunity to further study and understand these issues.

Finally, beyond the study of platforms per se, further insights into the journalistic profession and how its representatives navigate the tendencies uncovered here could be useful. For example, Powers (2011) point out that while nearly every technological development has eventually been adapted to journalistic practice, such processes have nevertheless been characterized by what Zelizer (1995) has referred to as “discourses of resistance” as novel work forms were being introduced. Keeping such historical instances in mind, researchers might find it fruitful to detail the ways that varieties of mobilizing journalism—journalism that apparently works well for amplifying purposes—are discussed in newsrooms and among media personnel more broadly. Given the success of hyperpartisan media actors demonstrated in this article, it would be interesting to gauge the prioritizations of those novel types of actors—and the prioritizations of their mainstream competitors.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

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

ORCID iD

Anders Olof Larsson

Author Biography

Anders Olof Larsson (PhD, Uppsala University, 2012) is professor at Kristiania University College. He was previously a postdoctoral fellow working in the Social media and Agenda-Setting in election campaigns project at the Department of Media and Communication, University of Oslo. During his PhD work, which was funded by The Swedish Research School of Management and Information Technology, he took part in the 2011 Oxford Internet Institute Summer Doctoral Program. His PhD thesis—“Doing Things in Relation to Machines—Studies on Online Interactivity”—was awarded the 2012 Börje Langefors award (for best Swedish thesis within the field of Informatics) and the 2012 FSMK Doctoral Dissertation award (for best Swedish thesis within the field of media and communication studies). His research interests include the use of online interactivity and social media by societal institutions and their audiences, journalism studies, political communication and methodology, especially quantitative methods.

References

Almgren

S. M.

Olsson

(2016). Commenting, sharing and tweeting news: Measuring online news participation. Nordicom Review, 37(2), 1-15.

Anderson

C. W.

(2011). Between creative and quantified audiences: Web metrics and changing patterns of newswork in local U.S. newsrooms. Journalism: Theory, Practice and Criticism, 12, 550-566.

Bennett

W. L.

Segerberg

(2012). The logic of connective action. Information, Communication & Society, 15, 739-768.

Bergström

Wadbring

(2015). Beneficial yet crappy: Journalists and audiences on obstacles and opportunities in reader comments. European Journal of Communication, 30, 137-151.

Bird

S. E.

(1998). News we can use: An audience perspective on the tabloidisation of news in the United States. Javnost—The Public, 5(3), 33-49.

Braun

Gillespie

(2011). Hosting the public discourse, hosting the public. Journalism Practice, 5, 383-398.

Bruns

(2010). From reader to writer: Citizen journalism as news produsage. In Hunsinger

Klastrup

Allen

(Eds.), Internet research handbook (pp. 119-134). Dordrecht, The Netherlands: Springer.

Bruns

(2018). Gatewatching and news curation: Journalism, social media, and the public sphere. New York, NY: Peter Lang.

Chadwick

Vaccari

O’Loughlin

(2018). Do tabloids poison the well of social media? Explaining democratically dysfunctional news sharing. New Media & Society, 20, 4255-4274. doi:10.1177/1461444818769689

10.

Chung

D. S.

(2008). Interactive features of online newspapers: Identifying patterns and predicting use of engaged readers. Journal of Computer-Mediated Communication, 13, 658-679.

11.

Cohen

(1988). Statistical power analysis for the behavioral sciences. Hillsdale, NJ: Lawrence Erlbaum.

12.

Deuze

Bruns

Neuberger

(2007). Preparing for an age of participatory news. Journalism Practice, 1, 322-338.

13.

Domingo

Quandt

Heinonen

Paulussen

Singer

J. B.

Vujnovic

(2008). Participatory journalism practices in the media and beyond. Journalism Practice, 2, 326-342.

14.

Esser

(1999). “Tabloidization” of news—A comparative analysis of Anglo-American and German press journalism. European Journal of Communication, 14, 291-324.

15.

Field

(2005). Discovering statistics using SPSS: (And sex, drugs and rock ‛n’ roll). London, England: SAGE.

16.

