JCMC 4 (4) June 1999
Message Board

Collab-U

CMC Play

E-Commerce

Symposium

Net Law

InfoSpaces

Usenet

NetStudy

VEs

VOs

O-Journ

HigherEd



Conversation as Publishing: the Role of News Forums on the Web

Ann Light and Yvonne Rogers

School of Cognitive and Computer Science
University of Sussex


Table of Contents

Abstract

The development of the World Wide Web (Web) has allowed publishers to move away from traditional newspaper models of news presentation to introduce more flexible products that offer both an information source and more scope for interaction with and between users. The opportunity to involve users more in the creation of news content has been exploited in various ways: for instance as discussion forums or as visitor responses to particular questions. This paper reports on an experiment investigating one form of this new kind of interaction, drawing on data from an e-mail questionnaire sent to visitors to The Guardian newspaper’s Election 97 website who observed or participated in a series of themed discussion forums in the run up to the 1997 British general election. We present an evaluation of the visitors’ behavior in the forums, their motivation and their perceptions of the discussion space. We discuss the findings in relation to the underlying model of the website, pointing out how, despite the flexbility offered by this new mode of interaction, the traditional publisher-contributor relationship remains preferable to both users and developers.

 

Introduction

Publishing has traditionally been a one-way process whereby publishers provide content for readers. However, the interactivity of the Web is redefining the relationship between user and producer. Publishers (and we shall subsume broadcasters within this term) have gained the opportunity to move away from just presenting material, towards a situation where content can be negotiated with their audiences allowing for more opportunity for public exchange. At the same time, the Web is providing the potential for producers to become ‘hosts’ - developing participation among users of the site, where the producer's role is to provide communication software and display the outcome of visitors' exchanges. There is some evidence that participation of this kind builds product loyalty and this may lead to a pool of customers [1]. Certainly, users are finding themselves with potential for more participation than before the development of online publishing.

At the moment publishers are experimenting. Many hosted web forums exist now, where publishers have supplemented their information-based sites with electronic space for discussion. So far, though, there has been little analysis of what users believe they are doing in these spaces, or how exchanges are managed. To investigate the benefit of providing this kind of interactive forum, an experiment was conducted by the British national newspaper, The Guardian, which hosted a series of themed discussion forums during the run up to the 1997 British general election as part of their substantial campaign news and information website. To assess the effectiveness of The Guardian’s choices in presenting an electronic discussion space, an analysis was carried out by the authors, examining the expectations, perceptions and behavior of the visitors to it. In particular, we were interested in seeing what kinds of people were attracted to these sites, and whether people who had never taken part in an electronic discussion were motivated enough to take part. We were also interested in whether visitors would return to the site and why. This paper presents the findings of the evaluation, exploring the make-up of the set of visitors in terms of their experience of participation and their motivation. It also contrasts these with the developer’s expectations and aspirations of how the web forums would work.

The Study

The study of The Guardian newspaper’s electronic discussion space formed part of a larger investigation into the usage and usability of The Guardian election website as a whole. This included analysis of logs, user interviews and usability testing [2]. However the focus of this paper is the findings of the questionnaire sent to people who had registered to use the discussion forums, supplemented by information gathered at interviews with Guardian staff involved in the design of the forums. Although the forums were observed as part of the study, no analysis of their structure and content is given here as our concern was with what users thought of the electronic discussion space, how it met expectations and the reasons they gave for their behavior in it, rather than an analysis of the kinds of discussion that took place during the debates (cf [3]).

Background

The role of the new web forums in publishing will depend very much on the perception of the visitors to the site as to whether it is worth their while contributing, together with how easy it is made by the publishers for them to do so. To begin, we compare traditional publisher-reader relationships with some of the emerging models for online electronic interaction.

Traditional Publisher-reader Relationships

After reading a column, listening to an interview on the radio or seeing a news story on TV, people often talk with each other about what they have just read, seen or heard, and whether they agree with it. Some go further and choose to participate in the forums that have traditionally been made available by the various media publishers to viewers, listeners and readers for discussion. However, the form of interaction that occurs in traditional publishing is very different from the discussions that take place in face-to-face settings. Because of the cost of production and distribution, the means of publishing have largely been owned by commercial interests who protect their investments by controlling the quality or message of prospective publications. Thus, even where the public or its chosen representative is involved in giving a view, a decision is made by the publisher about what to include and some reshaping may take place. Nevertheless, much energy is devoted to market research and establishing audience ratings or circulation figures. A variety of mechanisms exist for public interaction, including

  • Audience determined means, such as the right to reply, especially where the matter concerns the grievant; letters; complaint lines; phone-ins, such as donations, votes, competition entries or opinions.

  • Publisher determined means, such as Vox Pops and reaction pieces, where a range of people perceived to be affected by the news are asked for their opinion; community contributions; invited audience, with the opportunity to ask pre-arranged or spontaneous questions of guests.

Despite this set of mechanisms, much of the nature of traditional publishing interaction remains staged, restricted by both space and editorial concerns.

