The online debate viewing experience today

Some of the analyses and visualisations that we aim at including as enhancements to the debate viewing experience are already available to viewers.

In a previous post, we referred to Demos‘s analysis of tweets in terms of the personality and politics of the Nick Clegg and Nigel Farage in the second EU debate. As another example, Blurrt performed real-time sentiment analysis on the tweets during the first EU debate, producing the LBC Twitter Worm. This was shown live on the radio’s website next to the video of the debate.  The radio had its own live-blogging with comments from their politics correspondents, viewers’ comments were welcome on their website and also on their Youtube channel. There was a voting systems for viewers to express positive or negatively on the performance of each leaders at different points in the debate. In addition, Full Fact checked in real time the correctness of leaders’ claims . They tweeted their findings over the course of the debate and live-blogged in a dedicated section of LBC Radio’s website.


After the debate, Full Fact produced an indexed video which allows viewers to jump through the audience’s questions. Their findings are synchronised with the video and update automatically as the debate unfolds.


During the second debate, the BBC live-blogged political comment, select tweets and fact-checking on the politicians claims. Along the debate, these contributions were added in a single stream under the heading ‘Live Text’. The site also welcomed comments from the viewers which on occasion would be reproduced on the Live Text stream.

On both occasions, interested viewers had at their disposal a constellation of possibilities for enriching their experience. Just a few of these resources, updating constantly and independently as the debate progresses, can be overwhelming. One of our questions is then how to combine these sources of information in ways that facilitate – rather than potentially obscure – citizen’s access to the already complex debating activity.





On the 2015 General Election Debates format

As the BBC reports, it seems that the politicians are thinking about having three televised debates in the run-up to the next general election: one between the Conservative and Labour leaders; a second with the Labour, Liberal Democrats and Conservative leaders; and a third including the UKIP and Green Party leaders.

Why not? The more opportunities voters have to evaluate the parties and their leaders the better. I would go further; let’s think about different formats for each of the debates. Some of the participants in our focus groups came up with great ideas for improving the format. While politicians are busy negotiating who goes where on the debate platform, thought needs to be given to ensuring that the debates provide those who watch them with what they need and want, not just what suits the would-be leaders.

Clegg-Farage Debate Mapping: Prelim Steps

The televised Clegg-Farage debates on Europe [parts 1 and 2] provide a corpus of material to work with as we begin to test out ideas. Social media analyses, for instance, of the twitter stream, are being shared, such as this one from Demos which gives a very coarse grained classification of tweets as related to personality or politics:

Screen Shot 2014-04-04 at 21.01.53

The media is saturated with commentaries on the personalities that came through, who had ‘won’, and in some cases, the quality of the debate. In KMi, we’ve started exploring reviewing how journalists are making sense of the debate, and we’ve begun mapping ideas.

If we take the second debate, one could simply map the transcript visually. However, this does not really add much value: the nodes are too large to reveal interesting structure:


What’s needed is a finer grained breakdown of the moves. In our 2010 General Election maps we did Dialogue Maps driven by the questions, and the order of contender responses. These were produced primarily in real-time, as the broadcasts went out, with minimal editing. They seemed, from informal inspection and feedback from others, to be useful in showing more clearly what happened in a given session.

However, this time we’re aiming to do a more careful analysis. In this map, taken from the first 15mins of the transcript, we show which claims are being made and by whom (top), and who is supporting and challenging them. The order in which contributions were made is less influential, although can still be seen to some degree. The goal is to connect contributions as clearly as possible, so if Nick Clegg makes the same point three times, it might not appear as three nodes, but I might simply add quotations inside the detail of a summary node. See the Key at the bottom to read this map.

It’s not clear if we would want to render arguments like this to end-users, although feedback from the detailed focus groups run in January suggests that citizens would value ways to get an overview of such content. The first step is to understand what was going on in the debates, at least through this quite rational lens. (Screenshot below / Interactive Map)


More to follow!…

What would viewers want to say if they could?

We’re using the Clegg-Farage Debate on Europe as a testbed for some new ideas. One is to elicit from viewers how they’re responding to the debate, and what they would say if they could via a next generation user interface.

