Transparency to the Rescue?

Transparency has emerged as an ethical principle in contemporary journalism and is contended to improve accountability and credibility by journalists and scholars alike. However, to date, few attempts have been made to record the public’s views on transparency. This study enriches current knowledge by using data from an experiment, survey and focus groups in Sweden collected between 2013 and 2015. Overall, the results suggest that the respondents are not particularly moved by transparency in any form; it does not produce much effect in the experiments and is not brought up in the focus groups. While that is the key finding of this study, it should also be noted that various forms of user participation are evaluated negatively, while providing hyperlinks, explaining news selection and framing, and correcting errors are viewed positively. Implications for journalism practice and research are discussed.


Introduction
Transparency has emerged as an acknowledged ethical principle in journalism and is considered to both challenge traditional objectivity and journalistic practices, while also providing an opportunity to (re)connect with citizens to build trust and credibility (Phillips 2010;Lasorsa 2012;Morton 2015;Slattery 2016;Vos and Craft 2016). However, the general notion of transparency has to be embodied in order to impact everyday news production and be visible to the public. Previous research has found that news outlets use what is effectively known as transparency tools, such as explaining news selection and using corrections, which can be viewed as the transparency norm in practice. To date, few studies have investigated to what extent different transparency tools are providing the desired outcome according to citizens' perceptions and evaluations of news content and journalistic performance. Drawing on extensive empirical material from Sweden collected during a larger research project on transparency and credibility-using a survey, experiment, and focus groups-this study assesses how eight transparency tools influence how citizens view journalism.

Transparency as Theory and Practice
news is made and inviting the public to be a part of the news-making process (for a fuller discussion, see Singer 2007;Plaisance 2007;Karlsson 2010;Vos and Craft 2016). Transparency has been heralded as an instrument which allows journalists to be more accountable to the public (Lasorsa 2012;Lasorsa, Lewis, and Holton 2012;Singer 2007) which potentially increases trust and credibility (Phillips 2010;Lasorsa 2012;Morton 2015). Hence, many journalists and researchers consider that transparency can change journalism to better fulfil its role in society vis-à-vis the public.
Previous empirical research includes studies on how transparency has been appropriated by the journalistic field in trade journals and reviews and by journalistic associations (Vos and Craft 2016;Slattery 2016); journalists' views towards transparency (Hellmueller, Vos, and Poepsel 2013;Chadha and Koliska 2015); and how transparency is manifested in content and news organizations' publication platforms (Karlsson 2010;Phillips 2010;Lasorsa 2012;Lasorsa, Lewis, and Holton 2012;Morton 2015). However, the public's perceptions of transparency have not received sustained scholarly attention (notwithstanding van der Wurff and Schoenbach 2014; van der Wurff and Schönbach 2014; Karlsson, Clerwall, and Nord 2017a).
Thus, it is questionable how a more transparent journalism can improve the public's perception of journalism when we still know little about the public's views on transparent journalism. This is important because of journalism's relational character-it is shaped in interaction with citizens, policy-makers, and other stakeholders, and its role in society varies across time and space (Ward 2005;Sjøvaag 2010). Therefore, transparency must also have a relational dimension if it is to be recognized as a key ethical principle outside of the journalistic field. Moreover, in order for the public to be able to evaluate transparency and hold journalism accountable, it must be performed in a manner that can be observed (Tuchman 1972;Karlsson 2010); i.e. transparency as a normative concept needs to be converted into tools that can be used in everyday journalism and appear in relation to news stories. Otherwise, there is no way of knowing whether transparency is in use or not and, subsequently, to what extent the public appreciates it. Transparency tools pertinent to this study have been identified in previous research: explaining news selection and decisions, corrections of errors, utilizing hyperlinks, disclosing the preferences and motifs of the media worker, and giving people outside news media an opportunity to criticize and intervene in the journalistic process (Deuze 2005;Rupar 2006;Hayes, Singer, and Ceppos 2007;Kovach and Rosenstiel 2007;Karlsson 2010;Karlsson, Clerwall, and Nord 2017a).
Up until now, research has focused on how actors within journalism view or perform transparency in news. This study directs the spotlight outside of the profession and specifically on the views of the public that is supposedly better served by transparency. To inform this concern, the study asks: How does the public view transparency and to what extent does this impact credibility?