Online Forums

In contrast, one of the benefits of moving media-based forums online is that the Web has the potential to allow more voices to be heard. Electronic-based media are less bounded by spatial and temporal constraints: it is possible for hundreds of thousands of people to send messages to a website - as witnessed by the condolences sent following Princess Diana's death.

There is a relatively long history of computer-mediated communication: electronic discussion groups, preceding the appearance of the Web, have been around for several years in the form of newsgroups, conferences and forums. There are now thousands of them online, discussing a diverse range of topics. In the beginning, many of these discussion groups were created through individuals or groups setting them up within and outside of their workplace. As such, they were largely run as ‘bottom-up’ autonomous enterprises, with no commercial interest or organisational involvement. More recently, however, there has been a trend towards commercial organisations setting up ‘top-down’ forums, which are aimed at attracting readers and viewers to take part in online discussions about topics they cover. In particular, many newspapers and TV companies have set up websites that provide a combination of further information with various related ‘interactivity spaces’ for supporting more communication with their audience. These new forms of forum, therefore, are in contrast with the earlier kinds of electronic discussion groups, insofar as the nature of the relationship between the creators and the participants is very different.

New Forms of Publisher-reader Relationships

Within these ‘commercial’ electronic discussion spaces there have been various attempts to provide a structure to foster two-way exchange of ideas. Some have attempted to adopt the ethos of Usenet newsgroups, with an egalitarian model of communication, rather than of top-down publishing. They have sought to provide a space for participants to post messages and respond to other postings in given topic areas. Other publishers have changed the format to create new models of exchange, for example, ideas are supplied by the host of the site to which the visitors are invited to respond. An outcome of this approach, however, is that communication may be restricted to a single exchange: participants may only respond to a question raised by the host, sometimes at the expense of all inter-visitor discussion. Hence, in some ways this kinds of forum is much closer to the familiar relationship between media producers and their public rather than the interactions that go on in Usenet newsgroups. An example is the CNN website, which is part of the arsenal of one of the world's most popular news providers. A click away from all news pages on the site is the community section which offers a chat facility and a ‘message board’. Registration is required to post to the boards. Despite this, the message board is used by thousands daily, posting their messages in response to topical events chosen by the news service. Limited threading has been provided to categorize the topics, but the sheer volume of postings to each makes it unwieldy to respond to themes. Instead postings appear largely as disconnected broadcasts, unrelated to one another, with little discussion.

Another electronic space which shares similar features to the CNN message board is the BBC’s Talking Point, launched late in 1997. Like CNN, the BBC website introduces news topics of the day with a short explanation and a question. However, the BBC's approach is even closer to traditional publishing. Visitors are greeted by a ‘talking point’ topic and a choice of others in a menu on the side of the screen. ‘Talking points’ combine a news story with comments from the public, submitted on a form at the bottom of the page. This resides one click away from all the BBC news pages and requires no registration at all. A linked page, called ‘Your Reaction’, offers further short, pithy comments in each case. The quality of comment is guaranteed because the content of the site is moderated: a majority of postings do not appear. While contributions are not rewritten, views are chosen for succinctness, style and representativeness. The downside of this filtering , however, is that it makes it difficult for visitors to respond publicly to points raised by others.

These alternative electronic spaces indicate, therefore, that discursive tone and content is heavily constrained by the nature of the relationship between participants and producers. The communication tools employed also act to define the kind of interaction styles that are likely to arise. These forums, in being asynchronous, encounter problems not normally present in synchronous communication. For instance, poor threading can stop discussion and mediation can interrupt dialogue, while awkward submission mechanisms can increase the investment needed to post. In general, it appears that the more intrusive the communicative mechanism, the less fluid the exchanges [4].

The trade-offs involved in making decisions about how to involve the public make it likely that these electronic spaces will continue to evolve. To evaluate the benefit of one pioneering approach, we will now look at a British newspaper’s (The Guardian) election website forums, (collectively named 'The Debating Chamber' after the British Parliament forum for legislative debate), which in many ways was an attempt to combine features from both traditional and new forms of publisher-reader relationships.

The Design of The Guardian Site

The Guardian newspaper prides itself on its political coverage. The creation of the election website was viewed as a way of increasing its profile as a publisher on the Internet. It also provided a means of launching its other Web activities. The website was designed to run for a limited period, for the 40 days of the official British general election campaign through March and April 1997.

The orientation of the news carried on the website was liberal left, designed to appeal to professionals such as teachers and IT and media workers. The existing public image of the newspaper also helped to define from the start the kind of relationship it could establish with its forum visitors. The use of this kind of branding as a way of identifying the kind of website that was being established was extended further by involving one of the newspaper editors in taking an active role in the running of the website.

A main part of the website was the provision of a forum for discussions about news and politics. In all there were eight separate forums under the ‘Debating Chamber’ banner: six dealt with topical issues in Britain, such as 'Northern Ireland' and 'Education', introduced by an anonymous paragraph. The seventh was a general politics forum. The eighth featured Will Hutton, the editor of the sister Sunday paper, The Observer, and something of a celebrity in intellectual circles.

