Vous êtes sur la page 1sur 32

Has New Media Provided Effective Ways

to Politically Motivate and Engage


Citizens?
A study of the use of Twitter in the 2012 US
Presidential Election Campaign

- Matthew Trollope, 4739523


- School of Political, Social and International
Studies
- April 2013
- Sanna Inthorn

Word Count: 8785. Including the text only, not including appendices, footnotes,
endnotes, bibliographies, graphs, charts, diagrams, tables and their labels. (UEA Word
Count Policy, 2013)

Table of Contents
List of Figures .......................................................................................................................................... 2
List of Tables ........................................................................................................................................... 2
Acknowledgements................................................................................................................................. 3
Introduction ............................................................................................................................................ 4
Literature Review .................................................................................................................................... 4
Twitter Analysis ..................................................................................................................................... 13
Methodology......................................................................................................................................... 14
Data Collection ...................................................................................................................................... 15
Reliability Testing .................................................................................................................................. 16
Research Questions .............................................................................................................................. 17
Findings ................................................................................................................................................. 18
Discussion.............................................................................................................................................. 22
Conclusion ............................................................................................................................................. 27
Appendices............................................................................................................................................ 29
Bibliography .......................................................................................................................................... 30

List of Figures
Figure 1. Screenshot of data coding spread sheet in Microsoft Excel with categorisations and
coding added.
Figure 2. List of users who have retweeted a message. Twitter. (2013). 'RT if you agree...'.
Available: https://twitter.com/BarackObama/status/311545647436275712. Last accessed
16th April 2013.
Figure 3. Retweeted Message. Twitter. (2013). @common_genious: RT if you agree.
Available: https://twitter.com/common_genius. Last accessed 16th April 2013.

List of Tables
Table 1. Statistics of total number and frequency of tweets and how many retweets were
achieved.
Table 2. Data for tweets posted that called for offline action from the reader.

Table 3. Data for tweets posted that called for online action from the reader.
Table 4. Data for tweets posted that amounted to one way communication with the reader.

Acknowledgements

I would like to thank my dissertation supervisor Giuseppe Veltri for his help whilst
completing this report, reading over drafts, providing me with materials and offering
technical guidance. Along with all the authors referenced in my bibliography I would like to
thank Twitter for making large amounts of content available and for providing an easy to
use, entertaining and exciting social media platform. Also, I would like to thank the websites
timeanddate.com and springfrog.com for providing tools to calculate the amount of hours
between two inputted times and the conversions between decimal and standard time
respectively.

Introduction

This dissertation will begin with a literature review that will explore the arguments
surrounding the effectiveness and use of new media in electoral campaigns. I will review
literature explaining how electoral campaigns have changed and how new media has
become a part of the new, permanent campaign. I will move on to literature explaining how
the internet has been used by political parties. There is much debate around how web 2.0
sites such as social networks are being used by political parties and whether these sites can
be utilised to motivate and engage voters, increasing motivation and participation due to
interactive messages or whether new media is merely a modernisation of traditional
political communication. The debates in the literature focus on whether online activity by
individuals is likely to translate in to offline activity due to the small amount of effort
needed to become connected with politics online. I will investigate how the Barack Obama
and Mitt Romney campaigns in 2012 used Twitter and whether it was used merely as a new
device for one way communication of political messages. I will assess whether the campaign
harnessed the power of the site to create an echo chamber, with interactive messages
inspiring discussion and debate in the attempt to mobilise individuals. Or even further than
this, whether campaign used Twitter as a call to arms to create strong ties with supporters
and inspire offline action or strong engagement with the campaign such as financial
donations.

Literature Review

Norris (1999) explains; since the late twentieth century we have experienced the
post-modern campaign. Norris attributes this to the emergence of a more independent,
dealigned press (1999, 23) coupled with a diversification of electronic media outlets (ibid)
which have led to political parties seeking new ways to control the media through strategic
management in what Norris calls the permanent campaign (ibid). One key facet of the
post-modern campaign era is the rise of electronic media and narrowcasting (1999, 33).
Narrowcasting enables citizens to choose what media they wish to consume from a large

number of sources. The internet is a perfect example of this, with huge amounts of sites
offering political information. At the time of writing, internet usage was not as widespread
as it is today, Norris concludes; the internet promises to provide new recourses for the
politically mobilized (1993, 34).
Margolis and Resnick (2000) explain that although the internet reduces
organisational costs of political groups, they do not believe the Internet will be as effective
as a campaigning tool, stating; the flowers are not blossoming in the way the optimists
hoped (2000, 54).
Between 1996 and 1998 the authors contacted webmasters of American political
party websites and found; none believed their websites had significant electoral impact
(2000, 65). The authors explain they were told that websites were seen by political parties
as an excellent means of communication, however they did not raise large sums of money.
Margolis and Resnick explain virtual party headquarters are no substitute for money and
organisation on the ground (2000, 66) and the existence of party websites serve merely as
demonstrations of modernity (ibid).
However, Howard Deans 2004 US Presidential campaign accumulated over
$15,000,000 in small donations after a successful use of online social networking site
MeetUp.com which allows users to create events and move their online groups into the
offline world (Wolf, 2004). This has set the trend for more extensive use of new media for
political communication and fund raising in electoral campaigning, particularly in the USA.
Margolis and Resnick conclude; the average person is not motivated or actively
engaged in politics and although the internet makes it easier for individuals to get
information and receive messages from political parties we cannot expect people to
overcome their habitual indifferences unless parties motivate them to follow politics more
closely (2000, 72).
Hindman (2009) builds upon this study; explaining traffic to political websites is very
slight compared to traditional media sources of political information. He points out most
people get their political information online from a few sites which they revisit often; this
does not give them access to oppositional arguments or challenge attitudes. Hindman states
5

