Illness, return to work and yet more articles

Having been ill for the last ten days I returned to work yesterday and today have been back working on my masters. Frustrating to have such a long stint off but not in my control & I am quite pleased with what I have managed to achieve already since returning to the office.

I have found another 25plus articles in the last 24 hours, some of which are really excellent. In particular, it is encouraging to have found more articles on religiosity and internet use & also viral marketing - Word of Mouth (WOM) and electronic WOM (eWOM).

Having had some serious problems with the rate at which I could collect data for the samples I have found another suitable network to examine which is a more manageable size. This has allowed me to finish gathering the data overnight yesterday and today and so I finally have two complete networks which can be compared and studied. I am satisfied temporarily!!

My next task is to create a table showing all the data types that I have already and all than be collected and directly compared and contrasted. Obviously, the other main task is writing my Literature Review draft which is due end of February 2011. Please notice the deliberate mistake...

Progress, setbacks and large amounts of reading

Having had a good meeting with my supervisors yesterday, they are happy with how things are progressing and we have some exciting avenues to pursue with the data captured so far.

I had a bit of a setback yesterday when in the ninth hour of accessing the data source I made a mistake and caused the whole thing to fail! This has meant that I have started again from the beginning - frustrating, but not the end of the world. I am now six hours into the process again (being extremely careful not to make the same mistake twice) and hopfully should be done tonight/tomorrow.

Whilst my computer has been off periodically because of this accessing I have been doing a lot of reading. Today I read one of the most useful articles I have found so far: Twitter Power - tweets as electronic word of mouth, which covers lots of the areas I am; with similar tests planned and exposition of research in word of mouth, branding and viral marketing. I have also found a useful paper looking at religiosity affecting use of internet and social network sites in particular which is great for further references.

My aim for the day has been to finish the data aquisition and I'm not quite there but "you can't hurry [data], no you'll just have to wait, [data] don't come easy, its a game of give and take, no you can't hurry [data]..." etc. etc. Anyone reading this now thinks I'm mad unless you get the reference? Answers on a postcard please..

Big ideas...

I've just been thinking about the type of social network study I am reading about - where twitter or facebook are used and a very specific interaction is investigated. Some studies also examine in much more general terms the way in which that network functions. But why not approach these questions from another perspective - people.

Let me explain. Think about this: my egocentric social network is far bigger than my twitter network. My egocentric network on twitter is completely unrepresentative of my actual social network. At best it only represents those people that I know who are on twitter.

In reality, my ecocentric network is actually a whole mass of interconnected people, who I communicate with via Twitter, via Facebook, by email, by phone, by texting to a large extent, through letters very occasionally and most frequent of all face to face. It is also constantly changing and evolving both in how I contact them and who I am contacting and why.

I would love to see a comprehensive study of specific people's egocentric networks. Perhaps taking 10s or 100s of people. Imagine the diversity of methods used to communicate - imagine the potential findings and the document of our current communication forms that this would form! This sample could then be repeated years later and potentially conducted again multiple times over the years.

What would it look like?

Interview the participants in detail to identify the forms of communication they use. Use samples of people from deliberately chosen ages and demographics - for instance in the university it could be undergrad, postgrad, lecturer and admin person for instance... The networks within these various communciation forms could then be examined individually i.e. for each person you would come out with their twitter, facebook, email network etc etc and you would then end up with a comprehensive snapshot of the technology they used at that time and who they contacted, how etc. Each network could be analysed for its own merit but then also cross-analysed with each person's other networks and then between people or different groups of people...



It's probably wildly overcomplex & time intensive but anyway, just an idea!

Literature Review Update

Today's aim was to gather more research articles for my Literature Review which I am working on whilst also gathering data for analysis (alluded to below with pretty pictures..).