A Multimethod Approach to Study Citizen's Views of Transparency
This article is based on a synthesized analysis of a larger research project comprising three different methods conducted by the researchers in Sweden during 2013-2015. The empirical material consists of 13 focus groups (N = 82), representative data from an experiment (N = 1320), and a survey (N = 2091). This paper reports only the transparency tools that are consistent throughout the three different methods. These tools include disclosure transparency (Karlsson 2010)-explaining news selection and framing, the presence of hyperlinks to source documents and previous articles, journalists exposing their personal opinions, corrections of news; and participatory transparency (Karlsson 2010)-readers participating in news creation or replacing journalists entirely.
Since the experiment has already been reported in full elsewhere (Karlsson, Clerwall, and Nord 2014), the main focus in this paper will be on the survey and focus groups with findings from the experiments used supplementarily. The three methods were carried out using the services of Swedish opinion research institute TNS/Sifo. Details of the three methods and how the material was collected can be found in previous accounts (Karlsson, Clerwall, and Nord 2014, 2017a, 2017b. However, some important details from the focus groups have not been described elsewhere. More specifically, the focus groups started with a very broad approach where the participants were asked about their views on news and their news consumption. Then, they were asked what they considered to be "good," "ideal," and "credible" journalism in order to determine if, when, and how they would raise issues regarding transparency in journalism. Only later in the session were they asked specifically about different forms of transparency. At the end of the session, they were asked to rate (as described in detail in Figure 1) how they viewed different transparency tools. At the very end of the session, the respondents were explicitly asked whether any of the transparency tools currently influenced or could have a future influence on their news consumption; however, due to a misunderstanding, this procedure was only performed in eight of the 13 focus groups. While this mistake was unfortunate, it did not affect the general discussion about transparency.

FIGURE 1
Share of votes (per cent) for each transparency tool. Links = The use of internal and external links for more information; Rationale = Explaining the reason(s) behind publishing the story; Correction = Informing the audience about previous errors in the news; Angle = Explaining the angle used for the news piece; Opinion = Disclosing the reporter's personal opinion in the matter covered in the news piece; Comments = Allowing user comments; Reader participation = Inviting the reader (audience) to participate, e.g. material such as photographs for the news piece; Readers' text = Inviting the readers to write the news piece (e.g. letting an elementary school teacher report about the conditions at his/her school)

TRANSPARENCY TO THE RESCUE 1925
The two quantitative methods (experiments and survey) used a deductive approach to see how the public responded to transparency tools. In contrast, the focus group started off with a more inductive approach to see how the respondents would introduce transparency into the discussion, but closed with a deductive approach in which participants were asked for their assessments of transparency tools. Taken together, this approach explores how the respondents view transparency in journalism in general and eight transparency tools specifically.

Results
The data are very rich and we are only able to offer a rather short synthesis of the results in the space allowed. In order to relay the gist of the findings, the results are presented in three ways. The first is an interpretation of the overarching results from the research project. Secondly, an overview of how the eight transparency tools are viewed is reported, including the survey results (Table 1) and the aggregate ranking of how the focus groups rated them ( Figure 1). Finally, we walk the reader through the transparency tools independently, adding accounts from the focus groups and some observations from the experiments, starting with the transparency tools that the public view most positively.