Hutton’s inclusion was an experimental feature: at the time there had been celebrity interviews on the Web, but little sustained participation by a familiar name. Billed as the 'Will Hutton' forum, he authored the introductory blurb and followed this up with postings. Initially these postings appeared among others with no marked difference. About half way through the campaign, however, his contributions started to bear his name in capitals to ensure they stood out, possibly in response to criticism that he was not participating enough.

Initially, The Guardian development team were interested in attracting visitors who would want to return to the site and actively engage on a regular basis (Hunt 1997, personal communication). This was, in part, their rationale for the design: providing information and interactive activities intended to attract the visitor back. To achieve this they included a range of interactive elements alongside the provision of current news. These included the forums, quizzes, political self-assessment forms that matched visitors' views with manifesto pledges and a daily vote. They also promised prizes for the best contributors in the way of free books.

<I>Guardian</I>'s interface

The development team decided to build its own interface to organize the messages that were posted. The messages were presented as a series of embedded threads in a frame on the left hand side of the screen. Individual messages could be expanded in full on the right hand side of the screen (see Figure 1). The threading was designed to enable direct links from the postings to the replies.

Another design decision was the inclusion of a short registration form at the point of access. The reason for this was that The Guardian wanted to maintain some accountability in the expression of views presented in the postings. As part of registration, a warning about libel was displayed. Another feature was that each posting was headed by the contributor’s e-mail as a live link. This ‘e-mail link’ was intended to allow readers to reply directly to contributors as well as identifying them. A mechanism was included, however, to enable those who wanted to remain anonymous to remove the e-mail link.

Having this set of controlling mechanisms meant that visitors who wished to participate in the forums needed a certain level of commitment. At the same time it afforded some notion of privacy for the contributors. In sum, the website was designed to have features that were part of the traditional publisher-reader relationship, but also included a range of new features, intended to attract visitors to return to the site and take an active part in keeping it ‘alive’. In the next section we see what happened when the website went live and whether it met the developers' and visitors' expectations.

What Happened When the Website Went Live

When the website went live, many visitors (about 70,000) were attracted to it. The interactive sections of the website (like the quizzes) proved to be popular. However, far fewer took an active part in the forums. Of the total number of visitors during the campaign, only 2100 registered to visit the forums and of these only about 400 people actually posted to them. Nevertheless, the discussions that did take place were lively. The developers were happy with the standard of debate that took place, which they felt avoided the excess sometimes associated with electronic newsgroups discussions. They also felt that the discussions were very much in keeping with the general style of the newspaper (i.e. intellectual, liberal).

All the eight forums that were provided were used during the campaign. Some were more popular than others, especially the general forum (some 650 postings) and the one that included participation by the newspaper editor (Hutton) in it (some 400 postings). Threaded discussions branched into many initial replies and extended down several levels in some cases. Heated exchanges took place occasionally, especially where humor was misunderstood, but most were courteous.

The following examples, quoted from the various forum exchanges, show that some conversational interaction between participants was taking place:

"I'm not at my best today, hope that makes some sense_.. :-)'"

"What I don't understand from your earlier posting, is how you can square_"

"This offering is now pinned on the kitchen wall - our thoughts for the week".

Shortly after the election, seven forums closed: only the general forum stayed open and this was used sporadically by a small subset of participants until the end of 1997, though it was no longer promoted. This withdrawal of support reflected The Guardian’s perception of the debates' being based around an event. This view was borne out in its usage: there was a rise in activity as the election neared and immediately after it, followed by rapid decline. Nonetheless, a few people - who had come to know each other and staff at The Guardian quite well through e-mail and phonecalls - declared their intention of keeping discussions running between themselves after the election and these were some of the stalwarts who continued to use the site facilities to comment on the progress of the new Government.

Investigating Attitudes and Perceptions through a Questionnaire

In addition to interviews conducted with The Guardian website team, an electronic questionnaire was designed and sent to registered visitors. The developers wanted to know more about how visitors responded to the forums, with the aim of improving involvement in the future. Thus, we designed a questionnaire (appendix) specifically to find out more about forum visitors’ behavior, perceptions and motivation.

To analyse the different kinds of visitors we categorized them in terms of people visiting a discussion site for the first time (newcomers) and those with experience from other electronic discussion spaces (experienced). We also classified them in terms of what they did when coming to the site – those who simply read (non-participants) and those who made a contribution (participants).

Specifically, we wanted to:

  • examine the differences between newcomers and visitors experienced in electronic discussion – a) to see if new participants had been attracted and by what, and b) to look for trends in behavior that could inform future decisions;

  • examine the differences in behavior and views between non-participants and participants – to see if non-participants identify as strongly with the forums as the participants.

More generally we wanted to:

  • assess the benefits and disadvantages of adopting a hybrid model of interaction – leading to greater confidence in design decisions.