Apart from a saga involving an ATM eating my bank card which lasted 3 hours (all I wanted was a bite to eat for lunch) I've made some good progress. Some of the most interesting articles which I have located today have been on Twitter; how it is used, how it may be used in the future (as well as the web as whole) and how it was formed. I am planning a section on Microblogging and Twitter itself for my LR. As well as more general research, it is important that I present the academic research already carried out on and using Twitter.

Another of the more useful articles is on Religions going online (although slightly dated as it is 10 years old) and I have found a potentially interesting tool for improving the data collection process: VOSON (Virtual Observatory for the Study of Online Networks) which is a tool developed by The Australian National University. This can dovetail with the NodeXl addition to Excel (mentioned in previous posts) which could be really useful - we shall see.

Further Social Network Visualisations

Today I've been able to get quite a lot more data for eco-centric networks. The images on the left show some more complex social networks (not particularly well arranged) which give an insight into the sort of visualisations that can be created using the NodeXl tool.

These are both simple eco-centric networks which show how one user (me for example) follows others, these others may also be following me. Some of the people I follow also follow each other - hence the rather hectic lines connecting all the different actors in the network. These samples are limited to >200 people but each dot represents an account or person. The size of the dot in these visualisations has been defined by the 'betweeness centrality' value which it has within the network i.e. '..how many of the shortest paths between two others some actor lies on'. The colour of the dots has been defined by their Eigenvector Centrality - how important that actor is within the network. These can now be used to compare with other networks and to represent many other characteristics of the network visually.

Return to posting with Node XL success

Have had a successful day working with the social networking visualisation tool NodeXl which allows you to analyse social media networks. Twitter is the source of the network data and today I have finally got some data into the worksheet and been able to play around with it!

My supervisor will be happy to know that I have now mapped her ecocentric social network on Twitter!

GE2010 Extended Article. Part ii: The Social Networking Sites

At the time of the last General Election in the United Kingdom there were less than 5 million users of social networking site Facebook, compared with over 400 million in 2010 (Facebook, 2010). Twitter did not exist and YouTube was an embryonic company with its first video upload on 23rd April 2005. Although there were social networks and blogs available, the world-wide take up of social networking sites has only occurred in the last 5 years.

The Social Networking Sites

Each of the three main political parties had a presence on the main social network sites of Facebook and Twitter. They did not have a maintained presence on more than a handful of sites, particularly those that have the potential to reach voters in hard to reach demographics. The 2008 Obama for America campaign used many more in comparison:

"Pulling out all the Web 2.0 stops, [Obama's] campaign used not only Facebook and Youtube, but also MySpace, Twitter, Flickr, Digg, BlackPlanet, LinkedIn, AsianAve, MiGente, Glee - and others." (Fraser & Dutta, Guardian 2008.)

The main web sites of the Liberal Democrat and Labour Party's contained links to their content on Youtube, Facebook and Twitter. The Conservatives’ contained links to the same as well as links to ‘Livestream’, ‘delicious’ links and their podcasts on itunes. (Sources - the party websites)

Facebook

The three main parties had Facebook pages before the election campaign started but the changes during the campaign in terms of Facebook users ‘becoming fans’ or ‘liking’ each of the parties (Facebook changed its system from ‘fans’ to ‘likes’ during the course of the election campaign) are worthy of note. The content of the Facebook pages was roughly the same for each of the parties. Each contained links to their external websites, regular postings of latest videos created or events, as well as space to engage either through Facebook comments or by linking to an external mailing list or volunteering to work with the party in users’ local area through the organisational sites for each of the parties.

In terms of numbers of fans or ‘likes’, those recorded from the pages on 15th April were Conservatives (43,000), Labour (20,000) and Liberal Democrats (20,000). Contrast this with the 21st April where the Conservatives had climbed to 54,000, with the Liberal Democrats at 52,000 and the Labour party trailing at only 27,000. This trend in support for the Conservatives and the Liberal Democrats over Labour in active Facebook users was continued, as reported by the Guardian on 3rd May:

“The Conservatives were again leading with 74,500 over the Liberal Democrats with 74,000 and the Labour Party trailing at only 33,000.” (Charles Arthur, The Guardian 3rd May 2010).