in large part, citizens use search engines to seek out familiar sites and sources (2009, 132).
However, it is worth noting that for many people, social networks are sites they visit often,
these sites can expose users to the many different political views of their friends and from
the political news sources or parties they subscribe to.
Vaccari (2012) explains; electoral campaigns are finding new ways to improve the
reception and acceptance of their messages. The author explains the traditional view of new
media is that it merely reinforces political attitudes. Vaccari identifies the RAS (receiveaccept-sample) model first used by Zaller (1992), explaining there are three ways a political
message can be read and resisted; these are varying degrees of conscious and unconscious
resistance. See Vaccari (2012, 3). The most politically aware citizens will receive messages
but are less likely to accept them, and the least politically aware citizens are most likely to
accept messages but are less likely to receive them due to their media and news
consumption habits. The author explains; message intensity plays a part in how likely it is
that the message is accepted by individuals.
Vaccari (2012) explains that mundane internet tools (Nielsen, 2010) such as emails
from political parties are usually only accessed by those who are most politically aware who
have chosen to receive this media through mailing lists, potentially proving through the RAS
model that the online campaign has little other effect than reinforcement of views.
However, the author explains there have been qualitative changes in e-campaigning
such as the use of social networking sites. These sites offer a space where people can
publicly contribute to messages posted by politicians. The ability to view the actions of
friends on social media increases the likelihood of individuals encountering opposition
messages in non-political spaces and thus increases in intensity of messages. It is suggested
that e-campaigns have the potential to re personalise the campaign as individuals have the
ability to talk to politicians on social media sites (Gronbeck and Wiese, 2005, 529).
Vaccari (2012) used qualitative research methods to interview political consultants
about their use of new media. His findings were that the last few days of the campaign were
used to organise voters and get them to the polling station to vote, but for the most part,
new media use was for persuasion (2012, 7).

Vaccari (2012) explains the ability of a campaign to reach undecided voters depends
on how much effort the online platform demands from the user. For example, connecting to
a candidates page on Facebook is very easy; this could lead to more undecided voters being
reached by campaign messages, this is called a low threshold activity. These activities are
key for Vaccari as acts such as following a candidate on Twitter are not demanding and
therefore less motivated voters do not avoid these activities. Supporters can redistribute
messages to contacts which can create increasing reception of messages and possible
persuasion (2012, 11). The intensity of the electoral campaign message can increase
through redistribution by individuals as this serves as an endorsement. Interpersonal
communication can lead to higher reception rates and can also increase the acceptance
rates due to the message being communicated between friends who are trusted sources to
each other.
This is a central point in the study of the mobilising effect of new media in my
opinion. Shirky (2011, 4) recalls a study by Katz and Lazarsfeld (1955) looking into the mass
media effect at the 1948 US presidential election. They concluded opinion forming was a
two step process, firstly, opinions are transmitted by the media and then they are echoed by
friends and family who act as opinion leaders and increase the intensity of the message and
the likelihood of acceptance. Shirky (ibid) explains how new media is useful in this opinion
forming stage. When a politician shares a message on a social media site that encourages
responses, they will be able to create a buzz around a certain issue. When an individual
responds to this message, all the people they are connected to will be able to see both the
responding message from their acquaintance, and also the original message from the
politician. This creates an echo chamber and can raise awareness, as well as acceptance
rates of the campaign message.
Gladwell (2010) however, questions how effective the low threshold use of social
media actually is. Gladwell explores the recent Twitter Revolutions in Moldova and Iran
and explains Twitter was not as useful as it may have been first thought, suggesting the
majority of Tweets came from the West. Gladwell draws upon the work of Doug McAdam
(1993) and his work on the strong tie phenomenon. Gladwell explains one study of the
Red Brigades, the Italian terrorist group of the nineteen-seventies, found that seventy per
cent of recruits had at least one good friend already in the organization (2010). By
7

highlighting this research, Gladwell is suggesting that these strong ties are unlikely to
develop on the internet as opposed to through face to face communication with someone
already involved. However, as online social networks are merely used as modernisations of
communication techniques, it is my understanding that strong ties may still be able to
develop so long as the relationship between the members of a political group do not only
communicate through this method.
Linking to the thoughts of Vaccari (2012), Gladwell explains the weak ties that social
networks can create. Gladwell argues social networks are effective at
increasing participationby lessening the level of motivation that participation requires
(2010). Gladwell expands; Facebook activism succeeds not by motivating people to make a
real sacrifice but by motivating them to do the things that people do when they are not
motivated enough to make a real sacrifice (2010).
Haridakis and Hanson (2011) studied the 2008 American presidential electoral
campaign and explain the salience of new media. They explain that YouTube, Facebook and
Twitter were all new websites and Myspace.com was also less than one year old before the
previous election. The authors write; the rise of the media provided new conduits by which
information could flow in multiple directions between media professionals and the audience
(2011, 61). They argue the exponential increase in paths of communication has changed
the media distribution model fundamentally (ibid). The claims of the authors are backed up
by Hall (2009), who explains from a survey by the Pew research centre, it was found that 52%
of those with a social networking profile on a site such as Facebook used it for political
purposes (2009).
The authors suggest whilst traditional forms of media such as newspapers and TV
were dominant in terms of disseminating political information in 2008, the landscape was
beginning to shift (2011, 62). Haridakis and Hanson draw upon the work of Tedesco (2006),
who explains social media has a positive effect of political efficacy. As social media is
interactive, a voter online can engage with a message posted by a party, they can share it
with their friends and debate the message with other followers of the party. Haridakis and
Hanson explain young people who are more self-efficacious may be less politically apathetic
and more likely to vote (2011, 65).
8

Haridakis and Hansons research conclusions explain younger citizens used social
media significantly and also found the internet more useful in acquiring political. However,
the authors explain younger voters turned to the internet as a whole more than they did to
the individual functions of social media (2011, 77). The study concludes with the authors
explaining that traditional use of the internet (newspapers online for example), TV and
newspapers are the big 3 when it comes to political communication, but as young voters
adopt social networks as their chosen media outlet political wars of the future may play out
more and more in these mediated spaces (2011, 78).
However, Nielsen (2010) explains he believes that mundane internet tools such as
email are more important to political parties than emerging tools such as social media (2011,
4). Nielsen goes on to explains from his research, he found staffers used email to maintain
relationships with potential volunteers, explaining; people are simply much more likely to
work for a campaign if someone asks them to. (2010, 6). The author goes on to state that in
recent years there has been a resurgence in attempts to mobilise participants to help out in
electoral campaigns by organizers. To do this they have used email for personalized political
communication (ibid) to cut through the clutter of advertising and direct mail individuals
are subjected to during electoral campaigns.
Nielsen explains that tools such as web searches and email are increasingly
mundane, not developed specifically for political purposes, and equally available to staffers
and volunteers (2010, 9). The fact mundane internet tools such as email are available to
nearly everyone willing to participate in electoral campaigning means its reach is much
wider, unlike social networks where membership tends to consist of mainly young people.
Nielsen explains; mundane internet tools like email and search are integral to the everyday
work of identifying, recruiting, and retaining campaign volunteers. The ubiquity of these
technologies attests to their importance (2010, 25).
Nielsen continues; whilst Facebook was used in the 2008 congressional campaign in
the US, its effectiveness was limited. Whilst many groups were set up, they were often only
updated by staffers and rarely got comments and thus amounted to one way
communication. To highlight this, he recalls when he asked the volunteer coordinator in one

