Twitter journalism in Ireland: sourcing and trust in the age of social media^*

* An earlier version of this paper was originally published in the 2015 Proceedings of the International Conference on Social Media and Society: Heravi, B. R., Harrower, N., Sourcing and Trust: Twitter Journalism in Ireland, International Conference on Social Media and Society, Canada, September 2015.

ABSTRACT

Twitter has been widely adopted into journalistic workflows, as it provides instant and widespread access to a plethora of content about breaking news events, while also serving to disseminate reporting on those events. The content on Twitter, however, poses several challenges for journalists, as it arrives unfiltered, full of noise, and at an alarming velocity. Building on the results of the first national survey of social media use in Irish newsrooms, this paper investigates the adoption of social media into journalistic workflows, journalists’ attitudes towards various aspects of social media, and the content and perspectives generated by these online communities. It particularly investigates how Twitter shapes the processes of sourcing and verification in newsrooms, and assesses how notions of trust factor into the adoption of the Twitter platform and content into these processes. The paper further analyses relationships between journalist profile and adopted practices and attitudes, and seeks to understand how Twitter operates in the current journalistic landscape. While this paper draws its data from a survey of journalists in Ireland, the analysis of the relationship between trust, sourcing, and verification reveals broader patterns about journalistic values, and how these values and practices may operate in the field of journalism as a whole.

KEYWORDS:

• Social journalism
• social media
• news
• sourcing
• data journalism
• verification
• Twitter
• journalism
• trust

Introduction: Twitter and the news

Social media platforms have been widely adopted as communication tools across sectors, geography, cultures, and demographics. People communicate heavily through social media, and when seeking out news, increasingly turn to social media as a news source (Barthel, Shearer, Goettfried, & Mitchell, 2015). This practice of turning to social media for news is partly due to familiarity and convenience bred by the growth of social media adoption, but it is also due to the speed at which information can be disseminated through social media channels. Social media is a fundamental tool for sharing information in real time, and as a result, it has become a central platform for the discovery of breaking events, leading to a broad uptake by journalists. Social media platforms, and in particular Twitter, have been widely adopted in newsrooms over the last number of years, becoming an integral part of the journalist toolkit. This toolkit is put to use in a number of ways: journalists monitor social media for breaking news and content, they use it to find sources and eyewitnesses, and they harness its broad reach to crowdsource varied perspectives on newsworthy events. In a more general way, they also use social media for promoting content, attracting audiences, and driving traffic to personal or professional websites. The wealth of information social media provides is unprecedented in terms of volume, variety, and velocity, if arguably not in terms of quality, and it is shaping the way that news is gathered and disseminated by professional journalists.

In traditional, mainstream, journalism, information is sourced mainly from ‘official places’ and ‘bureaucratic beats’, with reporters stationed at the ‘institutions of society’, and drawing facts from ‘experts’ (Hindman, 1998). Official sources, therefore, are found in press releases from known institutions, respected newswire services, and first-person interviews with officials and experts. However, this traditional practice is changing as social media platforms and user-generated content grow in popularity; social media (and particularly Twitter) now plays an increasingly prominent role in the sharing of real-time information and breaking news (Bunz,2009; Newman, 2009; Doctor, 2011). Established a decade ago, Twitter has proved to be a central source of news for many important events (Broersma & Graham, 2013), with critics assessing the value of this centrality from different perspectives.

A short survey of how Twitter has been used in recent major news events reveals the variety of ways that the flow of information from online communities intersects with or disrupts traditional news practices. For example, the Mumbai terrorist attacks in 2008 have been called ‘Twitter’s moment’ (Caulfield & Karmali, 2008), because Twitter allowed readers to ‘tune into personal, real-time accounts of the attacks’ as mainstream media struggled to keep up with the flow of information. Taking a more sceptical response to Twitter’s role, other commentators noted that ‘the witnesses are taking over the news’ (Jarvis, 2008), or worse, that the free flow, or rather flood, of information aided the terrorists in further attacks (Oh, Agrawal, & Rao, 2010). The power of social media to disseminate news was also made obvious in Iran following the post-election protests in 2009, when it was increasingly difficult to get uncensored news out of the country, and individuals leaked mobile phone video and other information to international news agencies, heralding Iran as the ‘birthplace’ of citizen journalism (Baird, 2010).

Posts from citizen journalists and eyewitnesses can help fill the ‘news vacuum’ that immediately follows an event, when media organisations do not have a reporter in the field ‒ such as during the Haiti earthquake crisis in 2010 (Bruno, 2011). In Haiti, social media mapping tools were used to provide aid and warnings to people before traditional media could report the same (Heinzelman & Waters, 2010). During the 2011 UK Riots, journalists similarly turned to Twitter for crisis communication (Vis, 2013). Throughout the Arab Spring, studies have shown how Twitter facilitated greater inclusion of ‘non-elite’ voices during rapid and often confusing developments (Hermida, Lewis, & Zamith, 2014). The important role of Twitter in breaking news during the numerous uprisings of the Arab Spring in particular has been well documented (Bebawi & Bossio, 2014; Hermida et al., 2014; Lotan, Graeff, Ananny, Gaffney, & Pearce, 2011).