Finally, we were interested in the efficacy of the various mechanisms employed to encourage and control participation. These included:

  • registration and the e-mail link placed on contributions – to establish what impact they had;

  • involvment of a celebrity figure from the newspaper – to see whether this encouraged participation.

From these goals, a series of questions was specified (see Figure 2) in relation to whether they addressed visitors’ behavior, perception or explanations.

For example, recipients were asked to say whether they were visiting an electronic discussion space for the first time, if not, whether they habitually participated in discussions; whether they had participated on this occasion and at which point; and whether they had returned after an initial visit. Other questions focussed on their views. Both quantitative and qualitative data were collected, giving respondents the option to choose multiple answers to specified responses and encouraging them to add their own comments throughout.

 questionnaire analysis

The Questionnaire Findings

Just under 30% of the people receiving the questionnaire responded (572 returned from about 1960 received – out of 2166 registrations, some e-mail addresses being inaccurate or duplicated). A tolerance to filling in forms was shown already in their preparedness to register. There was a proportionately higher return from participants (190/approx 400: 49%) than from non-participants (342/approx 1550: 19%) - though whether higher engagement came from participating or vice versa cannot be determined. There were many eloquent comments, with about 40% of the respondents giving a detailed opinion to the open-ended question at the end. Several respondents thanked The Guardian for consulting them – especially those based outside Britain - and there was mostly a chatty tone. Many responses showed a degree of trust in the producers: ‘When it comes to The Guardian Group I trust that [registration] was for a good reason as I have never known you to be anything but honest. Generally I avoid giving any info about myself online.’ (Quotes in bold are all extracted from comments in response to the questionnaire).

Visitors who responded were split 2:1 between people who were familiar with discussion groups in electronic environments (399 respondents) and those who were new to the concept (190). The more experienced group was split almost evenly between those who posted regularly elsewhere (209) and those who visited discussions elsewhere just to read postings (188). Only somewhat more than half of respondents (54%) returned to the forums for a second visit (336): a majority of these were participants. Unsurprisingly, those identifiying themselves as one-time visitors were more likely to be critical, scoring higher on options such as ‘I was quickly bored’, and more of them cited ‘no time’ as a reason for not participating (101 against 45, which is 6% more).

The responses of participants as a general category (190) can be further analyzed in terms of:

  • people new to discussion groups – to see if they responded the same way as more experienced visitors (60 respondents)

  • experienced non-participants who chose to post on this occasion (readers-turned-contributors) – to find out what made them behave untypically in this forum (57 respondents)

  • any participant who had joined the discussions after a first visit to the site (late joiners) – to establish what prompted them to join in, since clearly it was not just the opportunity to do so (79 respondents).

The most popular option chosen by respondents to explain their participation was to ‘put my opinion’ (152 respondents or 80%). This could be interpreted as suggesting that many of the respondents were following a traditional publishing model rather than the newsgroup model of taking part in a discussion. However, disagreeing with a previous posting was the second most popular reason selected (77 respondents or 40%). This shows the value of including a mechanism in the forum that enables inter-visitor debate. A small number of people gave other reasons such as endorsing a view (26 respondents or 13%), moving forward the discussion (28 respondents or 14%) and asking for information (26 respondents or 13%), all of which are even more conversational in nature. The findings of the questionnaire support the observational evidence of the exchanges in the forums that people’s intentions were to discuss and influence, as well as to air their views.

Very few participants were motivated by the chance to win something (5 respondents or 2%), suggesting that participation was its own reward. Also very few had forgotten why they got involved (7 respondents or 3%). Perhaps those for whom participating was less memorable chose not to return the questionnaire.

It became apparent, though, that the motives of newcomers differed in certain respects from those of the experienced participants. We found that newcomers were a little more likely to post to ‘put an opinion’ (53 respondents or 88%), far more likely to ‘see what would happen’ (27 respondents or 45%, compared with 27% overall and 17% of readers-turned-contributors) and less likely to post because they disagreed with someone else (18 respondents or 30%), than the average. In contrast, some experienced discussion group browsers who rarely contributed (readers-turned-contributors) gave quite specific reasons for why they had been stimulated into posting: ‘Contributors were trivializing an important subject that affected the lives of many disadvantaged people.’ Whereas this last example reflects a commitment to changing people’s attitudes and was clearly a response to the content of other postings, by contrast, many of the people who were new to forums were responding to the mere fact of the forum, as is indicated by the answer ‘see what would happen’.

We also found that non-participating newcomers who responded were a bit more likely to express their behavior in positive terms, scoring higher on statements ‘would rather just read other people's opinions’ (70 respondents, which was 24% against 20% for all non-participants) and ‘read it for information’ (103 respondents, which was 35% against 30%) and marginally lower on those indicating lack of interest (by an average of about 3%). Not surprisingly then, this group was found to be more likely to return to the forums than non-participating visitors with greater experience, who perhaps chose to go elsewhere.

The most common reason given for not participating was lack of time (148 respondents, 43.5%) and some people amplified this by saying that they read the forums at work, where discussing would be frowned on. In general, non-participants answered less fully than participants.