The Leeds University team ran a group this week in which participants raised specially designed flash cards as the debate unfolded. Here’s some of the deck:

TV Debate Flash Cards for Viewer Reaction

— more news once we’ve analysed the session video…

Screen Shot 2014-04-24 at 22.30.30

Stephen Coleman gives evidence to House of Lords

Screen Shot 2014-04-23 at 11.52.59

EDV project lead Professor Stephen Coleman gave evidence to House of Lords’ Communications Select Committee Inquiry on Broadcast General Election Debates [Hansard transcript pdf].

As the clips below reflect, there is considerable interest in the potential of the web to help engage and educate citizens around the envisaged Prime Ministerial debates.

p.84 — Professor Stephen Coleman and Tim Gardam – Oral evidence (QQ 86-103)

Q99 Bishop of Norwich: We heard quite a bit about the way the US debates commission promotes awareness about the American debates and spreads knowledge of them, yet in Stewart Purvis’s evidence last week he said he thought that the broadcasters here had not promoted the debates very assiduously, that there was not much viewer education, and that there was no debates portal. Are there other ways other than social media in which the broadcasters themselves could extend the influence of the debates in relation to the campaign?

Professor Coleman: This is a very important question and was one of the most conspicuous gaps after the last election. I suspect that many people among the broadcasters were very nervous about these debates, perhaps wondering whether, after the expenses scandal, they were going to be taken seriously and whether they were going to be primarily for political insiders. They were a popular success. There is a tremendous opportunity for the public service broadcasters to get together now and use these debates to think very imaginatively and expansively about what the public needs around an election and the deliberation involved in the debate but also the information involved in voting itself.

I think there are a number of things that they could do. Websites are probably still the main source of good voter information. There is an opportunity to create opportunities for people to build Wiki-like aggregations of information and local knowledge. I think there are opportunities, and my university has been funded with the Open University to do this: to build platforms in which the claims made in the debates and the records of the debaters are open to public scrutiny. I think this is a very, very big one for the public service broadcaster, for all the broadcasters, to get together on and they probably need to start now.


Q102 Baroness Healy of Primrose Hill: With the millions who watched these debates, which was a very good thing, could greater public value be achieved from the broadcasters being encouraged to co-ordinate their activity? I know you have spoken about them possibly coming together around the broadcast general election debates, for example in a one-stop shop online for information and education about the debates and potentially live streams and archival content so that everybody can go back and find them. I am interested in what you think their role would be with 3 million, at least, not on the electoral register. These people, young people particularly, are watching and then they find they cannot vote in the end. Would it be a civic duty perhaps for the broadcasters to encourage them to make sure they are on the electoral register in their trailers for the debates perhaps?

Tim Gardam: It is important to amplify the points we were discussing a little while ago, which is that the broadcasters were taken by surprise by the resonance of the debates last time. That was my impression. I think, talking to them, that they would say the same. There was a nervousness about how popular they would be, although no one doubted their significance. What one then makes of that material as a broadcaster needs careful thought. It is normally the way of broadcasters that they go off on their own to work out what they can do with it.

Coming back to the points that we were making just now, in pure terms of ease of access, discoverability and the clear incentive to build a common point of access and, as I said, to open it up for material beyond that of the broadcasters, this would be a very, very powerful educational tool. I do not know how much thinking has gone on there. I know that the BBC is talking about the need to make its educational strategy much more self-evident. It seems, going back to the point I made before, that in the area of politics few institutions have their independence and impartiality so enshrined as to make such a tool seen to be something for the public good as opposed to something that may have implications of control.

Professor Coleman: There is something rather unusual happening here, which is that for once the universities might have a bit more money than the broadcasters to do this sort of thing. We have been given quite a large pot of funding to produce a platform for public education after the debates, if they happen next year. What we cannot do as universities— the University of Leeds is working in collaboration with the Open University on this—is what the broadcasters can do well, and that is publicise it and bring to it some of the creative design energy that we academics do not always bring to things. I think there is a tremendous opportunity here and it is because one of the research councils was imaginative enough to support this idea that we are in a position to be able to do this.

Focus Groups: how could the Election Debates be improved?

