Twitter’s influence on news judgment: An experiment among journalists

Abstract

The literature suggests that journalists give a substantial amount of attention to Twitter and use the platform widely, but the impact of that use on news judgment has not been assessed. We hypothesize that Twitter affects journalists’ news judgment, impacting coverage decisions. To test this, we conducted an online survey experiment on working US journalists (N = 212). We find that journalists who use Twitter less in their work discount news they see on the platform, potentially causing them to dismiss information that many of their colleagues identify as newsworthy. Our results also indicate that the routinization of Twitter into news production affects news judgment – for journalists who incorporate Twitter into their reporting routines, and those with fewer years of experience, Twitter has become so normalized that tweets were deemed equally newsworthy as headlines appearing to be from the AP wire. This may have negative implications, such as pack journalism, but we also see positives, as Twitter may conduit a wider array of voices into the mainstream news agenda. Twitter plays a key role in journalistic practices including, as we demonstrate here, influencing journalists’ news judgment. Twitter’s growing centrality in the news process warrants greater scrutiny from journalists and scholars.

Keywords

Digital journalism experiment news judgment newsworthiness Twitter

When Chris Christie was pictured staring uncomfortably at Donald Trump at a 2016 press event for the Republican candidate (and eventual president), some on Twitter made light of his dour look. As summarized in a Poynter.org article (Mantzarlis, 2016), the joke took off, and by the end of the day, Christie was the top trending topic on Twitter. News media, including USA Today, wrote about Christie’s gaze, interpreting it as a sign of a conflicted relationship with Trump. This despite the fact that, as Mantzarlis notes, those who cover Christie regularly said he ‘always’ has that look when standing behind someone at a podium.

Journalists may not always jump on what’s trending on Twitter, but examples such as this illustrate how the microblogging platform regularly exerts some influence on news coverage. Journalists use Twitter widely (Armstrong and Gao, 2010; Holton and Lewis, 2011; Lasorsa et al., 2012; Molyneux, 2015; Vis, 2013) and many are required by their employers to be active on Twitter, but while Twitter’s incorporation into reporting routines has been documented, the effects of their use of the social media platform on news judgment has not yet been assessed. Twitter, like most social media, represents a unique communication sphere where the media, political actors, and the public are all present. Unlike previous media, messages can flow to and from each of the actors. In a pre-Twitter era of news production, it is unlikely that journalists would spend much of their day connected to and at times talking with journalists at other news outlets and with the public. Twitter allows for a sort of virtual collocation, which journalists use to construct narratives and engage in community discourses (Mourão, 2014). Given that news production is often shaped by the daily routines of and influences on journalists (Anderson, 2009; Reese, 2001; Schudson, 1982), it follows that Twitter may have some effect on journalists’ news judgment. The purpose of this experiment is to assess the effect of Twitter on journalists’ news judgment.

Understanding the role journalists’ use of Twitter plays in news judgment and gatekeeping will shed light on a routine practice that is understudied in the literature. When online news sites first rose to prominence, a set of scholarly studies examined how new expectations and norms relating to online news shifted (or didn’t) the practices of journalists, including their gatekeeping function (Bruns, 2008; Gillmor, 2004; Robinson, 2007; Singer, 2006; Williams and Delli Carpini, 2000). Likewise, some studies have taken up the journalists’ use of Twitter, but mainly through examining the content of their tweets, with findings focused on reshaping journalistic norms (Armstrong and Gao, 2010; Holton and Lewis, 2011; Lasorsa et al., 2012; Lawrence et al., 2014; Molyneux, 2015; Mourão, 2014; Vis, 2013). Some scholars have ventured behind tweet content to examine journalists’ use of various digital media practices as part of their production routines (see Anderson, 2013; Bélair Gagnon, 2013; Usher, 2014), none have examined the use of Twitter specifically for its impact on news judgment. Nor have scholars attempted to assess the extent to which the routinization of Twitter into news production affects journalists’ news judgment. Early work on social media hinted at the potentially disruptive aspects of new media forms, while others, especially more recently, have noted that the news media and journalists have normalized, to an extent, the platforms (Lasorsa et al., 2012; Lawrence et al., 2014). Recognizing that the series of routines and influences on news judgment now include social media like Twitter, this study addresses these gaps in the literature.

Using an experiment, this study tests the influence of tweets on journalists’ news judgment. Via an online survey experiment, journalists were asked to rate the newsworthiness of stories that appeared to be from the AP wire or from Twitter. The working hypothesis, which we build to in the following sections, is that journalists pay such attention to Twitter (as part of their daily routine) that it influences their news judgment. That is, journalists will value story ideas gathered from Twitter above those gathered in more traditional ways.