As we mentioned earlier, The Guardian adopted the newsgroup practice of putting a direct e-mail link to the contributor on each contribution, unless expressly removed by the user. We were interested in finding out whether it left people feeling too exposed to want to post or whether it served to give some identity to postings. The answer is that it had little effect, deterring very few respondents (about 3%). Eighteen percent of respondents (106 respondents) did not even notice that the link existed.

Typically, participants saw the link more in terms of accountablity than non-participants - like ‘putting my address on a letter to the editor’ (93 respondents, or 49%), whereas one non-participant annotated her questionnaire: ‘Completely different… Few would mail-bomb a letter to the editor!’

It inspired a few positive comments from participants, who viewed it as a communication medium rather than just an annoying redundancy in this context:

‘Someone did write challenging what I had said, and our private exchange rapidly degenerated into name-calling, but I did not mind that.’

‘I expected some flame mail but got nothing. It seems your readers are quite civilized.’

‘The link led to contacts from interested readers that led to further discussions outside of The Guardian Forum.’

‘I was rather disappointed no one wrote to me whether to agree or not. I wasn't sure if it would have been correct protocol to contact someone directly.’

‘Who knows, it could become the latest way to pick up people who share your politics!’

Hence, one spin-off of incorporating the e-mail link was that a limited group of visitors were spurred on to private discussion. However, a large group - predominantly non-participants - added comments that they were worried about the possibility of receiving contact from outside the forums in the form of junk mailing lists.

We also wanted to know whether newcomers were approaching the forums with a different perspective from visitors who were familiar with other discussion environments. If so, did they express a different attitude to the e-mail link placed prominently at the top of each posting? Results were not significantly different between newcomers and experienced visitors, except on the subject of unwanted mailing, where fewer newcomer-participants than any other group seem concerned (only 24%, compared with an average of 36%).

The questionnaire went on to ask how visitors regarded interacting with newspapers, as the nearest comparable medium. Answers revealed that respondents were generally prepared to interact - nearly 75% would write or had written to a newspaper for publication. This is a high figure, showing that the respondents using the forums were not typical of newspaper readers, even The Guardian's, who are quite a vocal group. Interestingly, 12% of participants (23 respondents) said that they would never consider writing to a paper but had decided to try out the forums. One person commented that using electronic forums ‘was so easy and responsive and you were assured of publication of your ideas without them being lost in yesterday's news’.

A few respondents identified themselves as particularly interested in the political dimension of having political forums. Potential for a more democratic view of news was touched on: ‘It should be available at more than election time (I think it is the future for democracy, seriously!)’ to the point of running campaigns: ‘The internet is full of discussion groups, but few if any have contributions from experts or from decision makers. A national newspaper has the contacts and the standing to solicit contributions with the aim of focussing the debate. For example if you were to set up a discussion on education or health care you would get demands for higher taxes and other traditionally left of center positions being taken up.’

Will Hutton’s contributing impressed only 28% (161 respondents) and few identified his participation as one of the reasons for their own participation (26 respondents).

To summarize, visitors who responded had mixed views about using the forums, with many accessing them only once and even regular visitors blaming lack of time for their unwillingness to participate. This suggests that they may be too time-consuming for most people.

Analysis of Comments

There were many comments from all sides on access problems, speed of downloading and other network-related problems. There were also criticisms of the design of the debates: the posting framework, the use of frames to display the discussion, threads that disappeared off the screen as discussions grew, the display of only one posting at a time and the unmanageable size as the election drew closer. It was also pointed out that registration was not needed for reading postings if its purpose was to discourage libel.

Only one commentator felt that The Guardian should have steered clear completely from debate: ‘There is a myriad of newsgroups and IRCs covering such areas as politics and current affairs, and I didn't feel it was appropriate for a newspaper's Website to contain a discussion forum.’ His view, however, was implicitly endorsed by other experienced posters: ‘Online forums are not newsgroups and as such do not have the cut and thrust of newsgroups and local chat groups, the text based discussions that were around in the BBS days. This is partially because of the inexperience of many of the newer posters (and so many of them are new-) and hence their lack of understanding of the nuances of tone and structure that can be communicated through a short burst of electronic text’. and: ‘It's still print-orientated one way traffic, not interactive.’

Another, more supportive, view from experienced visitors was: ‘I prefer political newsgroups to forums generally but feel the latter come into their own when hosted by journalists or politicians. e-mail is quite easy to fake [ ] but if Will Hutton posts to The Guardian forum, it is more likely to be the real Will Hutton.’ In general, these comments commended The Guardian for attempting to exploit the advantage it had over newsgroups, or suggested the paper should go further by supplying background information. When they criticized Hutton's contributions, it was for lack of visible involvement: ‘Will Hutton did NOT participate in any meaningful forum sense. Don't promise what can't be delivered.’