While this is not a mathematically verifiable indicator (users can ‘like’ more than one party, ‘likes’ do not necessarily guarantee support or votes etc.) it acts as anecdotal evidence of the level of engagement and more importantly indicates the numbers of Facebook users who would have been receiving daily posts and updates from the parties via the social networking site.

Whilst all parties had pages about themselves, Gordon Brown did not have an official page for labour party leader where both Nick Clegg and David Cameron did so and gained the added reach and promoting power that those connections gave.
Facebook users themselves engaged in their own ways – beyond the control of any of the parties. Some of the strongest user-generated ‘viral’ campaigns on Facebook during the General Election were anti-Conservative Party campaigns. The poster campaigns launched by the Conservatives were quickly rubbished online and a group formed to share and comment on them, as Charles Arthur of the Guardian wrote:

“The social network has dozens of groups relating to the parties: though the "Vandalised Conservative Billboards" group, with 123,000 members, outranks the Conservatives' own (74,500)” (Charles Arthur, Guardian 3rd May 2010).

Another of the most popular groups on Facebook in terms of members was the “National not voting conservatives day” group created with 146,000 supporters on the May 5th (BlueRubicon, Facebook Election 2010 Stats). This was a popular protest movement against what the Conservatives had been trying to do on Facebook by raising their own groups for support on May 6th.

Facebook used its influence as a network to engage UK users of the site by setting up Democracy UK. This page was a deliberate move to act as a central hub for the information pertaining to the election and acted as a single space through which users could find out up to date information. Through a similar but separate effort ‘Voter Registration’ in partnership with the Electoral Commission, over 14,000 extra voter registration forms were downloaded through the site. The BBC’s Digital Election Correspondent Rory Cellan Jones commented on the use of Facebook during the election:

“Facebook, a much bigger network than Twitter, seemed uncertain how to engage its users at the beginning of the campaign - but by the end they were flocking to political pages, creating groups and generally showing more enthusiasm for politics than might be expected of a group often caricatured as news-averse and self-obsessed. Facebook's tie-up with the Electoral Commission to promote voter registration also appears to have been a success, with visits to the registration site soaring after links appeared on users' profiles.” (Rory Cellan-Jones, 7th May 2010)


Twitter

The Social Networking Site Twitter was launched publicly in July 2006. The site is widely reported in the press for its growth and the instantaneous nature of the platform for publicly available comment and opinion. Figures released in early 2010 demonstrate the reach and growth of the network:

“According to worldwide comScore figures released today, Twitter’s own site attracted 73.5 million unique individuals in January, up 8 percent from December, 2009 (when it had 65.2 million visitors). Its annual growth rate is still a phenomenal 1,105 percent. A year ago, Twitter.com attracted only an estimated 6 million visitors.” (Erick Schonfeld on Feb 16, 2010, Tech Crunch)

Twitter was used both by individuals who represent the parties (including their ‘celebrity’ supporters) and by direct accounts created to represent the views of each party and to raise awareness of events and news items. In the build-up to the election Reuters employed a ‘sentiment analysis’ firm to analyse the support for each party on Twitter:

“Edward Schneider, a consultant at Crimson Hexagon, said: “Twitter reveals the Tories to be the most talked about, as well as the most polarising, party in recent weeks. Over 40 percent of Twitter conversation on UK politics in the last two weeks has concerned the Conservative Party, compared to only 28% for Labour.” (Reuters, April 6th 2010)

Whilst individual and party controlled Twitter accounts created and disseminated partisan viewpoints, there were some serious consequences to using social media during the campaigns. Perhaps the most widely reported of these were the use of said social media tools, particularly Twitter and via the aforementioned Facebook group, to discredit and undermine the Conservative Party’s pre-election poster campaign:

“...the Tories have also found social media can be used to their cost. In January they launched their first campaign poster, featuring Mr Cameron’s face and the caption: “We can’t go on like this. I’ll cut the deficit, not the NHS.” In campaigns past, voters would have liked or lumped it – and told their friends. Yet with photo-editing software and a Twitter account, now they tell the country. Within days, doctored versions mocking Mr Cameron and his message were all over the web, TV and newspapers.” (Jon Swaine, Telegraph 6th April 2010.)