How information becomes news: Gatekeeping, news judgment, and newsworthiness

The concept of gatekeeping, as it is understood today, has to do with two things: people and content. As it relates to people, gatekeeping refers to asserting control and defining boundaries – in this case, between professional journalists and everyone else. Spoken of in this sense, gatekeeping is the process by which journalists preserve for themselves the authority to determine and produce the news. This authority faces incursions from citizen journalism (Lewis et al., 2010), social networks of users that can also serve a gatekeeping role (Coddington and Holton, 2014; Meraz and Papacharissi, 2013), and characterizations of professional journalism as ‘fake news’ (Nielsen and Graves, 2017). Separating professional journalism from these other forces is a form of boundary work, a process particularly prominent in Western journalism today (Carlson, 2015).

This study, however, focuses on a second aspect of gatekeeping, referring to the decisions that journalists make about what content to let through the gates. In this sense, scholars have described gatekeeping as deciding what makes it ‘into’ the news (Shoemaker and Vos, 2009). This form of gatekeeping is critical in shaping public knowledge about events or issues (DeFleur, 2010). Scholars have long sought to understand the array of influences on journalists’ decisions about what to let pass through the gates. The hierarchical model of newsmaking (Shoemaker and Reese, 1996, 2014) examines how ‘the message, or media content, is influenced by a wide variety of factors both inside and outside of media organization’ (p. 9), ranging from macro to micro-level influences. Research has also examined the interactions between journalists, as well as organizational factors and daily routines, and their effects on the content journalists produce (Singer, 2006; Tuchman, 1972). Now, journalists’ news judgment is additionally influenced by the use of web metrics in determining story production and placement (Anderson, 2011; Coddington, 2015).

If gatekeeping of content is the process by which some is selected for publication and some is not, then news judgment is the filter that is applied in making these decisions. A journalist’s news judgment may be the greatest predictor of whether a story passes through from journalists to the public (Shoemaker et al., 2001). In fact, according to journalists themselves, news judgment is the single most important professional skill separating them from everyone else. The term ‘news judgment’ is often used with the adjectives ‘professional’ or ‘expert’ in order to convey this. Journalists speaking about documents published via WikiLeaks distinguished their work from ‘information dumps’ by pointing out that they (and only they) were able to apply news judgment to the information before publishing it (Coddington, 2014). And in the case of journalists practicing news aggregation, deciding which stories to aggregate – that is, applying their news judgment – is what made the practice journalistic. This news judgment was seen as the most important skill an aggregator might have (Anderson, 2013). In summary, the process of passing through the filter of a journalist’s news judgment is what turns ‘information’ into ‘news’ (Karlsson and Clerwall, 2013).

Unfortunately, it has been notoriously difficult for scholars (or even journalists themselves) to define news judgment or clarify its constituent parts. Instead, it is treated as an assumed or taken-for-granted characteristic of the professional journalist. It is presumed to be developed through experience, or ‘long-term proximity to news events and actors’ (Coddington, 2014), and thus more experienced journalists have a more developed or more finely tuned sense of news judgment than newer journalists (Schultz, 2007). Scholars have also presumed that this sense of news judgment is developed through social control (Breed, 1955) and other forms of peer pressure, effectively learning by osmosis what is news. Still other scholars have attempted to identify particular news values (Gans, 1979), routines (Tuchman, 1972), and qualities (such as those often found in journalism textbooks: proximity, timeliness, and impact) that help identify what news is. In all these cases, however, what news is and what it is not is left up to the journalists, positioning them as the only ones capable of exercising news judgment.

The key point for this study is that a journalist’s news judgment is the filter that determines which bits of information are worthy of becoming news. That is, applying news judgment is a way of determining an item’s ‘newsworthiness’ in the eyes of journalists. Given the opaque nature of news judgment, this study cannot specify how journalists should arrive at that determination, or even which criteria to draw upon in making it. Instead, we ask journalists to apply their own news judgment in deciding whether an item is newsworthy. Recognizing that this question depends upon an array of influences, it becomes more complicated in the age of social media, which journalists now rely on extensively in their work.

Conceptualizing the role of Twitter in gatekeeping, news judgment, and newsworthiness

Some scholars have suggested that a network of users can serve a gatekeeping role, enabling people to get news through their network connections rather than from journalists directly (Coddington and Holton, 2014; Meraz and Papacharissi, 2013). Even so, journalists still enact a gatekeeping role, which shapes public knowledge about events or issues (DeFleur, 2010). If, as discussed above, gatekeeping is about what content passes through the press’ gates to the public via news stories, this process is clearly visible on Twitter. Journalists’ routine use of Twitter as a part of their work impacts what information and whose voices pass through the mythic gates.

First, Twitter is one place journalists look to be aware of what is happening, or in other words, what is news. Twitter can serve as a sort of early warning system for journalists, alerting them to stories under the mainstream news radar (Hermida, 2010). Journalists have described Twitter as essential to their work, a new and indispensable news wire (Lawrence, 2015). Twitter is not only a place to break news, but a place to spot breaking news (Hamby, 2013). Many newsrooms now subscribe to services such as DataMinr,¹ whose proprietary algorithm promises to create early alerts for breaking news by harnessing Twitter data. During the course of the 2016 US election, journalists relied heavily on DataMinr to identify breaking political news (McGregor, 2018). Journalists also know what is happening on Twitter in part because they keep tabs on the competition by engaging predominantly with other journalists (Molyneux, 2015; Molyneux and Mourão, 2017; Mourão et al., 2016). Simply put, Twitter plays a ‘critical role in the news production process’ (Usher et al., 2018) by making journalists aware of information that might become news.