Transparency Has Little Appeal and Is of Minor Concern to Citizens
The overall impression from the three methods is that transparency is a nonissue for most people. First, there were no or weak effects in the experiments, possibly suggesting that the respondents did not note, understand, or care about transparency. Second, many of the respondents in the survey chose the neutral option when rating the majority of the transparency tools and, furthermore, there was a rather high number of dropouts for these particular questions (up to 20 per cent). Last, but certainly not least, transparency was neither brought up in the focus groups in general terms nor were there any discussions about any specific transparency tools, except for a brief mention of corrections and journalists' bias. Thus, when the respondents had the opportunity to talk about journalism-i.e. what makes journalism good and credible, and what guides their own news consumption from their perspectives and concerns-transparency was not on the agenda. Transparency was discussed after the moderator introduced it as a subject at the end of the sessions (roughly 60-70 minutes into the sessions). Hence, when the respondents thought about their relationship with journalism and what makes journalism "good" and "credible," transparency was not discussed. Moreover, the discussion on transparency ended rather quickly in the focus groups when it was brought up. The respondents did not seem to have any developed ideas or great concerns about transparency. This was explicitly articulated in Focus Group 11 at the end of the session, when the respondents were asked to contemplate how, if at all, transparency related to their news consumption: "This [transparency] is not all that important for how we choose news outlets. 'In summary, when the lack of effects in the experiment, the dropout rate, and the preferred neutral option in the survey are combined with the disapproval and lack of attention in the focus groups, the main finding is that transparency has a much higher status among researchers and journalists than it has among the respondents. Table 1 and Figure 1 illustrate attitudes towards transparency tools from the survey and focus groups, respectively. Table 1 shows that the respondents are most positive towards being informed when and why news reports are erroneous (e.g. corrections) and least positive towards journalists mixing reporting with their own opinions and entirely replacing journalists with audienceproduced news content. It can also be noted that they do not seem keen to know the journalists' own opinions, even outside of the context of news stories.

Attitudes Towards Different Forms of Transparency
Moving on to the focus groups, the participants were asked to indicate which transparency tools they deemed important for credibility. Each participant was given seven votes that he/she could distribute freely among the eight tools used in the focus group discussion. Thus, a participant could choose to place all seven votes on one specific tool (e.g. the use of reader comments) or spread the votes across several tools (e.g. three for links, two for comments, etc.). Thus, rather than just asking about their attitude towards transparency tools in general (as in the survey), this exercise forced the respondents to prioritize what kind of transparency tools they would prefer. In total, 504 votes were cast and Figure 1 presents the results. The results from the focus groups give a somewhat different impression from those of the survey. Hyperlinks are the most appreciated transparency feature in the focus groups, while rationale (explaining why the story was published), angle (explaining the framing), and corrections score rather high here as well. The participatory features score much lower in relation to how they performed in the survey vis-à-vis the other features. The impression from the focus groups is that it is disclosure rather than participatory transparency that is appreciated. Next, the transparency tools are commented on in detail.

Hyperlinks
If media organizations seek to improve how citizens view journalism in general, and journalism's credibility in particular, hyperlinks are probably the best means of improving the public's perception. In the experiments, there were small but positive effects towards both source and message credibility. The survey reported neutral to slightly positive attitudes towards hyperlinks, and the inclusion of hyperlinks was the most appreciated transparency tool according to the ranking in the focus groups. When the respondents talked about hyperlinks in the focus groups, they said they added to journalism by offering more information on the issues; either by providing additional information for very interested readers or because some stories related to such complex issues that they cannot be fitted into just one news story. They also articulated that hyperlinks make it possible to track down original sources and documents. These can then be compared with how journalists portray events, because the participants do not always view journalists as knowledgeable or objective enough. It was also emphasized that hyperlinks should link to credible and/or relevant sources. Several examples were given about how difficult it can be for users to navigate between hyperlinks or that news sites sometimes link to hoax or blatantly commercial sites. At the end of the sessions, when explicitly asked for transparency tools that are/could be important for their news consumption, hyperlinks were by far the tool most emphasized by the respondents.
Overall, the three studies show that hyperlinks can be an asset to journalism when employed properly.