campaign how many volunteers they had recruited via social networks like Facebook, he just
laughed at my question (2010, 17).
Through mundane internet tools, Nielsen explains it is possible to meet the public
where they are (2011, 26). When parties do this they can actually realize some of the
much-vaunted mobilizing potential of the internet in practice, and get people involved.
(2010, 26)
Pippa Norris (2000) supports the idea of the mobilising potential of personalised
political communication online. The author draws upon the work of Putnam (2000) and his
famous work; Bowling Alone. Putnam explains that In the USA traditional social groups have
long been in decline. Putnam argues individuals no longer have large community networks
that contain politically mobilising influences, reducing their social capital which has
contributed towards long term erosion of American voter turnout (2000, 256). Therefore it
is up to political parties to act as mobilising forces. Norris explains that through direct
communication online, political parties can serve to increase party and candidate support
amongst electors and the propensity to turn out (2000, 257). However, Norris qualifies her
conclusion and states; political communication by political parties through new media is
more likely to have a reinforcement effect upon individuals, explaining; at present, politics
on the internet reinforces the activism of the activists, rather than magically transforming
the apathetic into engaged citizens. (2000, 266). At the time of writing (2000) however,
there were not the large amounts of social networks online as there are today. From
Nielsens (2010) and Norriss (2000) conclusions about direct communication online I believe
through positive party messages directed at voters on sites such as Facebook and Twitter
where only low threshold activity is needed, there is a very real possibility that new media
can have a mobilising, motivational and engaging effect upon individuals, particularly as the
use of these sites becomes more widespread.
In 2008, campaign site MyBarackObama.com was used effectively. This site built
upon the ideas of the Howard Dean campaign in 2004 using meetup.com to organise
supporters. The site allowed supporters to set up profiles to communicate with other
activists and donate to the campaign. Barko-Germany (2009) explains; MyBarackObama
was designed to use the internet to recreate as much as possible offline relationships by
10

allowing members to meet each other (2009, 156). Creation of offline relationships enabled
campaign activists and less motivated supporters to easily move between low threshold
activities of making a profile on the site and getting involved with public meetings and
campaign events. Barko-Germany explains this was an ingenious way of harnessing the
excitement and energy (ibid) of the campaign and utilising all the positive aspects of the
online echo chamber effect whilst also channelling energy into real activities that met the
goals of the campaign (ibid). Through MyBarackObama.com, the Obama campaign
managed to create strong ties between members of an online social network. Successes of
this campaign are shown by the large sums of money the Obama campaign received in
donations, the campaign raised a total of $650,000,000, crediting online donors for this
success.
Ancu explains how Twitter became an important communication tool for both
candidates and supporters in the 2008 US Presidential Election campaign. She explains;
Political candidates adoption of Twitter and similar social network Web sites recognizes
changing media consumption patterns of the American electorate (Ancu, 2011, 14). The
type of internet media consumption is also changing, with 2009 figures suggesting internet
users were spending more time on social networking websites than on email (ibid). This
surge in popularity seems to suggest social networking sites are moving in to the realm of
mundane internet tools rather than emerging tools in the definition of Nielsen (2010).
Ancu concludes; Twitter use by both candidates and voters in the 2008 campaign
was mainly used for disseminating information instead of engaged political discussion, but
explains that this communication mode has plenty of potential for important consequences
on political life (2011, 20). While some Twitter activity may be ill-informed ranting from
voters, it can create awareness of group activity (ibid) as it gives individuals a platform to
express personal opinions. Those that do not post Tweets can still become more engaged as
a result of being on the site due to the ability it hands politicians to disseminate campaign
information to large numbers of people and the ability of their supporters to retweet
(share) this information to their online friends. This argument extends to the idea that even
if politicians do not directly interact with followers they can still gain support online through
the use of Twitter.

11

To conclude this section, I will sum up the key debates that exist in the literature
surrounding the use of new media in electoral campaigning. Norris (1999) brought about
the idea of narrowcasting in the post modern campaign, where individuals can chose the
media they consume from a huge range of sources, and as such, political parties can focus
their mobilising attempts to sections of media where niche groups are likely to be paying
attention. Margolis and Resnick (2000) and Hindman (2009) broadly agreed that political
advertising in new media spaces such as online was broadly a modernisation of old
campaigning techniques where the old political elites would be the only ones receiving
these new media messages. However, with the advent of social media there became a real
possibility that parties could begin to communicate to new sections of the population.
Vaccari (2012) brilliantly explains the theory of how political messages are received,
explaining the paradox that those who are most likely to receive messages are those who
are less likely to be influenced. This is a key debate over whether these campaign messages
online are merely reinforcing ideas already held by politically engaged voters, or mobilising
new supporters. The debate over high and low threshold activities when identifying and
participating with a political message online is important. Most agree; simply linking your
social network profile to a political party online is a low threshold activity, Gladwell (2010)
explains that this will only create weak ties between the individual and political party, and
the political messages received are unlikely to make them act offline due to the voters
apathy. Haridakis and Hanson (2011) however explain this activity can increase political
efficacy as individuals engage with the messages of parties and as such, friends they are
connected to online will also become exposed to these messages. There is still the debate
however that strong ties between individuals and the party are needed in order to have any
mobilisation effect. Nielsen (2010) explains how mundane internet tools such as email are
more important than web 2.0 tools for parties due to their ubiquity and ability to
communicate personally with voters to get them to act in the offline world, creating strong
ties. The example of MyBarackObama.com in the 2008 presidential election however, seems
to have combined the echo chamber effect of creating hype about a campaign whilst
simultaneously creating strong ties between individuals offline. Whilst the use of Twitter by
Obama in 2008 was largely for one way communication as explained by Ancu (2009), I will
explore whether any lessons were learned from MyBarackObama.com in 2008 and whether
the usage of the site changed in attempt to create more strong ties from low threshold
12

activities. Ancus analysis of Twitter use in 2008 is interesting, but as Wortham puts it; If the
presidential campaigns of 2008 were dipping a toe into social media like Facebook and
Twitter, their 2012 versions are well into the deep end (2012, 1).