Journalists’ use of Twitter also impacts whose voices are featured in the news. First, journalists often turn to Twitter to report on or even embed tweets from elites and leaders, including President Trump. Journalists have always heavily featured elite perspectives in news stories (e.g. Bennett, 1990), but Twitter offers a new, simpler means by which to access them. Some research suggests that journalists’ use of Twitter may provide a pathway to expand the use of non-elite sources in news (Hermida et al., 2014). Still, citizens are not as often featured in the news, and when they are, they are most often represented passively (Lewis et al., 2005). One common way that the public is heard in the news is through vox pops, or so-called ‘man on the street’ interviews. As the media historian Heidi Tworek demonstrates, tweets are the new vox populi (Tworek, 2018, see also Beckers and Harder, 2016; McGregor, 2018). But this has not necessarily improved or expanded the public’s representation in the news. For instance, more than 30 outlets, including NPR, The Washington Post, and Buzzfeed, embedded tweets from the Russian disinformation group the Internet Research Agency (Lukito and Wells, 2018) between 2015 and 2017. To varying extents, journalists trust enough what they see on Twitter to include it in their reporting, in a multitude of forms.

Journalists’ reliance on Twitter for news awareness and news sources raises the question of what role the site might play in news judgment. Journalists report checking Twitter first thing in the morning (Hamby, 2013), using services like DataMinr to constantly monitor activity on the site (McGregor, 2018), and seeing it as an always-on ambient source for news (Hermida, 2010). Clearly, Twitter may now be counted among the array of influences on journalists’ news judgment, but research has yet to understand how this influence compares to more traditional forms of determining newsworthiness. Considering existing research on how journalists use Twitter, this study suggests that tweets may serve as a heuristic for journalists – that is, journalists’ own Twitter timelines may serve as a shortcut to determine what is newsworthy.

Both Twitter and traditional news wires, such as that offered by The Associated Press, offer streams of information from which journalists might gather ideas and against which they may compare their own news judgment as they seek validation of their choices (Donsbach, 2004). But it is important to note some key differences in how these services might be used in news production. Journalists have said that they use Twitter as a way to monitor two constituencies: the public and competing journalists (Lawrence, 2015). When monitoring what members of the public tweet, journalists may be looking for items that deserve coverage, but most members of the public don’t tweet news headlines such as the type used in this experiment. More often, content of this nature would come from other journalists, causing a journalist to consider whether that story might also appeal to her audience. When evaluating a tweeted headline, journalists might use not only their own news judgment but also the wisdom of the crowd – how many times the tweet has been favorited or retweeted, how many followers the Twitter user has, and so on. This extra information about how well the information may travel online is invaluable in a media landscape where attention is scarce. Finally, Twitter is arguably the backbone of an online news system in which updates come by the second. Live-blogging gave way to live-tweeting, which has now given way to live-streaming, all presented and distributed through Twitter. This sense of immediacy has long been a core value of journalism, and journalists have eagerly reported on what is trending on Twitter. While journalists might also look through newswire services such as The Associated Press to see what other journalists deem important, it is reasonable to expect that they would interpret headlines viewed there differently than those they view on Twitter – without the same sense of urgency and online momentum. This difference is the focus of our study.

Hypotheses

Given journalists’ heavy use of Twitter, as one part of evolving journalistic routines, we hypothesize that tweets will affect their news judgment.

H1: Journalists will rate news from tweets to be significantly more newsworthy than news from a traditional AP wire.

H2: The more a journalist relies on Twitter, the higher they will rate news from tweets as newsworthy.

Methods

We used an online questionnaire to conduct a 2 × 2 experimental design (story set × presentation format). US journalists above the age of 18 participated in the study (N = 212). The data were collected between 18 and 25 March 2016.

Participants

Participants in the experiment were recruited using email invitations. The sample was drawn using Cision, a media contact service whose database contains at least 1.5 million media professionals worldwide. The database was searched for editorial employees (those whose job descriptions contained at least one of these words: journalist, writer, reporter, contributor, correspondent, editor) working at two types of media: newspapers with circulations higher than 50,000 and news websites that attracted more than 100,000 unique monthly visitors. Since this study focuses on news judgment, these cutoffs were designed to capture news organizations with wide audiences that might exercise news judgment over a high volume of news pieces daily. This search yielded 12,319 journalists working at larger newspapers and 4210 journalists working at larger news websites. Using a random start point and a skip interval of 5, a stratified sample of 3306 of these journalists was selected, including 2464 newspaper journalists and 842 website journalists. These journalists were sent an email invitation to participate, offering as an incentive the opportunity to participate in a drawing for an Apple Watch.