Twitter plays a complex role in the news cycle, because it generates such massive and often unwieldy amounts of information. For journalists working with Twitter, the velocity at which information is circulated creates significant demands in the realm of verification. For example, during the chaos of the Boston marathon bombings in 2013, false information about victims and suspects circulated at an alarming rate (Reinwald, 2014), reminding us that information verification remains a crucial aspect of quality reporting. Broersma and Graham (2012) refer to social networks as ‘huge pools’ of ‘collective intelligence’ that allow reporters to obtain information about ongoing events, to diversify voices, and to get in touch with informative sources. These varied and increasingly complex deployments of Twitter’s rich online community as part of the news cycle point to the vital importance of social media in contemporary news production, and encourage further investigation into the details of how journalists adopt, value, or reject social media in their professional workflows.

This paper approaches this internationally relevant topic through a national case study, drawing on the results of a survey into Irish journalists’ use of social media, which revealed the country’s journalists to be some of the heaviest users of social media internationally (Heravi, Harrower, & Boran, 2014). The study reveals that social media (in particular Twitter) is used most for sourcing news leads and content, and that trust is the main deterrent for journalists for using social media. The purpose of this paper is to investigate the survey data in order to understand how traditional, ‘offline’ journalism norms and practices interact with relevant online communities. Several broad areas of inquiry present themselves: First, given the clear role that social media plays in the worldwide circulation of news, how have Irish journalists integrated the use of social media into their workflows? Second, what can we understand about the finding that journalists use Twitter quite heavily for sourcing leads and content, while maintaining low levels of trust in the very source they are using? Third, how do demographic differences amongst journalists impact their attitudes and professional behaviour with social media?

To address these areas, we pose four interrelated research questions:

RQ1: What are the primary purposes for which journalists use social media, and what are the primary deterrents?

RQ2: How does social media compare to other avenues for sourcing news leads and content and verifying information?

RQ3: What factors affect journalists’ levels of trust in social media as a source, and how does this compare to their trust in other sources?

RQ4: Cutting across the above questions, how does a journalist’s age, gender, knowledge of social media, and experience level affect their use and attitudes towards Twitter for journalistic purposes?

Studies on the impact of social media in the newsroom

Social media has become an integral part of the journalism workflow in newsrooms, and journalists are often looking to ‘scoop’ news from Twitter. Unlike Facebook, its largest social media competitor, Twitter posts are mostly open for anyone to read. Around half a million tweets are posted on Twitter every minute (Statistic Brain, 2015). Twitter’s combined openness, and the abundance of real-time information it generates pose a significant challenge for journalists when it comes to using its content in the news production cycle. At the heart of concerns about the integration of social media in journalistic workflows is the process of information verification; social media content is treated with particular caution for use in newsrooms (Lyon, 2012). As social media adoption and usage grow in the general public, confirming authenticity and provenance of social media content becomes increasingly critical; to deal with this change, journalists and newsrooms have been developing new mechanisms and workflows (both automated and human-centred) for verifying social media content (Browne, 2012; Diakopoulos, De Choudhury, & Naaman, 2012; King, 2012; Little, 2012; Lyon, 2012; Schifferes et al., 2014; Silverman, 2011; Turner, 2012).

In order to investigate the impact of the challenges associated with the ever-shifting social media landscape, a variety of studies have emerged, including a series of profession-wide surveys that have asked journalists about the adoption of social media into workflows. Studies have targeted the role of social media in national and regional contexts, and have run comparisons on international groups of select countries (CISION, 2013a, 2013b; Oriella, 2013). However, Ireland has only now been the subject of such study. The data gathered in the Irish survey provide the basis for analysing national trends, as well as for understanding how attitudes and practices shape key areas such as sourcing, trust, and verification.

In addition to surveys, the literature contains several studies of social media use in the general population, which show differences in use based on social profiles such as age, gender, and education level (Correa, Hinsley, & de Zúñiga, 2010; Perrin, 2015; Pew Research Centre, 2014). For example, younger people use social media more than older people (Ofcom, 2015), and women have slightly higher adoption rates of social media than men (Correa et al., 2010; Perrin, 2015; Pew Research Centre, 2014). When moving from the general population to the more specific cohort of journalists, studies have been conducted into gender differences amongst journalists who have adopted Twitter (Lasorsa, 2012) as well as into the influence of professional profile factors. Significant attention has been paid to journalists on a particular beat, such as the social media behaviour of foreign and political correspondents (Cozma & Chen, 2012; Lawrence, Molyneux, Coddington, & Holton, 2013). A study by Gulyas (2013) looks at the influence of the media sector, length of professional career, and size of an organisation as factors that affect attitudes and adoption of social media from different national cohorts. In this study, journalist age was more significant than professional career length for explaining social media use, and out of the three variables she focused on, media sector was determined to play the most significant role (Gulyas, 2013).

Taking into consideration the differences found in the behaviour of social media users in general, alongside the influence of personal and professional profile factors, our study includes a research question that evaluates the influence of four aspects of user profiles, namely gender, age, knowledge of social media, and experience as journalist.