Explaining News Selection and Framing
Another potentially useful transparency tool is an explanation of news selection and framing for the public; i.e. to explain why a news story is selected to be published, worthy of public attention, and framed in a particular way. In both the survey and focus groups, "explaining news selection" was received more positively than "explaining news framing". There was a small positive effect in the experiment with regard to message credibility when news framing is explained.
The transparency tools of explaining news selection and explaining new framing both sparked discussions among the respondents. It was pointed out that selection and framing are implied in journalism and the public should be aware of these approaches; thus, it is not important to explain them. Others said that it would benefit journalism and increase credibility if these were better explained. With regard to news selection, this tool was appreciated because it would put the news story in context with events, processes and problems in society, and make it easier for the public to understand the background and relevance of the news. In relation to background information, the respondents also made explicit references to how useful hyperlinks to previous news stories on the subject could be. Explaining framing was viewed as positive because this would enable the public to see the position from which the news pieces were written and other positions that could be possible. However, there were also some criticisms of explaining framing; it was argued that the readers themselves were capable of decoding the framing (and did not need to be told by the journalists). Several focus groups commented that they had never actually seen framing explained in relation to a news story.
A few of the focus groups (FG 2,4,12) indicated that explaining news selection and framing could be a useful transparency tool which could influence their choice of news outlets. The impression from all three methods is that explaining how news production is conducted is valued by the respondents and that explaining news selection is slightly more valued than explain news framing.

Corrections
While corrections had no effect whatsoever in the experiments, they were rated highly in both the survey and focus groups (Table 1, Figure 1). When discussing corrections, the respondents indicated that they are a good idea and that it increases credibility when news media admit wrongdoing. However, it was also stipulated that the corrections should be published equally prominently as the original article and that the reader's attention should be directed to the error. Many of the focus groups thought that the news media could perform better in this regard. While corrections where viewed positively, the respondents said that corrections should be kept to a minimum and that journalists, ideally, should get the story right/correct from the beginning.
Another study (Karlsson, Clerwall, and Nord 2017a) based on the survey data reported that the respondents were very sensitive towards errors and that corrections were not as effective when the errors were large. The same data show that the respondents were more critical the larger the error and the results from the focus groups confirm this. Corrections were mentioned in one focus group as an important tool that could influence news consumption.
In summary, we conclude that reporting when and why errors occur could be a somewhat useful transparency tool.

Journalists' own Opinions
The experiment produced small positive effects for messages of credibility disclosing journalists' opinions, but this was not the case in the other elements of the study. On the contrary, it was viewed rather negatively in the survey and it was not rated as high as the three former transparency tools in the focus groups (Figure 1). When the issue was brought up for discussion, some respondents were vocal in their support, while others were equally vocal in arguing that this was neither needed, nor did it have a place in journalism. Those who supported the idea and thought it would improve journalism's credibility, repeated the TRANSPARENCY TO THE RESCUE 1929 argument used with regard to explaining news framing; i.e. the reader would be served by knowing the journalist's own opinion in order to properly evaluate the news article. Those opposing the idea argued that this was unprofessional and a breach of journalism's mission to be impartial and objective. There seemed to be a tolerance of journalists' own opinions or having a stronger personal voice in some specific genres (e.g. sports, travel, history, cars, and opinion pieces) but not in issues pertaining to politics or social issues.
Taking data from all three methods into account, the results about journalists' opinions are somewhat ambivalent; the survey data indicates unpopularity (relative to the other tools) and the results from the focus groups demonstrate that some participants find this appealing while others strongly oppose it. Overall, this transparency tool could be useful, but would probably also be contested. The ambivalent result points to a need to further refine what kind of transparency is attractive to what kind of public for what reason-rather than a one-size-fits-all approach.