The sample was diverse in terms of years in the news media, ranging from 2 to 50 (M = 22.7, SD = 12.3). In today’s multiplatform media landscape, 99 percent said content produced by their organization appears online, 92 percent said it also appears in print; 18 percent said content from their organization appears on the radio, and 17 percent on television. Finally, 93 percent of the sample reported having a Twitter account.

Procedure

Participants selected for recruitment were sent an email introducing the study and asking for their participation. The email contained a link to the Qualtrics survey. After providing consent, participants were randomly assigned to view two sets of 10 news headlines each. These headlines were derived from actual news headlines on the AP wire but were altered so as not to reflect actual events. In the control group, headlines for Set 1 and Set 2 both appeared in the style of APexchange.com, the website where The Associated Press publishes wire content. In the experimental group, headlines in Set 1 were presented in the AP format, while headlines in Set 2 were manipulated so that they appeared as anonymous tweets. Respondents viewing the Twitter manipulation were told to ‘imagine these tweets are from people you follow and appeared in your Twitter timeline’.

The 20 headlines represented seven categories of news (as defined by the AP): national (6), international (4), economy (2), politics (2), sports (2), celebrity (2), and weather (2). Stories in each of these categories were divided evenly between the two story sets. These story sets were pilot tested on a small sample of professional journalists. The pilot test revealed that one set was marginally more newsworthy than the other, and so two stories swapped places so that each set of 10 stories was equally newsworthy (t(18) = .0003, p = .999). See Appendix 1 for the stimuli.

Participants were first shown both sets of stories in the designated format (all AP, or AP and Twitter). They then were presented each headline individually, followed by a series of questions measuring the dependent variable – newsworthiness. After completing the newsworthiness questions for both sets of headlines, a final block of questions asked about the participants’ experience in journalism, the type of outlet they work for, and their use of Twitter.

Measures

Newsworthiness

Three items were used to measure the newsworthiness of each headline. The first item asked journalists to rate each story on a 10-point scale where 1 = not newsworthy at all and 10 = very newsworthy. The journalists were prompted to answer two questions on a 10-point scale where 1 = not at all and 10 = very much so. The questions were, ‘to what extent does this story possess standard news values such as timeliness and impact?’ and ‘to what extent is this story important?’ For each of the 20 story presentations, these three variables were combined to form a composite newsworthiness score, with story means ranging from 2.80 to 7.81. These individual story scores were summed to create total newsworthiness scores for each set of 10 stories ranging from 10 to 100: Set 1, which both groups saw in AP format, and Set 2, which one group saw in AP format and one group saw in Tweet format.

Twitter use

Journalists were asked whether they had a Twitter account (93% did; those without a Twitter account were excluded from analysis because the questions asked them to imagine the stimulus tweets appeared in their own timeline). Next, participants were asked to report ‘how often are you on Twitter’ on a 7-point frequency scale from never (1) to connected all day (7) (M = 5.44, SD = 1.64). Finally, we asked respondents to rate how much they rely on Twitter for their journalistic work. Respondents rated four items on a 7-point scale ranging from 1 = strongly disagree to 7 = strongly agree: ‘Twitter is useful in my work’ (M = 5.9, SD = 1.2); ‘I often get story ideas and news tips from Twitter’ (M = 5.07, SD = 1.53); ‘Twitter helps me stay on top of my news coverage’ (M = 5.6, SD = 1.41); ‘I take the information I get on Twitter seriously’ (M = 4.77, SD = 1.24). The four items were used to create a composite Twitter reliance variable (M = 5.33, SD = 1.13, α = .86).

Statistical analysis

H1 was tested using independent sample t-tests to compare mean newsworthiness scores among the four conditions in the experiment (2 × 2, story set × presentation format). In order to find support for H1, the mean newsworthiness score for the Twitter condition should be significantly higher than for the AP condition. H2 was tested using independent sample t-tests to compare mean newsworthiness scores between high-frequency and low-frequency Twitter use groups. In addition, multiple linear regression was used to measure how well frequency of Twitter use and Twitter reliance predicted mean newsworthiness scores in the Twitter condition.

Results

The study design allows testing of H1 (Journalists will rate news from tweets to be significantly more newsworthy than news from a traditional AP wire) using both within-subjects and between-subjects approaches. For a between-subjects test, mean newsworthiness scores for story Set 2 were compared between the AP and Twitter conditions. For a within-subjects test, mean newsworthiness scores for story Set 1 (AP) were compared with mean newsworthiness scores for story Set 2. Means for all conditions are similar, and both within-subjects and between-subjects comparisons yield no significant differences. Participants judged each set of stories to be equally newsworthy, regardless of presentation format. These tests do not provide support for H1, which is rejected (see Table 1).

Table 1.

Comparisons of aggregate newsworthiness ratings by journalists.