By contrast, newcomers commented far less often on the effect of the newspaper link, knowing no other discussion space and possibly taking the host for granted. Their comments tended to ignore the difficulties of online discussion and suggested they had a different model in mind: one far more like traditional publishing. For instance, they also had a grievance about Hutton but it was expressed differently - he did not answer specific points they had raised: ‘I just wonder why Will Hutton did not respond. My guess is that he either did not see that particular posting or was busy with election coverage and other matters.’

Another comment about Hutton shows the full extent of potential ambiguity about whom participants were addressing and for what purpose. A newcomer who then became a regular participant: ‘I'd like to know if any of the issues posted get taken up by The Guardian and reported on. I rather felt like posting up an opinion was rather a waste of time. My particular posting was about education [ ]. I would have liked The Guardian to have taken up this story, but I didn't even get a reply from Will Hutton. Perhaps this is only to be expected, as the print rate of letters to The Guardian is probably very small also - a few % I suspect. So, all in all, I felt the discussion forums had very little impact.’ This comment strikes at the heart of the producers’ uncertainty about what they were achieving in the forums, which we shall return to consider later.

Both groups welcomed the idea of Hutton’s involvement as adding another dimension to the debate and as indicating how seriously the producers took it. However, experienced users’ comments suggest having a celebrity in a forum is like having a ‘super-participant’, similar to a conversationalist at a dinner party, then criticised for ignoring his guests. In contrast, the newcomers’ comments suggest that Hutton was seen more as an ambassador for the paper and should be answering questions on policy.

Thus, it becomes apparent that, while the quantative data revealed that there were some differences in the behavior of visitors experienced in other electronic discussion environments and newcomers, it is by looking at the comments made by the two groups that interesting differences in expectations are revealed. Both groups had positive things to say and criticisms to make, but their targets differed depending on their experience and, thus, their expectations.

One other interesting difference of approach emerged: between the responses of participants and of non-participants. There were many comments across the board about the political outlook of the debates. For example, from one of the experienced vistors: ‘Probably not a great idea to fragment discussion by having a forum on each newspaper … but ordinary mortals wouldn't visit uk.politics.’ Similarly from one of the newcomers: ‘the only disadvantage I can see to Newspaper (Broadsheet) sponsored discussion groups is that people appear to choose a paper that reflects their Political views. Therefore there is a danger of the debate becoming sterile - "I fink Maggie Thatcher woz a Git", "Nah, She wuz a mega-Git" "No, No, thrice No, she's a humungous git" - and so on’.

Consequently, several postings from participants were concerned with the lack of lively views in the forums and how the newspaper might stimulate debate: ‘Subtle editorial "provocation" or "interference" is actually quite good in these cases. Normally, no-one from the editorial department of a newspaper would dream of directly commenting on a letter writer's views (unless the person is famous, and their letter is newsworthy), but discussion groups are less formal and response is expected and required.’

Non-participants were also concerned with the liveliness of the experience, but their general comment was that they tended to get bored by the repetition and wordiness in the forums and - sometimes grudgingly - suggested editing1: ‘Somehow compelling but views of the punters never as interesting as the pundits.’ and: ‘The difference between online discussion sites and newspaper letters is that the latter are edited for brevity. The problem with online forums, as I see it, is that there are too many people talking the same thing, which makes for tautological reading. I sometimes wish someone could edit the content! (Do I hear cries of ‘Censorship!!’ ?).’

Discussion

There are many tentative conclusions to be drawn from the findings above, both in terms of providing suggestions for developing this kind of forum and also in developing other models for electronic discussion.

Firstly, we can note that some 200 people contributed the bulk of the discussion to these forums and these contributors seemed to get pleasure from exchanging ideas and information with each other. This was not a large number of people compared to the general website traffic, but it is a sizeable group to have an online discussion with. It implies that the website’s structure for posting was adequate to keep a discussion flowing, and that the content was motivating. However there were some drawbacks. Within this group there were identifiably different expectations of fellow participants and the host and this stimulated a few negative comments from experienced visitors, such as those about newcomers and their behavior. Equally, some newcomers described problems: for instance, admitting to feeling too intimidated to post. This implies that as with other forums, certain individuals’ messages will have more presence than others.

The specific content of the wider election site did not dominate the discussions in the forums, although visitors were very well-informed. More significant was the involvement of an editor on a near daily basis. As the questionnaire revealed, this was generally welcomed and much commented on, though it also led to some disappointment since his contributions tended to be brief. Furthermore, he did not seem to follow issues up in the actual newspaper of which he is editor. Nonetheless, The Guardian’s involvement of a senior editorial figure in discussions began to exploit the interesting position in which media providers can now find themselves, as both the public's desired audience and its impartial host.

Lessons Learned

The Guardian’s fielding of an editor to host a forum appears naïve in retrospect, because the commitment required taking time away from other duties during the busy run-up to an election in order to participate at the level many respondents expected. As it was, many of the criticisms of the forums refer to the perception that he did not fulfil his role, whether he was expected to be super-participant or ambassador for the paper. He could have limited people’s expectations and avoided fueling ambiguity by defining his role in his opening statement (though evidence elsewhere suggested that visitors ignored explanations in favor of using their expectations, so this might have made no difference). Equally, his involvement might have been more satisfying if it had been condensed into a shorter period of greater interaction, such as question and answer sessions at key points.