Twitter Analysis

I will analyse the use of Twitter by both Barack Obamas and Mitt Romneys
campaign teams in the week leading up to polling day for the 2012 US Presidential Election
on November 6th. For an in depth explanation of Twitter and the explanation of retweets,
see appendix A.
Retweets on Twitter are particularly important as they have the ability to draw the
attention of individuals who do not follow political candidates on the site. Vaccari (2012)
and Shirky (2011) explain how the opinion forming process is influenced by acquaintances
and how the endorsement of a friend can increase the intensity of a message. From the
example in Appendix A, a follower of @common_genius who is not engaged with the
political debate may see the message from Obama through the retweet shown. This
endorsement combined with the content of the tweet from Obama, may increase his
support for Obama. This is low threshold political activity online, but the dissemination of
messages by supporters can help to foster public support for a candidate amongst those
who are not specifically on the site to debate political matters. This can be seen to be a case
of politics attempting to meet the public where they are (Nielsen, 2010, 26). For this
reason I have used the number of retweets a tweet from either campaign achieved as the
main marker of how successful the message was.
Interactivity on Twitter can have a twofold mobilising potential. Firstly, the individual
connected to the political candidate will increase their political efficacy as they become
engaged in a political debate (Haridakis and Hanson, 2011). Secondly, their followers will
become aware the individual is communicating campaign information, and thus will become
more aware of the political messages. Whilst these are weak ties that are being created,
once the individual is convinced by the campaign message, the Twitter campaign can then
attempt to create stronger ties and demand more action from the individual. I will assess to
13

what extent the Obama and Romney campaigns used Twitter as a call to arms, using the
site to mobilise individuals to volunteer, donate money or merely just to vote. Secondly I
will assess to what extent each campaign used the interactive nature of the site to create an
echo chamber.

Methodology
My research method will be a content analysis of tweets, as described by
Krippendorff (2004) and Weber (1990). Krippendorff defines content analysis as a research
technique for making replicable and valid inferences from texts to the contexts of their use
(2004, 18). Weber (1990) lists a number of ways in which content analysis can be used, from
this, my content analysis will be used to; identify the intentions of the communicator, and
to describe trends in communication content (1990, 9).
My content analysis will be done through classification of tweets, coded by myself
and then checked by another coder to ensure reliability. My coding scheme will be to
analyse one tweet at a time. I will create classifications through a top down approach where
I will pre determine the type of tweets I am likely to encounter and then create categories
for these tweets. I will then move on to a bottom up approach where I will create new
categories for tweets that contain new types content not previously anticipated during data
collection. My categories are not mutually exclusive as a single tweet can contain many
defining features, for example the use of an embedded photo and a call for an offline action.
I will use thematic networks as explained by Attride-Stirling (2001) in order to group the
large numbers of basic themes into organising themes. For example, the basic themes of
tweets containing embedded pictures and embedded videos can be grouped into the
organising theme of containing visual media content.
My categorizations will be tested on a sample of tweets to ensure there are no
ambiguities in my coding scheme, also, through inter coder reliability testing I will ensure
not only that my data is coded correctly but also that my coding scheme is reliable.

14

Data Collection

To access the Tweets from the Barack Obama campaign, I searched Twitter using the
term Find:BarackObama. This enables the user to view all tweets posted from the account
@BarackObama. I then scrolled through posts to October 31st 2012, one week before
polling day on November 6th. I did the same, instead using the search term
Find:MittRomney to find tweets posted from the official campaign page of Mitt Romney.
I then created a spread sheet using Microsoft excel to code my data, adding
categorisations that can be seen in Figure 1.
At this point I began to collect my data, coding the results of the content analysis by
entering a 1 if the tweet did not contain the type of content explained in the category
column above and a 2 if the tweet did contain the type of content.
Tweet #1 read:
@BarackObama: Foreign policy heavyweights break down the stark contrast between the
President & Gov. Romney on Iran: http://OFA.BO/DKdUXm #RomneyNotReady
This tweet had 2s placed in the category columns for Embedded Video, Hashtag
Used and Policy Detail. As a note, whilst the tweet contains a link, Twitter automatically
embeds videos into the tweet, thus this link is not counted as an external link. The Tweet
also does not directly ask the user to watch the video, whilst this may be implied there is not
a direct call for online action so there was a 1 placed in the Calls for Online Action column.
The tweet clearly states the video contains details of Obamas policy on Iran so a 2 was
placed in the Policy Detail category. Also, the hash tag #RomneyNotReady was used in
this tweet. Hash tags are used commonly on Twitter and allow users to categorise their
tweets and when clicked, they allow the user to view all other tweets with the same hash
tag. Thus I included the Hash Tag Used category through a bottom up categorisation
technique and entered a 2 in the column.

15

Next to the Tweet number and date/time column, I added the number of retweets
the tweet achieved, which in this case was 494.

Reliability Testing

After analysing 30 tweets I felt I had an adequate amount of categories and began to
test my coding scheme on a further sample of 10 tweets, as well as giving another coder a
sample of 15 to assess my accuracy.
From content analysis of 15 tweets I made 34 categorisations, my coding partner
matched my results on 28 out of these 34 categorisations, an accuracy rating of 82%. My
coding partner suggested a new category for direct links to other users on Twitter. This
bottom up categorisation enabled me to create a new category of tweet in my coding
scheme.
The inter coder reliability test highlighted the ambiguity of the calls for online action
categorisation. My coding partner suggested that a tweet that asks the user to retweet
could also be categorised as a call for online action. From this debate I have edited my
coding scheme to only categorise a call for online action as actions that are asked for on
websites outside of Twitter. The other categories in the coding scheme such as asks for
retweet assess fully how the campaign teams attempt to get their followers to create an
echo chamber on Twitter. Once all these basic themes are detected and recorded in the
data, tweets that call followers into action both on Twitter and on other websites can be
grouped together into the organising theme of online action. I feel from this inter coder
reliability testing that my coding scheme has been tested well and is more accurate as a
result.

16

Figure 1. Screenshot of data coding spread sheet in Microsoft Excel with categorisations and
coding added.