Group	Format	Story set	N	M	SD
Control	AP	Set 1	91	61.191	10.610
	AP	Set 2	91	60.070	10.873
			t(180) = .482, n.s.
Treatment	AP	Set 1	107	60.146	13.152
	Tweet	Set 2	107	59.340	13.712
			t(212) = .661, n.s.

The between-subjects test (comparing the means highlighted by gray bars) was also non-significant, t(196) = .410, n.s.

H2 predicted that the more a journalist relies on Twitter, the greater newsworthiness they will ascribe to news from tweets, and the data provide support for this hypothesis. About 60 percent of journalists reported being on Twitter ‘several times a day’ or being ‘connected all day’, with the rest using the platform less often. Respondents were sorted into two groups, high-frequency users and low-frequency users. High-frequency users (M = 60.3, SD = 11.52) rated the stories in Set 1 (which both groups saw in AP format) roughly the same as low-frequency users (M = 61.17, SD = 12.91). But when considering the stories from Set 2, which were presented as tweets in the experimental group, differences are apparent. High-frequency users rated stories presented as tweets (M = 61.51, SD = 12.93) slightly but not significantly higher than stories presented in AP format (M = 59.02, SD = 10.87). Low-frequency users rated stories presented as tweets (M = 55.98, SD = 14.36) significantly lower than those presented in AP format (M = 62, SD = 10.79; t(72) = 1.982, p < .05). Looked at another way, both high-frequency and low-frequency users rated stories presented in AP format about the same (59.02–62). But high-frequency users rated stories presented as tweets significantly higher than low-frequency users did (M = 61.51, M = 55.98, t(105) = 2.07, p < .05) – (see Table 2).

Table 2.

Comparisons of newsworthiness ratings by high- and low-frequency Twitter users.

Group	Format	Twitter use	N	M	SD
Set 1	AP	Low	32	62.0	10.79
	AP	High	59	59.02	10.87
			t(89) = 1.252, n.s.
Set 2	Tweet	Low	42	55.98	14.36
	Tweet	High	65	61.51	12.93
			t(105) = 2.068, p < .05

This hypothesis may also be tested using a composite measure of reliance on Twitter. Four items – asking how much respondents rely on Twitter and trust what they see there, how valuable Twitter is for work and so on – were combined to form an index of Twitter reliance (α = .86, M = 5.33 SD = 1.13). We entered this index into ordinary least squares (OLS) regression model predicting ratings of the stories presented as tweets. Frequency of Twitter use was a significant and positive predictor for newsworthiness ratings of stories presented as tweets (β = .282, p < .01), but this association disappears once Twitter reliance was added to the model (β = .253, p < .05). The coefficient for Twitter reliance lowered slightly but was still significant when the number of years worked in news media was added to the model, which was a significant negative predictor of tweet newsworthiness ratings (β = −.203, p < .05). Together, these three variables accounted for 16.6 percent of variance in newsworthiness ratings of stories presented as tweets.

Finally, it is interesting to note that years worked in news media is negatively correlated with both frequency of Twitter use (r = −.268, p < .001) and Twitter reliance (r = −.179, p < .05) – (see Table 3).

Table 3.

Linear regression models predicting newsworthiness of tweets.

	Model 1	Model 2	Model 3
Frequency of Twitter use	.282**	.155	.110
Twitter reliance		.253*	.234*
Years in news media			−.203*
F (df)	8.719 (1, 102)**	7.320 (2, 102)**	6.547 (3, 102)***
R²(%)	7.9**	12.8*	16.6*

N = 102. Cell entries are final-entry OLS standardized Beta (β) coefficients.

p < .05.

p < .01.

***

p < .001.

Discussion

In this study, we present the results of an online experiment among a group of working journalists in the US designed to examine the impact of Twitter on news judgment. Participants were exposed to a series of news headlines – for those in the treatment group, half of the headlines were transformed into tweets. Since Twitter launched in 2007, journalists have widely used the site. At the same time, journalism scholars have studied and questioned its role in news and newsmaking. This study examines Twitter’s role in the journalistic processes of gatekeeping, news judgment, and newsworthiness.

This study was designed to find and measure a bias in favor of Twitter. The thinking was that journalists pay so much attention to what they see on Twitter that tweets might have an outsized influence on news judgment. The results presented here suggest that Twitter does not exert influence on journalists’ news judgment above and beyond more traditional formats of news presentation. Rather, the results suggest a negative bias against the platform among journalists who don’t use it as often as well as among journalists with more years of experience.² Twitter does appear to differentially influence journalists’ perceptions of news value, but it does so in a way divergent from what previous research suggested we might find (Armstrong and Gao, 2010; Holton and Lewis, 2011; Lasorsa et al., 2012; Molyneux, 2015; Vis, 2013). Future studies should further probe this relationship to see if tweets from different sources (the public, politicians, other journalists) differentially impact judgments of newsworthiness. For example, recent work suggests that journalists engage mainly with other journalists (Molyneux and Mourão, 2017; Mourão et al., 2016), at times creating gendered echo chambers and exacerbating gender bias (Usher et al., 2018). The journalists in our study may have projected various identities onto the anonymous tweets they viewed in the experiment, and so the role of source type (e.g. journalists vs the public) and gender beg further empirical examination.