Screen Shot 2014-04-11 at 19.10.38

The Leeds team (Stephen Coleman, Giles Moss and Jennifer Carlberg) have conducted a series of twelve focus group interviews, with groups ranging from the politically disengaged to party supporters to ask them about their experiences of the last televised elections debates in 2010 and how they think they can be improved next time. The aim is to identify some key themes – some of which will relate to all people and others to certain groups – and then to theorize how the informational and communicative capabilities of citizens might be served by new forms of election debate visualisation. A full report of the findings will be made available, but here are some brief reflections on what focus group members said:

i) People who watched the debates found them helpful in understanding the policies of the main parties and the characters of future Government leaders (there was a demand for both).

ii) Some people found the debates hard to follow and many believed that the debaters were being too strategic and didn’t work hard enough to be clear about their records or the ramifications of their policies. Several participants expressed concern about being addressed in ways designed to confuse and manipulate, but linked this directly to a desire to be able to scrutinise rhetorical and non-verbal strategies adopted by speakers and relate these to the speakers’ implicit values or explicit claims.

iii) Most people wanted a way of scrutinising claims made by the debaters and checking facts – especially when they related to their own lives. They were eager to have opportunities to evaluate political claims with reference to compelling background information, counter or correct false political claims, and evaluate the congruence between what is said by each speaker in the debate and the speaker’s integrity as a political actor. In short, they wanted help in understanding the meaning, background and historical record of political claims.

iv) Many people wanted to know more about the party leaders: where they came from; how they arrived at their political values; how far their actions and words were consistent.

v) Many participants were keen to have a sense of involvement in the proceedings and their broader consequences, including public discussion (live or after the event) and scope for efficacious action. They were particularly concerned to ensure that questions came at the debaters from as broad a range of people as possible without them knowing what to expect. While some participants wanted to see instant feedback from the public during the debates, most seemed  not to be very concerned about creating new ways of or spaces for discussing the debates; they felt that social media already provide that.

vi) Participants wanted opportunities to tell the speakers about their lives, values and preferences. In particular, they wanted to see how claims made in the debates affected their regions and their personal lives. Several also favoured making short films, giving the debaters an insight into their lives – or having the debates targeted to socio-demographics rather than policy themes.

EDV project launch: Visualizing the 2015 Prime Ministerial TV Debates

Screen Shot 2014-03-18 at 07.39.49During the 2010 UK general election, the first ever televised Prime Ministerial debates took place. University of Leeds and KMi research investigated public reaction to these debates, the roles they may play in democratic engagement, and the potential of mapping the debates visually. In 2015 the next election is expected, providing the opportunity to investigate how new kinds of knowledge media can deliver completely new ways to replay the debates and engage with the arguments at stake.

Funded by the EPSRC, this joint 3 year project has just kicked off jointly with University of Leeds (led by Stephen Coleman, Professor of Political Communication at the Institute of Communications Studies, with Giles Smith and Paul Wilson), and will investigate novel approaches to making the debates replayable for citizens, using their input to shape requirements, and computational techniques to analyse what happens in the debates.

The project team bring a unique combination of Information Science, Political Communication and Design (Leeds) with Computer-Supported Argumentation Visualization (KMi: Simon Buckingham Shum, Anna De Liddo and Brian Plüss).

University of Leeds coordinated in-depth research into citizen reaction to the televised debates (Leaders in the Living Room: PDF). This showed that:

  • the British public appreciated the debates
  • 2/3 said they’d learnt something new
  • they seemed to energise first-time voters
  • people would talk about them afterwards (esp younger voters)
  • media coverage shifted from focusing on the ‘game’ to the substance of the debates

Thus, while there was a significant public appetite for this means of learning about the candidates and their policies, many viewers were left feeling uncertain about the meaning of and relationship between the competing arguments they had witnessed.

Based on KMi’s real time mapping of the debates, and drawing inspiration from other projects that experimented with augmented TV replays, the project will develop a web application offering user interface visualizations of the debate replays, novel analytics on the argumentation, making the argument analysis available as open data for others to render or mash up in new ways. Extensive focus groups and user evaluations prior to, during and after the election will advance our understanding of how to make complex societal debates more accessible.