There was only a little confusion among participants about the role of forums: whether it was a place to discuss with others or to communicate with the paper. Most were clearer about what to expect from the forums than the developers themselves had been at the start. They had a vision of egalitarianism and free comment that sat uncomfortably with the commercial necessity of authoritative media and, thus, a heavily branded, high-status and compelling site.

Another factor affecting the level of interaction in the electronic discussion was the topic under discussion: politics. In the first place, the forums were designed around an event: an election campaign, which runs a specific course. Secondly, politics is an area where the expression of an opinion is closely linked with notions of democracy. Hence posting an opinion, aligning with a political group as represented by one of the British newspapers (which are partisan) and exerting influence may be seen as desirable goals by some of the contributors. This is specifically shown in a couple of participants’ calls for the paper to campaign on issues raised in the forums, but it is also implicit in other exchanges. Participants taking part in forums on other topics, like the plot of a soap opera, might have quite different expectations. For example, they may see their involvement only in terms of being part of a community, a support group or as a place to make new friends, rather than, more loftily, to be involved in British politics.

Because the host site was partisan, a major content-related criticism was the lack of different views expressed in the forums, though this was also seen as a virtue as it avoided the appearance of intransigence. The liberal left identity of the paper drew like-minded visitors, many of whom relished reading news and opinions with which they agreed. No one remarked on it, but disagreeing with a posting had motivated many to post despite this narrowness, so it may also have been seen as a safe environment to air differences, where nuances could be grasped and no one lambasted for general beliefs. This nonaggressive stance – despite an intimidating intellectualism – appears to have encouraged postings.

Our findings suggest that this kind of website branding determines who visits it and how they interact with each other. Some of the election site visitors were merely pleased at the presence of the forums, regarding them as a symbol of The Guardian’s democratic instincts; they didn’t need to use them. Hence, their identification was at a level above participation.

While the forums ran, the developers were disappointed with the numbers visiting them and becoming involved. What would have made them more popular? At one level, technical problems like difficulty in manipulating the customized software impacted on level of usage. Other visitors, not part of the questionnaire survey, suggested that the labeling of the forums was poor. For example, ‘The Debating Chamber’ sounded intimidating or obscure. Therefore more user-friendly terms and metaphors might have encouraged others to post.

Not requiring registration may also increase the numbers of visitors and participants, since our study suggests that a larger proportion of the 70,000 visitors to The Guardian election website would have joined the 2100 who registered for the forums, had it not been compulsory. Registration is known to be unpopular; many people do not like providing personal information and others are deterred by the effort involved. The ratio of participants to non-participants was about 1:5, but nearer 1:3 if we discount once-only visitors to the forums. This ratio will have been affected by the demand for registration even to read postings. However, the trade-off for not having to register – which The Guardian was only too aware of – was that it might invite libelous postings if a free-for-all postings policy was adopted. In this context, it was better for the website to demand accountability for content, to lesson the risk of irresponsible posting.

Registration did have its benefits, however. For example, it discouraged casual posting of trivial or repetitive messages which might have overwhelmed the simple threading system.

Finally, how might the group of visitors (including both participants and nonparticipants) be described, since they have the characteristics neither of a virtual community or participatory genre [5,6]? In many ways, they were brought together through the branding of the site and their association with The Guardian newspaper. From the developer’s perspective, it managed to build loyalty to their brand.

Future Developments

The Guardian’s early experiment in electronic discussion spaces met with partial success: a small group of visitors was impressed by the idea of forums and a smaller group enjoyed using them. Some good discussions and a few friendships resulted. People were brought together on the Web by The Guardian and might never have found their way together otherwise. But this was incidental to The Guardian’s plans; the forums were a greater success for the developers in terms of providing food for thought.

The team who went on to redevelop and expand the site have since decided to use it primarily as a means of providing news updates with more limited forms of interactivity. The news site has been expanded, linking particular news issues with a section called ‘Have your Say’. This now offers a topical question that is published alongside visitor responses. These responses have been chosen for posting by an editor from those submitted as e-mails through the site. In so doing, the publisher has changed the nature of the relationship between news provider and visitors to one that is now more similar to that of other broadcasting websites, e.g. the BBC's. The original goal of providing interactivity, like the forums, to establish a large number of loyal visitors proved to be elusive and was abandoned. Providing news reading as the core activity is considered to be a more effective way of bringing people back to the site.

The new website has made reading contributor postings straightforward, like accessing a further page of news. It has provided a very easy-to-use posting mechanism, involving mediation by the editor. A benefit of this 'filtering' model is that it provides status and feedback to those whose postings are selected: there is the cachet of seeing one's comment appear. The disadvantage of this approach for contributors, however, is that they have no choice but to publish or pass by; a good discussion is impossible. Although the audience may respond to the news, it is more like writing a reaction piece, and is far removed from the earlier attempts at providing a forum for negotiating opinion with the producers or with each other. In this respect it has become much more staged, more similar to the kinds of interaction which readers have with newspaper and television producers, but easier to conduct. Participants who want to exert an influence may be catered to, but those who want to argue or share beliefs appear to be excluded.