Our finding that journalists who don’t use Twitter as often in their work have a negative bias against the platform has several implications for the field of journalism. First, journalists who outright discount the usefulness of Twitter specifically, or social media more broadly, in their reporting may be dismissing news that much of the rest of their colleagues and the news media are indeed selecting as newsworthy. Giving less credence to news simply because it appears on Twitter means that these journalists, and the outlets they work for, may get left behind in their coverage, putting them at a competitive disadvantage in the news industry, where consumer choice has never been more varied. Individually, these journalists may ultimately become less valuable to their employers. What is more, the organization itself risks underserving its audience if its reporters choose to discount what has become a major news pipeline, particularly about social movements and politics.

On the other hand, events surrounding the 2016 presidential election in the United States have caused many to question the value of social media in our news ecosystem. The spread of false news stories (fabricated accounts designed to attract clicks and generate advertising revenue) was largely blamed on social media, and some journalists have sought to distance themselves from it – or at least regard it with a healthy skepticism. This study was carried out during the election, but before prominent acknowledgment of false news stories; in the wake of the election, more journalists may have developed negative views of social media, potentially widening the gap we observe here.

Perhaps this study’s most interesting finding is that, for journalists who incorporate Twitter into their reporting routines, and for journalists with fewer years of experience, Twitter has become so normalized that tweets were deemed equally newsworthy as headlines appearing to be from the AP wire. To be clear, the stimuli presented tweets from faceless, anonymous users with no links or additional context, and these were rated to be as newsworthy as headlines appearing in the format of APExchange.com. Journalists’ production of news is often shaped by the daily routines and influences (Anderson, 2009; Reese, 2001; Schudson, 1982), and our findings suggest that for some journalists, Twitter may be one of those routines on par with using the AP wire. This may cause concern for some in the field, as stories on Twitter can unfold so rapidly as to outpace important journalistic standards such as verification and sourcing. If a tweet appears to be newsworthy and journalists report on it, or even retweet it, this can contribute to a kind of pack journalism, where everyone is in a race to keep up with the latest social media trends. Concerns over a mob mindset in social media journalism find their base in examples such as misidentifying the Boston Marathon bombers. Immediacy is often a core value of journalism alongside accuracy, but many argue that the former should not come at the expense of the latter. Cases of false and/or unverified information often harm credibility more than any benefits gained by immediacy.

From another perspective, the extent to which journalists rated tweets equally with AP headlines suggests that there is now another channel through which stories may ‘pass’ (DeFleur, 2010). Gatekeeping could originally be viewed as a seat of power for those, such as journalists, in a position to wield it. To the extent that the public now constructs its own news feeds by combining traditional media, social media, and algorithmic recommendations, this power is redistributed. The benefit, from a democratic standpoint, may be that journalists could come to rely less on official or elite sources (a criticism often leveled at mainstream media – see, for example, Bennett, 1990) and begin to include a wider range of news sources coming through social media. If newsworthiness is a strong predictor of a story ‘passing through’ the journalists to the public (Shoemaker et al., 2001), then Twitter may be a conduit through which citizen journalists or other members of the public can influence the mainstream news agenda. Our findings suggest that, at least among journalists who use Twitter often, alternative voices may compete with traditional ones on equal footing. These normative benefits are tempered though by journalists’ use of Twitter to mainly follow and interact with other journalists and elites (Molyneux and Mourão, 2017; Mourão et al., 2016; Usher et al., 2018). Future research teasing out how tweet sources impact journalists’ use of those tweets as a cue or source for news can further assess the extent to which journalists’ gatekeeping power is impacted.

This research is, of course, not without limitations. Our conclusions are drawn from a single study, and replicating the findings could provide further evidence. Like all experiments, a certain amount of artificiality is necessary. In the real world, tweets are not anonymous, and seasoned journalists would likely discern between politicians, other journalists, and the public. Our planned future work in this area will identify different types of sources on Twitter – our attempt to account for that aspect of superficiality present in this study. As a function of our experimental design, the tweets necessarily mirrored the AP headlines, meaning that tweets appeared like headlines rather than the stream of commentary and observations that one is likely to encounter on Twitter. This presentation style could be varied in future studies. Also, this study is limited in its generalizability across journalists as a whole, as our invited respondents opted in to the survey experiment. Further replication of this work across different samples of journalists would add to the generalizability of these findings.

On the other hand, what some may see as our limitations suggest a lack of clarity on the part of journalists about what exactly determines newsworthiness. Journalists who routinely use Twitter on the job rated context-free, anonymous tweets that were devoid of links as newsworthy as headlines from The Associated Press. In this case, journalists’ news judgments may have been based on inferences about the importance of Twitter and apparently little else. Tweets were deemed newsworthy even though they did not contain other attributes by which journalists assess news value. These findings also hint at a profound disconnect between the concept of newsworthiness and other significant news values such as credibility, objectivity, and context.