In addition to national surveys, international comparisons, and analyses of social media adoption based on journalist profiles, a number of studies have investigated specific functions for Twitter in the newsroom, including the use of Twitter in promoting content (Bandari, Asur, & Huberman, 2012; Greer & Ferguson, 2011), the impact of social media use on professional values and journalistic norms (Hedman & Djerf-Pierre, 2013), and how social media may be redefining or reshaping the production and consumption of news (Bruns & Highfield, 2012; Newman, 2011; Newman, Dutton, & Blank, 2012). These studies often raise concerns over the core values of journalism and the importance of trust and verification when it comes to incorporating social media content in news production and storytelling (Bruno, 2011; Hedman & Djerf-Pierre, 2013; Newman et al., 2012). Calling for further studies on how reporters and journalists incorporate social media into their ‘reporting routines’, however, Lawrence et al. (2013) report little to no evidence that Twitter poses any new ‘gatekeeping’ exercise to political news reporters. They further report that Twitter is hardly at all used for fact-checking purposes. With the deluge of information produced on social media platforms, information verification has become more critical than ever before, and social media content are treated with particular caution for use in newsrooms (Lyon, 2012). Hermida (2012) argues that verification of sources and content is a vital part of every journalist’s job, and that the process of verification is central to the credibility of journalism as a profession. With this context in mind, this paper looks specifically at the triumvirate of sourcing, trust, and verification in the Irish context, and aims to contribute to a broader understanding of what motivates journalists to use social media for sourcing, and what concerns are raised in the process.

Methods

This study analyses the results of the national survey of journalists working in Ireland, conducted in 2013 across all types of organisations, beats, and media. The survey was open to all journalists ‒ full-time, part-time, casual, and freelance ‒ but results are limited to those who identify as professional journalists, which the survey defines as those who state that journalism forms a ‘significant part’ of their income. The survey was conducted following ethical approval by the National University of Ireland Galway’s Research Ethics Committee, and was carried out using the online tool Survey Monkey. It was promoted broadly through professional listservs, and community-specific social media channels. Journalists were also direct-targeted through a comprehensive list of media professionals from the National University of Ireland’s Press Office, which holds over 1445 email addresses of individual journalists, as well as national and local media companies. Several articles about the survey featured in local and national media.

A total of 421 people responded to the survey between 2 June 2013 and 20 September 2013, with 259 respondents filling it out to completion. To put this response rate in perspective, a similar survey conducted in the UK (CISION, 2013a) collected less than double the number of responses (589) from a population 10 times that of Ireland. The National Union of Journalists lists 1280 journalists in the Republic of Ireland, and 88 journalists in Northern Ireland, for a total of 1368, which likely includes a number of individuals who do not qualify as ‘professional’ based on the survey’s stated criteria. Using this figure, the response rate (421) is 31%, and the completion rate (259) is 19% of the relevant population. Based on survey completions, we can conclude that the results of this survey are accurate at the 95% confidence level with a 5.5% margin of error.

Respondent profile

Journalists in this study come from a fairly wide age range, with a generally high level of professional experience; 68% of respondents fall between the ages of 25 and 44 years, 80% of the respondents have five or more years of experience as a journalist, with 16% indicating 1–4 years of experience, and 22% saying they have been in the business for 20 years or longer. The gender balance is roughly even, with a slightly higher number of males (55%) in the final results. In terms of journalistic medium, journalists working for print newspapers make up the largest group (39%), followed by radio (17%), print magazine (12%), the online platform of a broadcast or print medium (11%), television (10%), online medium (9%), and personal journalism blog (2%). All areas of content production are represented, with journalists working in Irish news constituting the largest group (37%), followed by a relatively even mix of other beats, including sport, business, lifestyle, politics, science and technology, arts and culture, and world news. Journalists are distributed across organisation size, with 31% falling in the 10–49 employees category, and 23% working for very large organisations with over 500 employees. The vast majority of journalists in this study (59%) report their social media knowledge as ‘better than average’ or ‘expert’. Overall the demographics are diverse, capturing a range of journalists working in different areas, media, and organisations. However, the sample does reflect a group that is experienced in the profession, and savvy in the use of social media, which could be partially accounted for by the likelihood that journalists who use social media more frequently were more likely to complete the survey.

Results

The results of the survey reveal that journalists in Ireland are amongst the heaviest users of social media platforms for work purposes. Almost 99% of journalists who participated in this survey use social media in a professional capacity, with 92% of respondents turning to Twitter at least once a week. This frequency of use is amongst the highest studied, matching the UK rate, surpassing France (91%), Canada (89%), Australia (85%), the U.S.A. (79%), and Sweden (77%) (CISION, 2013b).

Regardless of platform, journalists in Ireland are frequent users of social media; almost half of those surveyed interact (retweet, share, ‘like’, etc.) with content on social media platforms at least once a day, while a further 24% do so several times a week. Those who describe their social media use as infrequent (fewer than one retweet/share/‘like’ per week) are in the minority (21%). This heavy adoption provides a good pool for understanding how adoption factors into the professional journalistic workflow.

The results of the survey reveal that Twitter is by far the most popular social media platform for professional journalists; two-thirds of journalists name Twitter as their most-used social media platform, whereas only 18% name Facebook and 8% name Google+ as their platform of choice for professional purposes. Due to the evident popularity and importance of Twitter in this cohort, and the reported significance of Twitter as the most important social media platform for journalism as a field (Bruns & Highfield, 2012; Hindman, 1998; Newman, 2011), this paper focuses on journalists’ use of Twitter.

The high adoption of Twitter by journalists contrasts with the Irish general population’s use of Facebook as the primary social media platform: polling statistics show that, at the time of the survey, 57% of Irish adults were using Facebook, whereas only 27% had Twitter accounts (Ipsos MRBI, 2013). Overall, the survey reveals that survey participants are significantly more connected to social media than the average general population in Ireland, and that Twitter, and not Facebook, is the platform of choice.