Conclusion

Our study of The Guardian website came early in the development of commercial interests on the Web. It showed evidence of different kinds of visitors to website discussion forums, having different expectations and interacting in different ways. The key distinguishing features were:

  • newcomers to electronic discussion spaces have different expectations of what they want to do in the forums compared to those who are familiar with other electronic discussion environments; they are less discursive, more likely to post in order to present opinions, and more likely to visit the forums again to read other people's contributions

  • non-participants - both those experienced and those new to forums - were more likely to support the idea that postings should be edited to make them more informative, rather than lengthy, repetitive exchanges. Participants, on the other hand, were more likely to have the view that Web culture should uphold freedom of speech.

The number of visitors coming to the Web with little or no experience of electronic discussion has risen since the study and will soon be dwarfed by a further wave, so we can expect the number passing through news websites relative to the people who have been involved in newsgroups to increase substantially. This suggests that news and politics websites will be visited by many people who want a snapshot of the state of affairs, some searchable history and a section showing other people's responses, with an option but no compulsion to contribute to this. It is likely that proportionately fewer people will want to argue online with other people about events. It would seem, therefore, that despite a move toward developing interactivity on publishers’ websites, there is still going to be a largely silent audience reading the words of a vociferous few. It follows that web forums which adopt a degree of control over what gets published will end up being more successful. This may not be the future of democracy, but perhaps the future of readability.

In Web forums that support asynchronous text exchanges, our study suggests, public discussion of politics and news will continue to be a minority activity. The Web accelerated the migration of publishing to the Internet, extending its scope so that more people had the means to publish. Some are doing so, posting opinions and hoping that others will read them. However, a change in both attitudes and opportunities will be needed if it is also to extend the art of conversation.

Acknowledgments

This study could not have been conducted without the help of Robin Hunt of The Guardian New Media Laboratory and his team; in particular Emma Caldwell, Rebecca Fox and Ali Gordon.

The Sites

The Guardian Debating Chamber is no longer open to the public, but the main site can be found at:

http://www.Guardian.co.uk

BBC Talking Point at: http://news.bbc.co.uk

CNN Interactive at: http://www.cnn.com

Footnotes

1 On one occasion The Guardian did invoke its powers of moderation to remove a strand of discussion about a libel case involving two other media parties. This brought protest from participants in the forum and was brought up again in responses to the questionnaire.

References

[1] J. Nielsen, "Loyalty on the Web," Alertbox, August 1, 1997.

[2] A. Light, Expectations of New Media: An Evaluation of the Use and Usability of The Guardian’s Election ’97 Website,, doctoral dissertation, 1998.

[3] R. Riedl, "Patterns in Computer-mediated Discussions," Mindweave, R. Mason and A. Kaye, eds., Pergamon Press, Oxford, 1989.

[4] R. Bentley and P. Dourish, "Medium versus Mechanism: Supporting Collaboration Through Customisation," Proceedings of the Fourth European Conference on Computer-supported Cooperative Work, H. Marmolin, Y. Sundblad, and K. Schmidt, eds., Kluwer Academic, Dordrecht, 1995.

[5] H. Rheingold, The Virtual Community: Finding Connection in a Computerised World, Secker and Warburg, London, 1994.

[6] T. Erickson, "Social Interaction on the Net: Virtual Community as Particpatory Genre," Proceedings of the 30th Hawaii International Conference on Systems Sciences, Hawaii, January 1996.

About the Authors

Ann Light is completing a Ph.Dd in the School of Cognitive and Computing Sciences at Sussex University in the UK, following a career as a political journalist. Her research concerns interaction between producers and users on the Web, looking at issues of design and communication. Her academic publications explore electronic publishing, the use of the Web in teaching, user-centred design and methodological questions.

Yvonne Rogersis Associate Professor in the School of Cognitive and Computing Sciences at Sussex University in the UK, where she teaches Human Computer Interaction and Cognitive Science. She has published widely in the area of HCI and CSCW, including the use of field studies and distributed cognition to inform system design and the design of interactive representations in software and web-based applications. A common focus of her research is concerned with developing a theoretical account of the 'external cognition' that occurs when we create, manipulate and read different representations for various kinds of activities (e.g. learning, problem-solving, communicating).

Copyright Information

Copyright © 1999 Institute of Electrical and Electronics Engineers. Reprinted from the CD ROM-based "Proceedings of the Thirty-Second Annual Hawaii International Conference on Systems Sciences," 1999 (January 5-8, 1999, Maui, Hawaii. This material is posted here with permission of the IEEE. Internal or personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution must be obtained from the IEEE by sending a blank e-mail message to info.pub.permission@ieee.org. By choosing to view this document, you agree to all provisions of the copyright laws protecting it.