If journalists’ news judgment is a defining factor in separating the profession from others and from their audiences, these findings suggest that journalists and scholars might reconsider its foundations. In less than a decade, Twitter has come to play a key role in journalistic practice including, as demonstrated here, influencing journalists’ news judgment. This places the platform, its administrators, and its policies (or lack thereof) much nearer the center of journalism than might be expected, calling for greater scrutiny from journalists and researchers alike.

Footnotes

Appendix 1 Acknowledgements

The authors wish to thank Dr Renita Coleman for her feedback on this study when the first author developed it as a proposal in Dr Coleman’s Experimental Methods course. The authors also wish to thank Dr Regina Lawrence and the Twitter Research Group at the University of Texas for their invaluable feedback on the development of this study. Thank you also to Dr Kathleen Searles for her advice and feedback throughout the process, and to the anonymous reviewers whose thoughtful comments improved the study.

Funding

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

Notes

ORCID iD

Logan Molyneux

Author biographies

Shannon C McGregor (PhD, University of Texas) is an Assistant Professor in the Department of Communication at The University of Utah. Her research interests are political communication, social media, gender, and public opinion.

Logan Molyneux (PhD, University of Texas) is an Assistant Professor in the Klein College of Media and Communication at Temple University. His research focuses on digital media and mobile technology, specifically as they relate to journalistic practices and products.

References

Anderson

(2009) Web production, news judgment, and emerging categories of online newswork in metropolitan journalism. ISOJ. Available at: https://www.isoj.org/wp-content/uploads/2018/01/Anderson09.pdf

Anderson

(2011) Between creative and quantified audiences: Web metrics and changing patterns of newswork in local US newsrooms. Journalism: Theory, Practice & Criticism 12(5): 550–556.

Anderson

(2013) Rebuilding the News: Metropolitan Journalism in the Digital Age. Philadelphia, PA: Temple University.

Armstrong

Gao

(2010) Now tweet this how news organizations use twitter. Electronic News 4(4): 218–235.

Beckers

Harder

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

Bélair-Gagnon

(2013) Revisiting impartiality: Social media and journalism at the BBC. Symbolic Interaction 36(4): 478–492.

Bennett

(1990) Toward a theory of press-state relations in the United States. Journal of Communication 40(2): 103–127.

Breed

(1955) Social control in the newsroom: A functional analysis. Social Forces 33(4): 326–335.

Bruns

(2008) The active audience: Transforming journalism from gatekeeping to gatewatching. In: Paterson

Domingo

(eds) Making Online News: The Ethnography of New Media Production. New York: Peter Lang, pp. 171–184.

10.

Carlson

(2015) Metajournalistic discourse and the meanings of journalism: Definitional control, boundary work, and legitimation. Communication Theory 264(2015): 349–368.

11.

Coddington

(2014) Defending judgment and context in ‘original reporting’: Journalists’ construction of newswork in a networked age. Journalism: Theory, Practice & Criticism 15(6): 678–695.

12.

Coddington

(2015) Telling secondhand stories: News aggregation and the production of journalistic knowledge. Unpublished Dissertation, School of Journalism, University of Texas at Austin, Austin, TX.

13.

Coddington

Holton

(2014) When the gates swing open: Examining network gatekeeping in a social media setting. Mass Communication and Society 17(2): 236–257.

14.

DeFleur

(2010) Mass Communication Theories: Explaining Origins, Processes, and Effects. Boston, MA: Allyn & Bacon.

15.

Donsbach

(2004) Psychology of news decisions: Factors behind journalists’ professional behavior. Journalism 5(2): 131–157.

16.

Gans

(1979) Deciding What’s News: A Study of CBS Evening News, NBC Nightly News, Newsweek, and Time. Evanston, IL: Northwestern University Press.

17.

Gillmor

(2004) We the Media: Grassroots Journalism by the People, for the People. Sebastopol, CA: O’Reilly Media.

18.

Hamby

(2013) Did Twitter kill the boys on the bus? Searching for a better way to cover a campaign. Shorenstein Center Report. Available at: https://shorensteincenter.org/wp-content/uploads/2013/08/d80_hamby.pdf?x78124

19.

Hermida

(2010) Twittering the news: The emergence of ambient journalism. Journalism Practice 4(3): 297–308.

20.

Hermida

Lewis

Zamith

(2014) Sourcing the Arab Spring: A case study of Andy Carvin’s sources on Twitter during the Tunisian and Egyptian revolutions. Journal of Computer-Mediated Communication 19(3): 479–499.

21.

Holton

Lewis

(2011) Journalists, social media, and the use of humor on Twitter. Electronic Journal of Communication 21: 1–22.

22.

Karlsson

Clerwall

(2013) Negotiating professional news judgment and ‘clicks’. Nordicom Review 34(2): 65–76.