Profile of journalists on Twitter

An analysis of Twitter use by respondents reveals that the microblogging platform is equally popular amongst male and female journalists. However, the age of a respondent corresponds negatively with frequency of use: as age increases, frequency of use decreases. Whereas 82% of 18–24-year olds tweet multiple times a day, this figure falls to 50% in the 55+ age category (which is still a notably high uptake). It is not surprising that younger journalists tweet more frequently, based on the high overall adoption of social media by members of their age cohort. Social media is accessible and immediately comprehensible to younger generations, as it has been part of their communication patterns prior to entering the workforce.

When tweet activity is viewed by the size of media organisation, it becomes clear that ease and accessibility may not be the most important factors in adoption. Where we may have hypothesised that freelancers and small organisations with the most limited resources rely more on free, open, high velocity social media for news than traditional sources, the survey instead reveals that the larger the organisation, the more likely the journalist is to use Twitter multiple times a day for work purposes. When observing Twitter use by primary area of reporting, the results do not reveal any identifiable pattern: Twitter use is least frequent amongst journalists of world news, and most popular with those covering sport, science, and technology. The frequency of Twitter use by journalists covering other areas falls somewhere in between. In terms of publication medium, those working for online publications (including the online platforms of broadcast or print media outlets) unsurprisingly use Twitter more frequently than those working in other media, with just over 80% stating they engage multiple times a day. These findings on publication medium are consistent with comparable surveys conducted in Sweden (Hedman & Djerf-Pierre, 2013) and the UK (CISION, 2013a).

There is a moderate positive significant correlation between frequency of Twitter use and the number of followers. For example, almost 90% of those with a high number of followers (5000 followers or more) state that they use Twitter daily or multiple times a day, while only 50% of those with a small number of followers (less than 100) use Twitter daily or multiple times a day. This result is expected, in that activity on social media tends to generate followers.

Uses of social media: sourcing vs. verification

The adoption of social media in Irish newsrooms is clearly high; however, the results show that the function of social media in a journalist’s traditional workflow varies. To investigate the first research question (RQ1), the survey named a number of uses for social media (sourcing leads, sourcing content, verifying information, networking, publishing/promoting own work, and monitoring trends), and asked respondents to use the following scale: ‘social media is the main thing I use for this’, ‘I use social media daily for this’, ‘I use social media around once a week for this’, ‘I only rarely use social media for this’, or ‘I don’t use social media for this’. The results reveal that social media is most popular for sourcing news leads, with 66% saying ‘I use social media daily for this’ or ‘social media is the main thing I use for this’, and the second most popular for sourcing content (55%). The least popular option was for verifying information, with only 40% using it frequently, and a significant 45% stating they never use social media for this purpose.

The reluctance to use social media for verification could be a consequence of the demands that are placed on journalists’ time when verifying social media content – the process is very time-consuming – a factor that has been recently occupying discussions at broadcaster debates and conferences (Phillips, 2012). However, our findings suggest that the reluctance could also be a consequence of the sense that many journalists believe information on social media cannot be trusted. This is further discussed below in the ‘Trust in Social Media’ section.

To investigate the impact of journalists’ profiles on their attitudes and professional behaviour towards social media, that is, Research Question 4 (RQ4), the study looks at four factors in journalists’ profiles: age, gender, knowledge of social media, and experience as journalist. When these profiles of journalists are considered in relation to the use of social media in the newsroom, the most outstanding correlation appears between journalists’ knowledge of social media, and their overall adoption of social media for traditional journalistic purposes. Journalists with higher self-reported expertise in social media use it more for sourcing news leads, sourcing content, and verifying information. Perhaps unsurprisingly, social-media-savvy journalists also use social media more for monitoring social media trends, publishing and promoting their own work, and networking ‒ all activities that intrinsically require a fairly high level of engagement with social media networks. These three activities are also taken up more by younger journalists, who tend to use social media more than older journalists for monitoring trends, publishing/promoting their work, and networking. Note that 82% of respondents who self-identify as above average or expert in their knowledge of social media are of age 44 years and under. Even though presenting only a small correlation, the gender of journalists also plays a role in their use of Twitter for networking; men seem to use Twitter for networking more than women do. Unlike age, which seems to affect three activities, experience of journalists shows a weak correlation only with networking. Less experienced journalists tend to use Twitter slightly more than their more experienced colleagues for the purpose of networking. Table 1 presents the correlations between the aforementioned user profile factors (as defined in RQ4) and reasons for using social media for professional purposes.

Table 1. Correlations between user profile factors and reasons for using social media.

	Monitoring trends	Sourcing news leads	Sourcing content	Verifying information	Publishing and promoting work	Networking
Age	−.118*				−.109*	−.144**
Gender				.127*		.118*
Knowledge of social media	.321**	.217**	.238**	.207**	.290**	.303**
Experience						−.120*

Question: ‘Rate your reasons for using social media in your job’. Response ranks: 1 =‘Don’t use social media for this’, 2 = ‘Only rarely use for this’, 3 = ‘Use it around once a week for this’, 4 = ‘Use it daily for this’, 5 = ‘Social media is the main thing I use for this’. For gender: 1 = female, 2 = male. The measure of association is Kendall’s tau-c. Tested for each reason separately. Significance test: *p < .05, **p < .01 (two tailed). Only significant correlations are presented.

Social media vs. other avenues for sourcing

To examine our second Research Question (RQ2), how social media compares to other avenues for sourcing and verification, respondents were asked to rate how frequently they use various methods for sourcing news leads, placing social media alongside more traditional sources (Figure 1). Social media, with 64% reporting daily use for sourcing news leads, fairs better than newswires (54%), television (45%), and radio (58%), but is outranked by several other popular traditional sources: ‘direct contact with individuals’ (73%), press releases (71%), and print media (67%).