Fletcher

Cornia

Graves

Nielsen

R. K.

(2018). Measuring the reach of “fake news” and online disinformation in Europe (Fact sheets). Oxford, UK: Reuters Institute for the Study of Journalism.

17.

Foot

K. A.

Schneider

S. M.

(2006). Web campaigning. Cambridge, MA: MIT Press.

18.

Gans

H. J.

(1980). Deciding what’s news: A study of CBS evening news, NBC nightly news, Newsweek, and Time. New York, NY: Vintage Books.

19.

García-Perdomo

Salaverría

Kilgo

D. K.

Harlow

(2017). To share or not to share. Journalism Studies, 19, 1180-1201.

20.

Graham

Broersma

Hazelhoff

van’t Haar

(2013). Between broadcasting political messages and interacting with voters. Information, Communication & Society, 16, 692-716.

21.

Hermida

Thurman

(2008). A clash of cultures: The integration of user-generated content within professional journalistic frameworks at British newspaper websites. Journalism Practice, 2, 343-356.

22.

Ihlebaek

K. A.

Krumsvik

A. H.

(2015). Editorial power and public participation in online newspapers. Journalism: Theory, Practice and Criticism, 16, 470-487.

23.

Ipsos/MMI. (2018). Ipsos SoMe-tracker Q3’18. Retrieved from https://www.ipsos.com/nb-no/ipsos-some-tracker-q318

24.

Kalogeropoulos

Negredo

Picone

Nielsen

R. K.

(2017). Who shares and comments on news? A cross-national comparative analysis of online and social media participation. Social Media + Society, 3(4), 1-12. doi:10.1177/2056305117735754

25.

Kalsnes

(2016). The power of likes: Social media logic and political communication. Oslo, Norway: University of Oslo.

26.

Kalsnes

Larsson

A. O.

(2018). Understanding news sharing across social media. Journalism Studies, 19, 1669-1688.

27.

Klinger

Svensson

(2015). The emergence of network media logic in political communication: A theoretical approach. New Media & Society, 17, 1241-1257.

28.

Krumsvik

A. H.

(2017). Redefining user involvement in digital news media. Journalism Practice, 12, 19-31.

29.

Kümpel

A. S.

Karnowski

Keyling

(2015). News sharing in social media: A review of current research on news sharing users, content, and networks. Social Media + Society, 1(2), 1-14.

30.

Larsson

A. O.

(2012). Understanding nonuse of interactivity in online newspapers: Insights from structuration theory. The Information Society, 28, 253-263.

31.

Lee

A. M.

Lewis

S. C.

Powers

(2012). Audience clicks and news placement: A study of time-lagged influence in online journalism. Communication Research, 41, 505-530.

32.

Lilleker

D. G.

Koc-Michalska

Schweitzer

E. J.

Jacunski

Jackson

Vedel

(2011). Informing, engaging, mobilizing or interacting: Searching for a European model of web campaigning. European Journal of Communication, 26, 195-213.

33.

Lindebø

(2018, March 16). Nå skal Resett, HRS og Radikal Portal holdes utenfor PFU [Resett, HRS and Radikal Portal to be kept out of the Norwegian press ethics board]. Journalisten. Retrieved from https://journalisten.no/2018/03/na-skal-resett-holdes-utenfor-pfu

34.

Lorentzen

D. G.

Nolin

(2017). Approaching completeness: Capturing a hashtagged Twitter conversation and its follow-on conversation. Social Science Computer Review, 35, 277-286.

35.

Mahrt

Scharkow

(2013). The value of big data in digital media research. Journal of Broadcasting & Electronic Media, 57, 20-33.

36.

Marwick

boyd

(2011). To see and be seen: Celebrity practice on Twitter. Convergence: The International Journal of Research into New Media Technologies, 17, 139-158.

37.

McManus

(1995). A market-based model of news production. Communication Theory, 5, 301-338.

38.

MediaNorway. (2018). 2017 readership statistics for Norwegian online newspapers. Retrieved from http://www.medienorge.uib.no/statistikk/medium/avis/253

39.

Montells

(2017). What is Facebook EdgeRank or how Facebook algorithm works. Retrieved from https://metricool.com/what-is-facebook-edgerank-or-how-facebook-algorithm-works/

40.