Candidate suspensions were another high-profile consequence of opinions expressed through social network sites. Both the Labour and Conservative Party’s had to suspend candidates standing in the General Election. The Scottish newspaper ‘The Daily Record’ reported:

“Stuart MacLennan became the first Twitter casualty of the general election when he was axed by the party for his offensive behaviour. Scottish Secretary Jim Murphy took the decision to fire MacLennan as the party's candidate in Moray after reading the online comments in which the 24-year-old branded one woman a "boot" and joked about slavery.” (Torcuil Crichton, Daily Record 10th April 2010.)

Reuters reported that “The party withdrew its support for Philip Lardner's candidacy in the Scottish constituency of North Ayrshire and Arran after he wrote: "Homosexuality is not 'normal behaviour'." (Reuters, 28th April 2010). The freedom of speech and information available publicly on the social networking sites allowed individual voters and the wider press to examine and follow candidates, increasing the ‘spotlight’ under which candidates and pundits alike were scrutinised.

Three Prime Ministerial Debates were held for the first time during the course of the General Election Campaign. Twitter played an important role in reporting on reaction to the debates and how each of the leaders fared. Rory Cellan-Jones wrote, of the engagement with the Prime Ministerial Debates:

“...a growing number of people have followed these major events via two screens - a laptop perched on their knees as they sit in front of the television, connecting with social networks to comment live on what they're seeing. This has in turn made possible an industry of what's being called "sentiment analysis", which attempts to understand what has been said online.” (Cellan-Jones, 30th April 2010)

Tweetminster, a site which exists to analyse and encourage political engagement and current affairs on Twitter actively recorded data from tweets during the debates. According to their figures, in the first debate over 36,000 people ‘tweeted’ about the debate, with a peak of 41 ‘tweets’ a second. In the final debate of the three the engagement level was lower at 33,000 people, although this was higher than the second debate. The ‘sentiment’ (an internal algorithm is used to measure the amount of positive tweets vs negative tweets to gauge overall reaction to an individual or topic) indicated that Nick Clegg received roughly the most positive reaction over all three of the debates.

The debates saw a high level of engagement by around 30,000 individuals on Twitter. This does not appear to be a significant number as a percentage of the population of the UK, or even compared to the twenty-three million estimated UK users of Facebook, however polls can be created extrapolating views from one thousand people that accurately reflect the population of the USA (NCPP), so it does have potential as a means to accurately gauge reaction. This reaction was largely out of control of the political parties, although does include the tweets from the parties themselves and their supporters to create these ‘sentiment’ statistics.

The main parties’ engagement was largely equal in use of Twitter, although some sources would indicate that the Labour Party had the most ‘influential’ users of the site. Prior to the General Election Campaign, the independent public relations agency Edelman released a list of the top two hundred political users of Twitter in terms of influence. Of the top twenty-five on the list ten were Labour (Edelman, 23rd April 2010), four were Conservatives and four were Liberal Democrat party members or representatives, indicating that within that particular sphere Labour had the most influential Twitter users of the three main parties.

This section has explored the use of Facebook and Twitter by the political parties. While these are two of the most widely used social networks, there are many others. The direct engagement by parties with other social networks was very low. The Conservative Party website for instance, included links to profiles on Facebook, Twitter and YouTube. If you compare this to President Obama's site 'Organize for America' (Formerly Obama for America) there are currently links to more than 15 different social networking sites. These are mostly US based sites, but are used by key voting groups within the electorate, often harder to reach through traditional media methods.