Figure 1. Where journalists source their news, N = 259.

In contrast to the most popular sources in our respondent set ‒ traditional press releases, contact with individuals, and print media ‒ social media is published nearly instantaneously, which lends it a speed advantage over traditional news formats for communicating breaking news or new developments in running stories. In the current global media landscape, where readers can switch between news sites with the click of a mouse or swipe of a finger, readers expect continual updates. The constant flow of information on social media provides an excellent forum for sourcing story leads in this pressured and fast-paced milieu where consumers have insatiable appetites. In the survey, two-thirds of respondents report that they use social media heavily for sourcing leads. At first glance, this suggests that social media is taking over as a way of providing story leads. However, if we look more closely at the provenance of these social media leads, it is clear that many of these leads are from sources that journalists can access outside the world of social media. Seventy per cent of journalists indicate that they draw leads from the social media accounts of people they know in the real world (e.g., friends and colleagues), 65% take leads from ‘influential individuals’ in their area, and 51% use information posted on the official social media accounts of media companies. In contrast, only 27% of journalists indicate that their social-media-based story leads emanate from sources they are connected to solely through social media. This finding suggests that, for sourcing leads, social media is currently providing an additional or alternative platform for accessing sources, rather than providing new root sources of information.

To examine the profile factors defined in Research Question 4 (RQ4) in relation to sourcing, the gender of a respondent presents a weak significant correlation when it comes to sourcing news from online sources, namely social media, blogs, and ‘other’ online sites (excluding the online platform of media organisations). Men use these sources more frequently than women as a means for sourcing news (Table 2). The age and experience of a journalist both have a positive correlation with more frequent use of traditional media as a source; older and more experienced journalists more heavily use television, radio, print, direct contact with individuals, and most significantly, reporters in the field. Age and experience parallel each other when it comes to the popularity of sources chosen by journalists. Analysis of the cohort of respondents reveals that age and experience themselves are highly correlated; hence, it is likely that the same people are answering this question.

Table 2. Correlations between user profile factors and where they source the news.

	Social media	Newswire	TV	Radio	Print	Reporters in the field	Direct contact	News media website	Press release	Blog	Other online
Age			.131*	.137*	.118*	.200**	.112*			−0.044
Gender	.116*									.156**	.127*
Knowledge of social media	.214**									.256**	.143**
Experience			.127*	.165**	.125*	.203**	.184**

Question: ‘What do you source your news for professional purposes?’ Response ranks: 1 = ‘I never use this source’, 2 = ‘I use this once a month’, 3 = ‘I use this once a week’, 4 = ‘I use this every day’, 5 = ‘This is my main source of news’. For gender: 1 = female, 2 = male. The measure of association is Kendall’s tau-c. Tested for each source separately. Significance test: *p < .05, **p < .01 (two tailed). Only significant correlations are presented.

Knowledge of social media also has a positive correlation with more frequent use of social media, blogs, and ‘other’ online sources, for sourcing news.

However, the massive number of sources available on social media means that knowledge of the platform does not equate to trust in the source, as journalists are less likely to know the person connected with a relevant Tweet, and familiarity with a source helps to create trust in that source. Hermida et al. (2014, p. 481) argue that ‘the more familiar journalists are with a source, the more likely they are to be considered credible’. From a slightly different angle, our findings support Bruno’s qualified conclusion that despite the heavy incorporation of user-generated social media content by ‘citizen journalists’ into major news agencies’ websites (CNN, B.B.C., and The Guardian), reporters in the field still have an important role to play, because they bring credibility to information sources (Bruno, 2011). Part of that role for journalists could be to investigate beyond known sources on social media, to the wider chorus of voices that only social media can provide. The next section looks more closely at how the concept of trust factors into the use of social media in newsrooms.

Trust in social media

Facts, truth, and reality are considered to be the ‘god terms’ in journalism by Zelizer (2004). The process of verification is considered to be one of the most important aspects setting apart journalism from other forms of communication, and social media poses a challenge for verification. This challenge creates a tension for professional journalists who rely on social media, but who also uphold the ethical imperative to verify all sources (Hermida, 2012). This tension is the basis for our third Research Question (RQ3) on trust and sourcing, and is clearly evident in the survey results: even though the daily adoption of social media for sourcing leads (64%) and content (55%) is relatively high, overall, journalists report very little trust in social media as a source. In fact, lack of trust in social media is reported as the strongest deterrent in using social media in our study, with 64% of respondents agreeing that they are deterred from using social media because information cannot be trusted.

The identification of trust issues as a deterrent in adopting social media for sourcing is corroborated by another question, which compares social media to other commonly used sources for news. In this response, only 12% of respondents stated they are ‘fairly confident’ that social media is a trustworthy news source, and only one respondent stated without qualification that he/she trusts social media as a news source. The only listed option considered less trustworthy than social media was blogs, at 9% for ‘trusted’ or ‘fairly confident’.

The trust level of social media as a source contrasts strongly with other commonly used news sources, revealing that overall, the trust level in social media is not the result of generalised journalistic scepticism, but reflects medium-specific concerns. As presented in Figure 2, ranking highest for ‘I feel fairly confident this source is trustworthy’ or ‘I trust this source’ are reporters in the field at 85%, followed by newswires (80%), direct contact with individuals (77%), radio (77%), print media (76%), websites of traditional media outlets (71%), and television (69%). Other online publications (44%) and Press Releases (34%) were generally considered less trustworthy as sources. Note that this figure is 12% for trust in social media.