Participation
Finally, the least useful or even counterproductive transparency tool is user participation. The experiment shows that negative user comments lowered the trustworthiness of the journalist writing the story and had a negative impact on the perception of the story being based on facts. On the other hand, positive comments decreased the respondents' sense of partisanship on behalf of the journalist. Both the survey and focus groups showed that the more the audience was involved in the news stories, the less appreciated it was by the respondents. In particular, the results from the focus groups reveal that participation is viewed with scepticism (as elaborated elsewhere; e.g. Karlsson, Clerwall, and Nord 2017b). Furthermore, participation in the form of user comments was mentioned in one of the focus groups (FG 9) as a possible asset for journalism if those people commenting could refrain from writing "stupid things," but that this was not currently the case.
The differences between the dissimilar outcomes in the survey and focus groups might be due to, as Bergström and Wadbring (2015) suggest, the disparity between participation as an idea and the actual outcome. The two methods might trigger different connotations of participation, as described by Borger, van Hoof, and Sanders (2016); the survey may have triggered "participation for me" and the focus groups may have triggered "participation in general/participation as outcome".
In any case, user participation is at best something that does not engage a large part of the public, but could instead hurt journalism.

Transparency Not to the (Immediate) Rescue
Based on the results of this study, there is no major transparency effect on credibility, nor are there urgent demands for it by the public. With this in mind, it is probably too optimistic to expect that transparency will help restore journalism's credibility. Rather, if transparency is going to contribute to the backbone of journalism, it is more likely a question of teaching the public to recognize and appreciate it. Nevertheless, the study shows that some transparency tools are more likely to improve the public's perception of journalism than others. The great divide seems to be between disclosure transparency, which is valued, and participatory transparency, which is disdained.
However, if the public at large needs to be educated about the benefits of transparency, why then is transparency specifically championed-by researchers and journalists alike-as a tool to bring the public closer to journalism and not any other form of normative framework, be it old or new? Also, if transparency is not the answer to (some of) journalism's problems, what is? While these questions are beyond what can be satisfactorily answered by our data, we encourage scholars to exercise a more critical perspective towards transparency and look more broadly into the relationships between the news media and the public, and what makes or breaks that relationship.
With regard to the role that transparency can play in journalism, we think a more detailed research agenda is needed. To date, transparency has, in our view, been treated as too abstract and general, and assumed to have some underlying benefits. Specifically, we suggest that scholars could study what kind of transparency will serve what kind of role for different segments of the public and other stakeholders under different circumstances.
Arguably, people will seek more transparency when a news organization is experiencing heavy criticism or in a crisis, but not, as our results might indicate, under normal circumstances. As our results suggest, some transparency tools will most likely be better at addressing public concerns than others. Then again, it seems that people with different expectations and credibility assessments might request different transparency tools, because there seems to be mixed views on the need for journalists to disclose their own views. The results may also be impacted by the cultural context in which they were collected (i.e. Sweden-a high trust, welfare state with relatively high news consumption) compared with studies in other.
While most transparency tools, save participation, are received positively, it should also be asked if working with transparency tools is worth the resources required, because transparency seems to be of relatively little overall importance to the respondents. The lack of effect of and disinterest in transparency by the public begs the question of whether the performative level (i.e. journalists' actual work) is the driver of changes in norms, credibility and trust and, if so, what should the timeline of change be?
On a methodological note, the three different methods convey rather different images of what transparency means to the audience. In the survey, when the respondents were primed to think about transparency, it seemed that they had at least some idea of its role in journalism. In contrast, the experiment exposed the respondents to different transparency tools at the performative level (i.e. in use in actual news item), but this did not produce much effect and only small effects when they occurred. Likewise, in the focus groups, transparency was a nonissue until introduced by the moderator and it was largely met with unconcern when it was brought up. In future research, scholars should be careful about relying on single method studies, especially surveys which seem to trigger responses that are not validated by other methods.

DISCLOSURE STATEMENT
No potential conflict of interest was reported by the authors.
TRANSPARENCY TO THE RESCUE 1931