Research Questions

1. Barack Obamas campaign will attempt to mobilise voters and create offline action
through his tweets more frequently than Mitt Romneys campaign.
2. Obamas campaign will use interactive tweets more frequently than the Romney
campaign; conversely, the Romney campaign will use non interactive tweets more
frequently than the Obama campaign.
In order to determine the frequency of tweets attempting to mobilise followers into
offline action, I shall create a frequency table to determine how often each campaign
attempted to foster offline activities from voters. From these frequency tables I can read
which campaign was most committed to the creation of strong ties through its Twitter
campaign. I will also be able to sort my tweets so all tweets calling for offline action are
placed at the top of my spread sheet, from this I can view how successful in terms of
retweets these tweets were. My secondary hypothesis is that these tweets will gain more
retweets than the mean across all tweets.
By viewing the coded data I can analyse how often each candidate used tweets that
amounted to merely one way communication and conversely, how often they attempted
to get a response from the public. By organising my data into thematic networks by creating

17

categories for the organising themes of both interactive and non-interactive tweets I can
create frequency tables to read my data.

Findings

An initial finding from my data was the extent to which the Barack Obama campaign
was used Twitter compared to the Romney campaign. In the same one week period, the
Obama campaign twitter account posted 325 tweets, whereas the Romney campaign
account posted 37.
Tweets posted from the Mitt Romney Twitter account received a mean of 2953
retweets and on the Obama Twitter account a mean of 2952 retweets was recorded. When
assessing total support, @BarackObama received 959,295 retweets in total, whereas
@MittRomney received 109,250.
The Barack Obama campaign posted 325 tweets between 12.33am on the 31st
October 2012 and 10.01pm on the 6th November 2012. This time period accounts for 165.5
hours (rounded up by 2 minutes). However, the latest tweet posted from either the Mitt
Romney or Barack Obama account was at 1.07am (@BarackObama on 6th November 2012)
and the earliest was posted at 6.07am (@MittRomney on 6th November 2012). Removing
this 5 hour time period where both Twitter accounts were inactive on all days of the
investigation means we are left with an active time period of 135.5 hours. Thus, a tweet was
posted from the Obama account every 0.42 waking hours, converting to a tweet being
posted every 25 minutes.
The Romney account posted 37 tweets in this time period with the first posted on
the 31st October 2012 at 10.51am, and the final tweet posted at 5.55pm on November 6 th
2012. Using the same 135.5 hour time scale, the Mitt Romney campaign posted a tweet
every 3.6 waking hours, converting to a tweet being posted every 3 hours and 41 minutes.

18

Mitt Romney

Barack Obama

Total number of tweets

37

325

Frequency of tweets

every 3 hours and 41

every 25

(average)

minutes

minutes

Mean number of retweets

2953

2952

Total number of retweets

109,250

959,295

Table 1. Statistics of total number and frequency of tweets and how many retweets
were achieved.
My first research question assesses to what extent each campaign used Twitter to
foster offline action. This excludes tweets asking the reader to go out and vote, these tweets
ask the reader to participate in action such as calling friends persuade them to vote, donate
money to the campaign or to go out and canvass support. The Obama campaign posted 37
tweets which asked the reader to engage in an activity offline. This amounted to 11.38% of
tweets, meaning for every 9 tweets posted (rounded up from 8.783), one would ask the
reader to engage in offline. This relates to a tweet calling for offline action being posted on
average every 3 hours and 47 minutes.
In the Romney campaign, 6 tweets were posted asking the reader to engage in
offline action. This amounted to 16.21% of tweets posted, meaning for every 6 tweets
posted (rounded down from 6.169), one would ask the reader to engage in offline action.
Despite the Romney campaign posting these types of tweets more frequently as a
percentage of total tweets, a tweet calling for online action was posted on average every 22
hours and 12 minutes.
Also, the Romney campaign achieved a total of 11,127 retweets on all tweets asking
for offline action, whereas the Obama campaign achieved 112,968 total retweets.
My secondary hypothesis for this research question was that these tweets calling for
offline actions from supporters would gain more retweets than the average tweet. The
Obama campaign achieved a mean of 3053 retweets with these tweets compared to the
total mean number of retweets of 2952. In the Romney campaign, tweets calling for offline
action received a mean of 2117 retweets, less than the total mean number of retweets of
2953.
19

Tweets calling for Offline


Action

@MittRomney

@BarackObama

Total number of tweets

37

Percentage of total tweets

16.21%

11.38%

every 22 hours and 12

every 3 hours and 47

Frequency of tweets

minutes

minutes

Mean number of retweets

2117

3053

Total number of retweets

11,127

112,968

Percentage of total retweets

10.18%

11.78%

Table 2. Data for tweets posted that called for offline action from the reader.
My second research question assessed to what extent each campaign aimed to
engage the individual by promoting low threshold online activity and actively attempting to
create an echo chamber for messages. Interactive tweets attempt to engage the user by
asking them to retweet posts, submit replies, follow links, watch videos or sign up to mailing
lists. By using the method of grouping categorisations into organising themes as explained
by Attride-Stirling (2001), I grouped the categorisations of; asks for response, asks for
retweet and calls for online action into the organising theme of Interactive Tweets.
The Obama campaign posted 42 tweets grouped into the organising theme of
interactive tweets, this amounted to 13% (rounded up from 12.932%) of all tweets
meaning for every 8 (rounded up from 7.738) tweets posted, one asked the reader to
engage in a low threshold online activity. An interactive tweet was posted on average every
3 hours and 6 minutes by the Obama campaign. As a side point, these tweets received a
mean number of retweets of 9477, significantly higher than the mean of all Obamas
campaign tweets which was 2952.
The Romney campaign posted 8 tweets grouped into the organising theme of
interactive tweets. Interestingly, all of these tweets asked the reader to complete an action
online such as finding out where their local polling station was or watching a video rather
than asking them to retweet or respond to a question. These 8 interactive tweets accounted
for 21.62% of all tweets, meaning for every 5 (rounded up for 4.625) tweets, one would be
20

interactive. This means an interactive tweet was posted by the Mitt Romney campaign
every 16 hours and 39 minutes. Again, as a side note; the mean number of retweets for
these interactive tweets was 2952, slightly below the mean number of retweets for all posts
from the @MittRomney account which was 2953.
Online Interactive Tweets

@MittRomney

@BarackObama

Total number of tweets

42

Percentage of total

21.62%

13%

every 16 hours and 39

every 3 hours and 6

Frequency of tweets

minutes

minutes

Mean number of retweets

2952

9477

Total number of retweets

23,612

398,028

Percentage of total retweets

21.61%

41.49%

Table 3. Data for tweets posted that called for online action from the reader.
My secondary hypothesis was that the Romney campaign would use non interactive
and one way tweets more frequently than the Obama campaign. To correctly assess the
number of non-interactive tweets, I took the same organising theme of interactive tweets
and added the categories of calls for offline action, calls for funds and specifically asks for
vote to it. These tweets still recognise the audience. Once this larger organising theme of
instructional tweets had been created, I could view all other tweets as non-interactive,
non-instructional and as one way communication.
The Obama campaign posted 245 tweets which can be seen as one way
communication. This accounted for 75.38% of all tweets, meaning 3 out of every 4 tweets
were merely one way communication. On average, every 33 minutes a tweet that was
neither interactive, nor instructional and did not communicate with the reader was posted.
147 of these 245 non interactive tweets were quotes from Obamas speeches, making up
60% of the category. The mean number of retweets for these posts was 1772 exactly, less
than the overall mean.