Figure 2. Trust in various sources of news, N = 259.

In terms of the four defined profile factors in Research Question 4 (RQ4), younger journalists declare higher levels of trust in the online sources of social media, blogs, and ‘other’ online sources, than older journalists, as well as slightly higher trust in press releases. Interestingly, there is almost no correlation between trust and age for any of the listed sources, which perhaps reveals a general increase in scepticism as journalists age. These findings may have little to do with journalism as a profession, as they are reflective of broader societal trends around trust in the media. Levels of trust in social media have been shown to decline as users age, with 25–34-year olds showing greater trust than any other groups in social media, and indeed in other aspects of the media (Reuters, 2013).

Similarly, the survey reveals that males show slightly higher trust levels in all the online sources named in the survey than females, including social media, news media websites, and blogs. Trust in other sources of news ‒ typically considered more ‘traditional’ sources such as television, radio, press releases, newswires, and reporters in the field ‒ does not reveal statistically significant differences based on respondent gender.

The number of years a respondent has practised as a journalist shows a significant correlation with trust in reporters in the field. More experienced journalists reveal a statistically significant higher level of trust in these reporters than less experienced journalists. This divergence from findings related to the age of the journalist may indicate that life experience – in the world in general – has more of an effect on one’s trust in sources than specific experience as a journalist.

The relationship between trust and knowledge of social media shows a positive correlation, with more social-media-savvy respondents revealing higher trust in social media, as well as in blogs. Knowledge of social media also shows a strong negative correlation with age ‒ self-designated ‘experts’ tend to be younger. Increased expertise in social media does not show a statistically significant correlation with trust in any other source of news. Correlations between user profile factors and most trusted sources for news are presented in Table 3.

Table 3. Correlations between user profile factors and most trusted sources for news.

	Social media	Newswire	TV	Radio	Print	Reporters in the field	Direct contact	News media website	Press release	Blog	Other online
Age	−.132*					.120*			−.120*	−.108*	−.114*
Gender	.127*							.153**		.147*
Knowledge of social media	.175**									.157**
Experience						.148**

Question: ‘What are your most trusted news sources?’ Response ranks: 1 = ‘I don’t trust this source’, 2 = ‘I will use this source, but don’t fully trust it’, 3 = ‘I am neutral on whether this is a trustworthy source’, 4 = ‘I feel fairly confident this source is trustworthy’, 5 = ‘I trust this source’. For gender: 1 = female, 2 = male. The measure of association is Kendall’s tau-c. Tested for each source separately. Significance test: *p < .05, **p < .01 (two tailed). Only significant correlations are presented.

Thirty-seven per cent of journalists who say they do not trust social media for sourcing news also identify themselves as infrequent users of social media (those who post original content to social media platforms less than once a week). These infrequent users are twice as likely to distrust social media as users who post original content to social media sites multiple times a day. The correspondence between heavier use of social media and greater levels of trust in its content may shed light on the ongoing debate about whether or not journalists who are very active on social media deviate from traditional professional practices and norms (Hedman & Djerf-Pierre, 2013; Hedman, 2015; Lasorsa, Lewis, & Holton, 2012). However, such an investigation is beyond the scope of this current paper.

To put trust as a deterrent in context, the second strongest deterrent for using social media (45%) is concerns about copyright and ownership, or intellectual property restrictions on social media. This is followed by concerns around privacy or reputation, with personal privacy ranking higher as a deterrent (37%) than concerns over the source’s privacy (25%). Privacy issues are understandable given that the most-used platform, Twitter, comprises almost entirely public content. Very few respondents (3.5%) feel too constrained by their organisation’s guidelines to use social media for work purposes, and only 10% state that social media was not used widely enough by their relevant contacts to be useful. Given these figures, the lack of trust stands out even more clearly as a key deterrent. Regardless of the stated lack of trust, journalists are using social media for sourcing, so what steps do they take to fill the ‘trust gap’?

How social media content is validated

When social media content is deemed worthy of further inquiry, journalists pursue a number of methods for validating information. The most popular method adopts the long-standing practice of direct contact with trusted individuals. Eighty-three per cent of respondents state that, in order to validate content on social media, they contact official bodies, such as press offices or the police. Sixty-five per cent of journalists tap into existing ‘real-world’ networks for validation, 52% say they contact the social media source directly. Outside of direct contact with individuals, 63% reveal that they validate by cross-referencing information online, or by cross-referencing on social media (40%). A small percentage (15%) use technical means (i.e., tools) to verify information, which could be the result of the relatively small number of tools available for these purposes (e.g., processing metadata and geo-tagged information, or using available services for checking authenticity of images). Overall, the validation of content, relying more on direct, individual contact than other means, remains a time-consuming process, even if the time required to find the lead is shortened by the instantaneousness of social media publishing.

Factors in determining trustworthiness of a source

When asked what is used to determine the trustworthiness of a social media source, 73% of journalists choose a ‘link to an institutional or company website’ as important, very important, or extremely important, whereas 59% choose external validation ‒ going beyond what is provided by profile information in the source (Figure 3).

Figure 3. Factors in determining trustworthiness of source, N = 259.

These figures suggest that social media is being adopted as a tool for accessing information partly because the veracity of that information is corroborated outside of social media. However, in contrast, the quality and number of posts on a social media account are named as important, very important, or extremely important by 70% of respondents, suggesting that the quality of the account activity itself significantly contributes to trustworthiness. The source’s Avatar, that is, profile image (24%), and the age of the social media account (40%) were considered the least important aspects amongst the options.