23.

Lasorsa

Lewis

Holton

(2012) Normalizing Twitter: Journalism practice in an emerging communication space. Journalism Studies 13(1): 19–36.

24.

Lawrence

(2015) Campaign news in the time of Twitter. In: Farrar-Myers

Vaughn

(eds) Controlling the Message: New Media in American Political Campaigns. New York: New York University Press, pp. 93–112.

25.

Lawrence

Molyneux

Coddington

, et al. (2014) Tweeting conventions: Political journalists’ use of Twitter to cover the 2012 presidential campaign. Journalism Studies 15(6): 789–806.

26.

Lewis

Wren

Inthorn

(2005) Citizens or Consumers: What the Media Tell Us about Political Participation. Maidenhead: Open University Press.

27.

Lewis

Kaufhold

Lasorsa

(2010) Thinking about citizen journalism. Journalism Practice 4(2): 163–179.

28.

Lukito

Wells

(2018) Most major outlets have used Russian tweets as sources for partisan opinion: Study. Columbia Journalism Review. Available at: https://www.cjr.org/analysis/tweets-russia-news.php

29.

McGregor

(2018) Social (media) construction of public opinion by Elites. Doctoral Dissertation, University of Texas at Austin, Austin, TX.

30.

Mantzarlis

(2016) Christie stared, Twitter rejoiced: But was it a story? Poynter.org , 2 March. Available at: http://www.poynter.org/2016/christie-stared-twitter-rejoiced-but-was-it-a-story/399526/

31.

Meraz

Papacharissi

(2013) Networked gatekeeping and networked framing on# Egypt. The International Journal of Press/Politics 18(2): 138–166.

32.

Molyneux

(2015) What journalists retweet: Opinion, humor, and brand development on Twitter. Journalism 16(7): 920–935.

33.

Molyneux

Mourão

(2017) Political Journalists’ normalization of Twitter: Interaction and new affordances. Journalism Studies. Epub ahead of print 5 October. DOI: 10.1080/1461670X.2017.1370978.

34.

Mourão

Diehl

Vasudevan

(2016) I love big bird: How journalists tweeted humor during the 2012 presidential debates. Digital Journalism 4(2): 211–228.

35.

Mourão

(2014) The boys on the timeline: Political journalists’ use of Twitter for building interpretive communities. Journalism. Epub ahead of print 6 October. DOI: 10.1177/1464884914552268.

36.

Nielsen

Graves

(2017) ‘News you don’t believe’: Audience perspectives on fake news. Reuters Institute for the Study of Journalism Report. Available at: http://reutersinstitute.politics.ox.ac.uk/sites/default/files/2017-10/Nielsen&Graves_factsheet_1710v3_FINAL_download.pdf

37.

Reese

(2001) Understanding the global journalist: A hierarchy-of-influences approach. Journalism Studies 2(2): 173–187.

38.

Robinson

(2007) Someone’s Gotta be in control here. Journalism Practice 1(3): 305–321.

39.

Rogers

(2010) Diffusion of innovations. Simon and Schuster.

40.

Schudson

(1982) The Power of News. Cambridge, MA: Harvard University Press.

41.

Schultz

(2007) The journalistic gut feeling. Journalism Practice 1(2): 190–207.

42.

Shoemaker

Reese

(1996) Mediating the Message: Theories of Influence on Mass Media Content. White Plains, NY: Longman.

43.

Shoemaker

Reese

(2014) Mediating the Message in the 21st Century : A Media Sociology Perspective. New York: Taylor & Francis Group.

44.

Shoemaker

Vos

(2009) Gatekeeping Theory. New York: Routledge.

45.

Shoemaker

Eichholz

Kim

, et al. (2001) Individual and routine forces in gatekeeping. Journalism & Mass Communication Quarterly 78(2): 233–246.

46.

Singer

(2006) Stepping back from the gate: Online newspaper editors and the co-production of content in Campaign 2004. Journalism & Mass Communication Quarterly 83(2): 265–280.

47.

Tuchman

(1972) Objectivity as strategic ritual: An examination of newsmen’s notions of objectivity. American Journal of Sociology 77(4): 660–679.

48.

Tworek

(2018) Tweets are the new vox populi. Columbia Journalism Review. Available at: https://www.cjr.org/analysis/tweets-media.php

49.

Usher

(2014) Making News at the New York Times. Ann Arbor, MI: University of Michigan Press.

50.

Usher

Holcomb

Littman

(2018) Twitter makes it worse: Political journalists, gendered echo chambers, and the amplification of gender bias. The International Journal of Press/Politics 23: 324–344.

51.

Vis

(2013) Twitter as a reporting tool for breaking news: Journalists tweeting the 2011 UK riots. Digital Journalism 1(1): 27–47.

52.

Williams

Delli Carpini

(2000) Unchained reaction: The collapse of media gatekeeping and the Clinton–Lewinsky Scandal. Journalism 1(1): 61–85.