21

The Romney campaign posted 17 tweets which can be seen as one way
communication. This accounts for 46% (rounded up from 45.94) of all tweets. More than
half of the tweets posted by the Romney campaign attempted to engage the reader in some
way. On average, a non-interactive tweet was posted every 7 hours and 50 minutes. The
mean number of retweets for these posts was 2226 (rounded up from 2225.70), less than
the overall mean.
One Way Tweets

@MittRomney

@BarackObama

Total number of tweets

17

245

Percentage of total

45.94%

75.38%

every 7 hours and 50


Frequency of tweets

minutes

every 33 minutes

Mean number of retweets

2226

1772

Total number of retweets

37,837

434,029

Percentage of total retweets

34.63%

45.24%

Table 4. Data for tweets posted that amounted to one way communication with the reader.
I will now discuss these findings and work towards concluding whether new media
has provided effective ways to engage and motivate citizens.

Discussion

My rationale for carrying out this content analysis of election campaign tweets was
to assess for what purpose new media was used by politicians. Margolis and Resnick (2000)
stated; the existence of party websites served merely as demonstrations of modernity
(2000, 66). My data can be seen to further this argument. Over 75% of tweets from Barack
Obamas Twitter account were classified as one way communication. A majority of these
tweets were quotes from party rallies. This is a clear example of a modernisation of the
dissemination of traditional political communication. Previously, a citizen would have had to
attend the rally to hear the speech, or wait for it to be broadcast. Through Twitter, they can
read the speech as it is being given. Whilst this new media creates opportunities for a larger

22

audience to receive a message, the use of web 2.0 tools in this way ignores the social
network and interactive aspect of the site.
Bringing in the comparative nature of my study, the Mitt Romney campaign posted
far fewer one way tweets than the Obama campaign. 46% of tweets from the Mitt Romney
campaign amounted to one way communication. This can be attributed partly to the
significantly lower total number of tweets posted by the Mitt Romney campaign however.
Ancu (2011, 13) noted how in over 5 weeks of the presidential election campaign in
2008, the Obama campaign posted 261 tweets and the McCain campaign posted 26. In one
week in the 2012 presidential election campaign both Obama and the new republican
candidate, Mitt Romney, posted more tweets than this respectively. This proves new media
use is being taken increasingly seriously by candidates. Ancu also noted Obamas Twitter
page was mainly used as a one way information push (2011, 16) but still managed to gain a
large audience on the site in 2008 and briefly became the most popular Twitter account on
the site. Obamas account is still one of the most popular on the site and can still be seen in
part as a one way information push. I believe whilst many posts on the Barack Obama
Twitter account amount to one way communication, it is the sheer amount of posts that
make the site popular and this furthers the popularity of the president. The constant
updating of the site act as a kind of transparency as the reader can always see a recent post
relating to something Obama is doing, or a policy he is promoting, the use of pictures and
video create a more personal relationship with the president and make him seem less
anonymous.
Vaccari (2012) explains the paradox of the most politically aware voters being the
only voters receiving messages online and the least aware voters remaining ignorant can be
overcome by Twitter. The Obama Twitter account has amassed a huge amount of followers,
29,279,859 as of the 5th April 2013, and posted a phenomenal number of tweets during the
electoral campaign. It is assumed most of these followers are at least slightly politically
aware due to their choice to follow the president. However, each tweet posted from the
account received a mean number of retweets of 2952, a total of 959,295 in just one week of
the electoral campaign. Both Shirky (2011) and Vaccari (2012) explain how messages shared
by friends can increase the intensity of a message and increase the possibility that opinions
23

can be formed and changed as a result of new information provided by friends. As so many
tweets from Obama were shared, the chance that these messages were received and
accepted by individuals with fairly low political awareness is great. Whilst the Romney
campaign received the marginally higher mean rate of retweets of 2593, due to the
relatively small number of tweets in total, the overall figure for the number of retweets was
109,250. Whilst the same logic applies to the Romney campaign, as of the 5th April 2013 the
account only has 1,598,858 followers, this makes it likely that a smaller, core group of
supporters were responsible for these retweets, meaning a smaller overall network of
Twitter users would come in to contact with these messages.
Gladwell (2010) explains pessimistically that new media can increase participation in
political activity by reducing the amount of motivation needed to participate. This relates to
the idea of low threshold political activity. My research aimed to assess the extent to which
this was true in the 2012 election. The Barack Obama campaign posted 42 tweets (13% of
the total) which specifically asked users to complete a task online which supported the
campaign, for example at 7am on November 6th 2012 @BarackObama posted:
Its Election Day! This is your last chance to help win this thingRT this link so your friends
know where to vote: http://OFA.BO/chM7LH
This tweet received over 10,000 retweets, significantly more than the mean for all
tweets, showing the followers were happy to follow instructions to complete a simple task
to support the campaign. The same result was clear across all tweets which asked the
reader to complete an online task. The mean number of retweets for this category of tweet
was 9477, significantly higher than the total mean number of retweets which stood at 2952.
These 42 tweets accounted for 13% of all the tweets posted in the research period yet
received 41.5% of the total retweets for period (398,028 out of 959,295). Interestingly,
tweets asking the reader to complete an offline action such as calling friends in key states to
ask them to vote, or tweets asking for donations did significantly worse in terms of gaining
retweets in the Obama campaign. 37 tweets (11.38% of the total) were posted asking the
reader to complete an offline action; these tweets achieved a mean number of retweets of
3053, only marginally above the mean rate of response. This significant drop in willingness
to share tweets which call for action requiring more motivation and commitment to the
24