If we break down the responses around trustworthiness of a source by our four factors of age, gender, knowledge of social media, and experience as journalists, there is a significant, if weak, negative correlation between the age of journalists and most of the factors used to determine trustworthiness of a source. Younger journalists attribute more weight to a social media account’s number of followers than older journalists, as well as to external factors such as the ‘verified account’ badge assigned by Twitter. An account’s avatar is also a more important factor in determining trust for younger journalists than older journalists, even though as noted above, avatars are the least important factor overall, with only 24% of respondents choosing them as a factor at all (Figure 3). In fact, the only factor that younger journalists did not find very important in this category was the age of an account.

Interestingly for older journalists, there is no significant correlation between higher age and any of the trustworthiness factors. This may mean that older journalists are in general more sceptical about the trustworthiness of social media content in general, and do not believe that any of these factors will eventually help them in determining trustworthiness of sources on social media. A journalist’s level of experience, determined by the number of years working in the profession, presents similarly to age, in that higher levels of experience equate with placing less weight on most trustworthiness factors. As we have seen that there is a strong correlation between age and experience, it is difficult to determine which of these two factors may be more important in creating a general scepticism around the markers of trustworthiness. There is a notable correlation between less experienced journalists and the belief that an account’s number of followers engenders trust ‒ less experience correlates with more weight given to the number of followers. Paralleling the findings on journalist age, overall there is a correlation between less experience and more weight being given to various markers of trust offered in the survey. No significant correlation between gender and factors to determine trustworthiness of a source was noted, whereas knowledge of social media presents a weak correlation with the following trust factors: number of posts, number of followers, location, and external validation, such as verified account badge. Table 4 presents the correlations between user profile factors and factors for determining trustworthiness of a source on social media.

Table 4. Correlations between user profile factors and factors for determining trustworthiness of a source on social media.

	Age of account	Number and quality of posts	Number of followers	Link to institutional website	Location	Avatar	Verified account
Age		−.147**	−.282**	−.130*	−.181**	−.176**	−.199**
Gender
Knowledge of social media		.127*	.185**		.106*		.129*
Experience		−.133*	−.257**		−.113*	−.140**	−.150**

Question: ‘When determining whether or not a social media source is trustworthy, rate the importance of each of the following’. Response ranks: 1 = ‘Not important at all’, 2 = ‘Somewhat important’, 3 = ‘Important’, 4 = ‘Very important’, 5 = ‘Extremely important’. For gender: 1 = female, 2 = male. The measure of association is Kendall’s tau-c. Tested for each factor separately. Significance test: *p < .05, **p < .01 (two tailed). Only significant correlations are presented.

Perhaps not surprisingly, a correlation analysis between the sources a journalist uses, and the sources he or she names as trustworthy reveals a consistency between a preferred source and trust in that source (Table 5), as they all present a positive significant correlation between use and trust. For example, journalists who rely on newswires, radio, and reporters in the field show a moderate significant correlation to trusting newswires, radio, and reporters in the field over other sources, respectively. The relationship between the source used and the trusted source holds for social media as well, with those adopting social media as a source trusting it more than other sources.

Table 5. Correlations between where journalists tend to source their news from and which sources they trust the most.

Where do you source your news?	What are your most trusted news sources?
	Newswire	Social media	TV	Radio	Print	Media org. website	Blog	Other online	Reporters in the field	Direct contact	Press release
Newswire	.427 **		.184**	.255**	.124*	.162**			.283**
Social media		. 188**					.165**	.119*
TV	.143*		. 296**	.272**		.168**			.169**
Radio	.224**		.249**	. 304**	.178**	.213**			.227**
Print	.197**		.200**	.247**	. 229**	.229**			.127*
Online platform of media outlets	.216**		.148**	.180**	.162**	.210**		.142*	.117*
Blog		.170**					. 285**	.139*			−.119*
Other online						.128*		. 137*
Reporters in the field	.174**							−.109*	. 474**
Direct contact							−.120*		.167**	. 289**
	.214**				.110*			.115*	.189**		. 282**

Question 1: ‘Where do you source your news for professional purposes?’ Response ranks: 1 = ‘I never use this source’, 2 = ‘I use this once a month’, 3 = ‘I use this once a week’, 4 = ‘I use this every day’, 5 = ‘This is my main source of news’. Question 2: ‘What are your most trusted news sources?’
 Response ranks: 1 = ‘I don’t trust this source’, 2 = ‘I will use this source, but don’t fully trust it’, 3 = ‘ I am neutral on whether this is a trustworthy source’, 4 = ‘I feel fairly confident this source is trustworthy’, 5 = ‘I trust this source’. The measure of association is Kendall’s tau-c. Tested for each reason separately. Significance test: *p < .05, **p < .01 (two tailed). Only significant correlations are presented.

Conclusion and discussion

Our findings suggest that the full adoption of social media as one amongst many available quality sources of news is in a transitional phase, with journalists integrating Twitter into their workflows enthusiastically, while noting a series of deterrents to its use. Journalists frequently adopt social media to find stories and content for their stories, but they are reluctant to rely on it for verification. They have integrated it to varying degrees in all areas of their journalistic workflow, but they quite clearly state that it cannot be trusted, despite the fact that they draw most of their information on social media from known or authoritative sources, and almost a third of respondents report that they ‘strongly agree’ they would not be able to carry out their work without social media. The stated distrust combined with high adoption and the clear necessity of social media for work could point to the pressure that journalists are under to compete with various online communities in breaking stories, and would be an interesting avenue for further research.