Nahon

Hemsley

(2013). Going viral. Cambridge, UK: Polity Press.

41.

Nielsen

R. K.

Ganter

S. A.

(2018). Dealing with digital intermediaries: A case study of the relations between publishers and platforms. New Media & Society, 20, 1600-1617.

42.

Nightingale

(2011). The handbook of media audiences. Malden, MA: Wiley-Blackwell.

43.

Pallant

(2001). SPSS survival manual: A step by step guide to data analysis using SPSS for Windows (Version 10 and 11). Buckingham, UK: Open University Press.

44.

Papacharissi

(2002). The virtual sphere: The internet as a public sphere. New Media & Society, 4, 9-27.

45.

Picone

(2016). Grasping the digital news user. Digital Journalism, 4, 125-141.

46.

Picone

(2017). Conceptualizing media users across media. Convergence: The International Journal of Research into New Media Technologies, 23, 378-390.

47.

Powers

(2011). “In forms that are familiar and yet-to-be invented”: American journalism and the discourse of technologically specific work. Journal of Communication Inquiry, 36, 24-43.

48.

Rieder

(2013, May 2-4). Studying Facebook via data extraction: The Netvizz application. Paper Presented at the WebSci’ 13 Conference, Paris, France.

49.

Rieder

Abdulla

Poell

Woltering

Zack

(2015). Data critique and analytical opportunities for very large Facebook pages: Lessons learned from exploring “we are all Khaled said.” Big Data & Society, 2(2), 1-22.

50.

Rosen

(2006). The people formerly known as the audience. Retrieved from http://archive.pressthink.org/2006/06/27/ppl_frmr.html

51.

Singer

J. B.

Domingo

Heinonen

Hermida

Paulussen

Quandt

. . . Vujnovic

(2011). Participatory journalism: Guarding open gates at online newspapers. Chichester, UK: John Wiley.

52.

Sjøvaag

Stavelin

Karlsson

Kammer

(2018). The hyperlinked Scandinavian news ecology: The unequal terms forged by the structural properties of digitalisation. Digital Journalism. Advance online publication. doi:10.1080/21670811.2018.1454335

53.

Skogerbø

Krumsvik

A. H.

(2015). Newspapers, Facebook and Twitter: Intermedial agenda setting in local election campaigns. Journalism Practice, 9, 350-366.

54.

Tandoc

E. C.

(2014). Journalism is twerking? How web analytics is changing the process of gatekeeping. New Media & Society, 16, 559-575.

55.

Tandoc

E. C.

Lim

Z. W.

Ling

(2017). Defining “fake news.” Digital Journalism, 6, 137-153.

56.

Trilling

Tolochko

Burscher

(2017). From newsworthiness to shareworthiness: How to predict news sharing based on article characteristics. Journalism & Mass Communication Quarterly, 94, 38-60.

57.

Tuchman

(1978). Making news: A study in the construction of reality. New York, NY: Free Press.

58.

Vaage

O. F.

(2018). Norsk mediebarometer 2017 [Norwegian Media Barometer 2017]. Oslo: Statistics Norway.

59.

van Dijck

Poell

. (2013). Understanding social media logic. Media and Communication, 1, 2-14.

60.

H. T.

(2014). The online audience as gatekeeper: The influence of reader metrics on news editorial selection. Journalism, 15, 1094-1110.

61.

Webster

J. G.

Ksiazek

T. B.

(2012). The dynamics of audience fragmentation: Public attention in an age of digital media. Journal of Communication, 62, 39-56.

62.

Welbers

van Atteveldt

Kleinnijenhuis

Ruigrok

Schaper

(2016). News selection criteria in the digital age: Professional norms versus online audience metrics. Journalism: Theory, Practice and Criticism, 17, 1037-1053.

63.

Zamith

(2016). On metrics-driven homepages. Journalism Studies, 19, 1116-1137.

64.

Zelizer

(1995). Journalism’s “last” stand: Wirephoto and the discourse of resistance. Journal of Communication, 45, 78-92.

65.

Zhang

Wells

Wang

Rohe

(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, 3161-3182. doi:10.1177/1461444817744390