cause seems to show that the Twitter followers of Obama are much more comfortable with
low threshold support of the campaign.
The same relationship can be seen in the mean number of retweets of posts calling
for online action and those calling for offline action from the Romney campaign but to a
lesser degree. The Romney campaign received a mean number of retweets of 2952 for
online action tweets but 2117 for offline action tweets.
These figures seem to prove that supporters of the campaign online are much more
willing to support the campaign when it involves doing only a small amount of work.
Margolis and Resnick (2000) explained from interviews with webmasters of campaign
websites in the late 1990s, the general consensus was that websites and online activity was
useful for communication but was no substitute for money and organisation on the ground
(2000, 66). They explain; websites must do much in the way of motivating individuals online
in order to make them overcome their habitual indifferences (2000, 72). These figures
seem to prove the conclusions of Gladwell (2010). Twitter can be seen to be excellent in the
communication of messages that inform the individual about what is happening in the
campaign, and disseminating the contents of speeches. Whilst there is a small attempt to
motivate citizens to volunteer to help the campaign and to donate money online, the
audience is more reluctant to support this. However, as Haridakis and Hanson (2011),
Tedesco (2006), Vaccari (2012) and Shirky (2011) explain, whilst the response to overt calls
from the campaign to volunteer for the campaign are not particularly well received, through
the reception of messages either directly from the campaign or from friends, greater selfefficacy can be evoked leading to higher levels of political motivation offline.
These figures show that activists seem to be active away from Twitter. Whilst they
will be in the core number of people who will share almost all messages posted from the
account of their preferred candidates, it is most likely they will mainly operate in other
places online, through mundane internet tools and most importantly, offline. In the case of
the Barack Obama, the specific social network mybarackobama (MyBO) was set up to
harness the power of sites like Facebook and Twitter to create an echo chamber for
messages but also move beyond this and create strong ties between members in the offline
world (Barko-Germany, 2012, 156). The Obama campaign posted 71 tweets with links to
25

other sites, the majority of these were links to MyBO, this related to 22% of all tweets in the
week leading up to Election Day. On this site, users are given detailed policy information,
added to mailing lists, asked for donations and are prompted to add their zip code so their
location data could be accessed and could be added to campaign events local to them. A
particularly interesting further study could be to analyse the features of this site and how
responses to calls to offline action varied to responses on other social media such as Twitter
and Facebook.
Interestingly, the use of mundane internet tools is strong once a user signs up to the
MyBO site. As a part of my research I added my email address to the site and since signing
up on the 18th March 2013, I have received 4 emails addressed to me personally asking for
donations to the campaign. More research is needed in this area and it could be a very
enlightening future study. Nielsen (2010) explains that mundane internet tools such as email
are the most useful tools for online mobilization of support. From my research into the
Obama campaign, Twitter can be seen as being used as an excellent communicating device
to update the casual supporters of the campaign, whilst partly actively attempting to create
an echo chamber for messages and enlighten new supporters, but also as a recruitment
drive to get casual supporters to connect with the campaign on another site where the real
campaign organisation begins.
A fantastic example I feel demonstrates the relationship between the use of Twitter
in the campaign and the offline campaign comes from a speech Barack Obama made in Iowa,
telling the story of an activist he met in the primaries before the 2008 election. Obama
explained how Edith Childs had a chant she would use to motivate other activists when out
campaigning. This chant became famous and Barack Obama had invited her to the rally in
Iowa to lead the chant and to end the campaign. Obama explained; She said, Id love to see
you, but I think we can still win North CarolinaI dont have time just to be talking about it.
Ive got to knock on some doors. Ive got to turn out the vote. Im still fired up, but Ive got
work to do (Obama, 2012). Twitter can be seen as the part of the campaign where action is
just talked about, there is the possibility of motivation of citizens through exposure to
messages, but the real campaign work is still done offline, through interpersonal
communication, and this is where votes are won.

26

However, whilst we can use this data to argue the point that activism is not fostered
on Twitter, nor do activists use the site as a key way of gaining votes, we are still left
wondering about what effect the largely one way communicative nature of the campaign
Twitter accounts has on the public. There is strong theoretical backing to suggest the active
creation of an echo chamber can increase knowledge and thus increase political
participation. I have also hypothesised that the frequent use of Twitter by the Obama
campaign creates a more personal relationship with his followers and supporters and can
greatly increase his popularity. To fully explore this however, it would be useful to assess a
selection of responses to tweets posted by citizens and to assess whether these sparked
political discussion. Also, interviews with followers of either Twitter account would be
particularly enlightening as conclusions could be drawn as to how big a part the site played
in their political motivation. The drawback of merely using retweets as a marker of how
successful a piece of political communication was is that we do not know why the user has
chosen to do this. A more qualitative approach, attempting to find out the motivations
behind political twitter use from individuals and what effect the site has on an individuals
political efficacy is the next logical step in this field of research.

Conclusion

To conclude this dissertation, I believe there is clear evidence that for the most part
in the 2012 US presidential election; Twitter was not actually used to actively create an echo
chamber or to actively attempt to get the reader to engage with the campaign offline. The
majority of tweets from the Obama account were simple campaign messages and quotes
from speeches, both of which can be seen as merely demonstrations of modernity. Whilst
the Romney campaign posted a majority of tweets attempting to engage with the reader,
the limited number of tweets overall and the significantly smaller reach of the Twitter
campaign compared to Obama means I do not believe this had a significant effect. Whilst
the Obama account posted many tweets, there seemed to be a reliance on the users of the
website creating an echo chamber of messages without being prompted. From the Obama
campaign, tweets merely asking the user to retweet a link, watch a video or reply to a
question were significantly more likely to be shared by readers of the site. Calls for low
27

threshold action were much better received by the audience than messages calling for
offline action. However, the limitations of my research were that this conclusion is only
based on the number of retweets these messages achieved.
From this data I believe the ideas of Norris who explains; politics on the internet
reinforces the activism of the activists (2000, 266) can be challenged. Twitter use in the
2012 US presidential campaign seemed to reinforce the lethargic support of the majority of
fairly indifferent individuals who are slightly politically engaged. There are attempts by both
campaigns to foster offline support with the Obama campaign encouraging the audience to
sign up to the MyBarackObama website where stronger ties are developed. However, as
Vaccari (2012), Shirky (2011), Haridakis and Hanson (2010) explain, Twitter use as a means
of one way communication to a large group of followers can still convince people to vote a
certain way and aid learning about policy debates, this is no bad thing, but the grassroots
activism is occurring away from the site and in more specialised places online and more
importantly, offline. Twitter is used as an excellent communicative device and has the
possibility to politically engage and motivate citizens to an extent, and its use by politicians
in a campaign can have a positive effect. However, the key to a successful campaign is not
the use of Twitter. Obama has used new media effectively to engage his supporters through
a strong presence online; this informs individuals and creates a more personal relationship
with the campaign. The Romney campaign did not do this to anywhere near the same
extent, yet still gained a large share of the vote at the 2012 election (47.2%). New media use
can entertain, engage and motivate citizens to an extent, but as Margolis and Resnick
explain, new media use is no substitute for money and organisation on the ground (2000,
66). My conclusion states that whilst Twitter is an important factor in the awareness of a
campaign, other efforts are needed to foster motivation to get out the vote.