To address our first research question (RQ1), the most common use of social media by Irish journalists is for sourcing, with two-thirds of the population using it daily or as their primary tool for identifying leads, and 55% using it in the same way for sourcing content. On the opposite side of the spectrum, almost half of the respondents were clear that they turned to other methods for verifying information, stating they never use social media in this capacity. The clearest deterrent in using social media is a sense, despite the notably high adoption rates for sourcing leads, that social media cannot be trusted. Given the strength of this response (only 12% stated they were ‘fairly confident’ social media could be trusted), future research would benefit from addressing this issue in a more granular way, as the ability to fully understand the definition of trust in this context is a limitation of this particular study.

If social media has been mostly heavily adopted for the purpose of sourcing leads and content, the paper’s next query was to put this in context, and determine how this adoption factors into existing processes for finding leads. Our findings on this question (RQ2) show that social media has become one amongst many avenues for sourcing leads, being used slightly more than some traditional sources (newswires, television, and radio) and slightly less than others (print media, press releases, and individuals). At first glance this may be an unremarkable finding; however, the level of adoption for social media lead-sourcing is significant when considered in relation to the short length of time that social media has existed. For the current news landscape, social media plays a significant role in one of the key areas of journalism’s workflow, suggesting that it has embedded itself as a crucial part of the journalist’s contemporary toolkit. Online communities have become part and parcel of newsroom activity.

However, the issue of trust in social media still remains. Investigating our third research question (RQ3), this study shows that most journalists still list ‘authority sources’ as their most trusted for sourcing – turning to reporters in the field, newswires, print media, and their own direct contact with individuals, especially if they are older, or more experienced in the profession. The top factor in deeming a social media source as trustworthy is a link in the social media account to a company or institutional website, and indeed indicators from ‘outside’ the social media world are key to the verification process. When it comes to validating content found on social media, the strongest tendency is to go outside of social media – to official sources and contacts in ‘real-world’ networks. Similarly, journalists most consistently use the social media accounts of people from their own real-world networks, ‘influential individuals’, and trusted media organisations, suggesting that social media may not be providing new sources of information, but perhaps just new ways of accessing that information. This finding is commensurate with the idea that social media is a ‘social awareness stream’ (Naaman, Becker, & Gravano, 2011) perceived mostly as a new tool for carrying out the standard journalistic task of ‘finding out what is going on in the world’ (Hedman & Djerf-Pierre, 2013). The relationship between the availability of a plethora of new voices as potential sources, and the effort to actually incorporate these voices, therefore, requires further investigation. Social media users contribute content and leads, but given our findings it is not clear how much this content is being adopted into mainstream media stories.

Considering Research Question 4 (RQ4), it is interesting to note that younger journalists display higher levels of trust in social media, and they engage with it more heavily for peripheral areas of their jobs (monitoring trends, promoting their work, and networking). Nonetheless, these ‘peripheral purposes’ to journalism are in fact the core purposes of social media in general, and by engaging with them, younger journalists are developing their profiles in ways that could feed back into the development of sources. Hedman and Djerf-Pierre (2013, p. 5) argue that if

a journalist has many followers on Twitter, and is retweeted by the right persons, or hosts a blog that is linked to by others, it enhances the individual’s social capital in the professional field, at least in the eyes of other active social media users.

If younger journalists have a greater sense of social media’s power, then this trend of greater use and adoption will over the course of time impact the profession. In general, our findings indicate that heavier use of social media relates to stronger statements of trust, and those who report the use of social media as a source are more likely to trust social media, which suggests that adoption leads to more familiarity with the markers of trust. Similarly, the level of experience with social media correlates with heavier use for both core areas of journalism (sourcing and verification) and peripheral areas.

Despite the stated lack of trust in social media, high adoption rates indicate that social media is important for journalists, and perhaps further indicates a changing attitude towards social media. This change is supported by the fact that the second most important factor in determining a social media account’s trustworthiness is the number and quality of posts found on the account. Number and quality of posts are a factor that exists entirely within the framework of social media – it does not rely on external links, connections to authority institutions, or connections to pre-existing sources. The current literature on Twitter adoption in the news cycle points to a process of normalisation (Hermida, 2013), where journalists adapt social media to fit professional practices and norms, while at the same time adjusting their practices to accommodate evolving technologies (Lasorsa et al., 2012), and our findings suggest that conceptions of trust may be evolving in a similar vein.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Dr Bahareh Heravi is the leader of Insight News Lab and a Research Fellow at Insight Centre for Data Analytics in National University of Ireland, Galway. She is also a Lecturer in Information and Communication Studies at University College Dublin. She is the former Lead Data Scientist at the Irish Times, and the co-founder of Irish Times Data.[email: [email protected]].

Dr Natalie Harrower is the Director of the Digital Repository of Ireland, Ireland's national trusted digital repository for social and cultural data. [email: [email protected]].

Additional information

Funding

This publication has emanated from research conducted with the financial support of Science Foundation Ireland (SFI) [grant numbers 12/TIDA/I2389, SFI/12/RC/2289, 13/TIDA/I2723, 13/TIDA/I2743 and 14/TIDA/2419].

Notes

* An earlier version of this paper was originally published in the 2015 Proceedings of the International Conference on Social Media and Society: Heravi, B. R., Harrower, N., Sourcing and Trust: Twitter Journalism in Ireland, International Conference on Social Media and Society, Canada, September 2015.