28

Appendices
Appendix A
Twitter is an online microblogging service and also a social network, it is free to use
and anyone can create an account. The user can then post tweets, of 140 characters which
become public. Users on the site can follow other users, which means they subscribe to
another accounts tweets and these will be posted on the users homepage. Users can follow
whoever they wish to personalise their homepage, meaning different users will see a unique
homepage with only messages from users they have decided to subscribe to. Users can post
public messages to each other by using the @ sign and then typing the username. Users
can also interact with other users by using retweets, this means a user forwards a message
posted from another account. One example of this from my research is shown below.

Figure 2. List of users who have retweeted a message

This tweet was posted from Barack Obamas account. Once you click on a tweet, details are
given of the number of retweets and the users that retweeted the post. If we click on one of
the names we can see how the message is reposted by another user.

29

Figure 3. Retweeted Message


This user, @common_genius, has retweeted the original message posted from
Barack Obamas account and thus it has appeared on his personal timeline. This message is
then forwarded to his 273 followers and will appear on their homepages, despite the fact
that they may not follow Barack Obama.

Bibliography

Ancu, M. (2011). From Soundbite to Textbite: Election 2008 Comments on Twitter. In:
Hendricks, J. A. and Kaid, L. L. Techno Politics in Presidential Campaigning: New Voices, New
Technologies and New Voters. London: Routlegde. pp. 11-21.
Attride-Stirling, J. (2001). Thematic networks: an analytic tool for qualitative
research. Qualitative Research. 1 (3), pp. 385-405.
Barko-Germany, J. (2009). The online revolution. In: Johnson, D. W. Campaigning for
president 2008 strategy and tactics, new voices and new techniques. Hoboken. Taylor &
Francis.
Civic Youth. (2012). Youth Voting. Available: http://www.civicyouth.org/quick-facts/youthvoting/. Last accessed 19th April 2013.
30

Gladwell, M. (2010). Small Change: Why the revolution will not be tweeted.. Available:
http://www.newyorker.com/reporting/2010/10/04/101004fa_fact_gladwell. Last accessed
19th April 2013.
Gronbeck, B. E. and Wiese, D. R. (2005). The Repersonalization of Presidential Campaigning
in 2004. American Behavioral Scientist. 49(4), pp. 520534.
Hall, M. (2009). Internet engaged people in '08 election, survey shows. Available:
http://usatoday30.usatoday.com/tech/webguide/internetlife/2009-04-15-online_N.htm.
Last accessed 19th April 2013.
Haridakis, P. and Hanson, G. (2011). Campaign 2008: Comparing YouTube, Social Networking,
and Other Media Use Among Younger and Older Voters. In: Hendricks, J. A. and Kaid, L. L.
Techno Politics in Presidential Campaigning: New Voices, New Technologies and New Voters.
London: Routlegde. pp. 61-82.
Hindman, M. S (2009). The Myth of Digital Democracy. Princeton: Princeton University Press.
Katz, E. and Lazarsfeld, P. F. (1955). "Personal Influence". New York: Free Press
Krippendorff, K. (2004). Content Analysis: An Introduction to its Methodology. Thousand
Oaks, California: Sage.
Leip, D. (2012). 2012 Presidential General Election Results. Available:
http://uselectionatlas.org/RESULTS/index.html. Last accessed 19th April 2013.
Margolis, M. and Resnick, D. (2000). Politics as Usual: The Cyberspace "Revolution". London:
Sage.
McAdam, D. and Paulsen, R. (1993). Specifying the Relationship Between Social Ties and
Activism. The American Journal of Sociology. 99 (3), pp. 640-667.
Nielsen, R. K. (2010). Mundane internet tools, mobilizing practices, and the coproduction of
citizenship in political campaigns. New Media and Society. 13 (5), pp. 1-27.
Norris, P (1999). On Message: Communicating the Campaign. London: Sage.

31

Norris, P (2000). A Virtuous Circle: Political Communications in Postindustrial Societies.


Cambridge: Cambridge University Press.
Obama, B. (2012). Fired Up! Ready To Go! Speech. Available:
http://www.politicalspeeches.net/barack-obama/barack-obamas-fired-up-ready-to-gospeech. Last accessed 19th April 2013.
Putnam, R. D (2000). Bowling Alone: The Collapse and Revival of American Community.
London: Simon & Schuster.
Shirky, C. (2011). The Political Power of Social Media Technology, the Public Sphere, and
Political Change. Available: http://www.foreignaffairs.com/articles/67038/clay-shirky/thepolitical-power-of-social-media. Last accessed 19th April 2013.
Tedesco, J. C. (2006). Web Interactivity and the Young Adult Political Efficacy. In: Williams,
W. P. and Tedesco, J. C. The Internet Election: Perspectives on the Web Campaign in 2004.
Lanham, MD. Rowan and Littlefield. Pp. 187-202.
Vaccari, C. (2012). From echo chamber to persuasive device? Rethinking the role of the
Internet in Campaigns. New Media and Society. 0 (0), pp. 1-19.
Weber, R. P. (1990). Basic Content Analysis. 2nd ed. California: Sage.
Wolf, G. (2004). How the Internet Invented Howard Dean. Available:
http://www.wired.com/wired/archive/12.01/dean.html. Last accessed 19th April 2013.
Wortham, J. (2012). Campaigns Use Social Media to Lure Younger Voters. Available:
http://www.nytimes.com/2012/10/08/technology/campaigns-use-social-media-to-lureyounger-voters.html?_r=2&. Last accessed 19th April 2013.
Zaller, J. R. (1992). The Nature and Origins of Mass Opinion. Cambridge: Cambridge
University Press.

32

Vous aimerez peut-être aussi