Framing Trump: How do The Trump Administration, The Guardian and Greenpeace USA frame issues on Twitter during the first 100 days of the Trump Presidency?

(1)

• Framing Trump: How do The Trump Administration, The Guardian and

Greenpeace USA frame issues on Twitter during the first 100 days of the

Trump Presidency?

(2)

2 Abstract:

This paper analyses four Twitter users - @RealDonaldTrump, @POTUS, @Greenpeaceusa, and @GuardianUS - during the first 100 days of Donald Trump’s Presidency, and highlights the methods used by each user to frame contemporary political and social issues to their followers. It is found that each user frames issues differently to other users in the study, with contrasts most clearly observed between @Greenpeaceusa and @GuardianUS, on one side, and @RealDonaldTrump and @POTUS, on the other. These findings are supplemented with humanities scholars such as Marres, Scheufele, Tewksbury, and more. The paper relies on the Digital Methods Initiative Twitter Capture and Analysis Tool (DMI-TCAT) for data accumulation and subsequent research investigations.

Key Words:

Donald Trump, POTUS, The Guardian, Greenpeace USA, Framing, Agenda Setting, Twitter, US Politics.

(3)

3

Section 1: Introduction

1.1 Introduction and Research Question

In his recent book, Insane Clown President, authored during the most recent US Primary and Presidential elections, journalist Matt Taibbi offers a fascinating, behind-the-scenes insight into one of the most polarising democratic processes in recent memory. Donald J. Trump, once of reality television fame, has become The Leader of the Free WorldTM_{, and is in the process of leaving his}

indelible Trumpian stamp on the role. Taibbi, whose views are at times almost aggressively liberal, combines furious self-searching and disappointment with a clear narrative of events, in the process providing many starting points for a thesis such as this. For instance, his claim that “Trump will one day be in the Twitter Hall of Fame. His fortune-cookie mind … is perfectly engineered for the medium” (28), although disparaging in the extreme, has at its centre a degree of truth. Although the implications are not fully clear yet, social media has been widely attributed great importance in the 2016 US elections, with many voters utilizing platforms such as Twitter and Facebook, both as sounding-boards for their own opinions, and for unverified - in the traditional sense - sources of news and election coverage (The Verge). New Media academics have recognised the growing importance of social media to current events, and have begun to research in detail the ways in which social media have changed the landscape for national debates (Bruns and Burgess, 2012).

Therefore, the following research question is posed:

- How do The Trump Administration, The Guardian, and Greenpeace USA frame issues on Twitter during the first 100 days of the Trump Presidency?

1.2 Clarification of Research Question and Research Limits

Frames, framing theory and the discussion of wider agendas, all of which will be defined in the opening sections of this paper, are the key focus of this study. Twitter offers many avenues of exploration to researchers; however, this paper is concerned with the frames themselves, and trying to identify the key framing techniques used by the users studied over President Trump’s opening 100 days in office. By extension, it is not concerned with other potential explorations made possible by the Twitter platform. Therefore, a successful paper will; identify the key characteristics of each user’s tweets; discuss how these characteristics compare and contrast to other users in the study; and discuss the impact of such efforts with relation to the ideas of academics in the wider field of Twitter studies. This will culminate with some descriptions of opportunities for future research.

(6)

6

1.3 Definition of Platform

The Research Question detailed above requires some framing of its own, as well as some explanation of its origins. The mention of Twitter in the research question implies a platform-specific analysis is to be conducted. Therefore, a solid working definition of the term ‘platform’ is needed. Here, Tarleton Gillespie is useful. In his text, Politics of Platforms, Gillespie outlines a four-pointed understanding of the concept, and thus highlights the intricacies and connotations of such a loaded term. Gillespie’s four ‘platform’ definitions - Computational, Architectural, Political, and Figurative - point readers to the complications that abound when using the term (349-351). Gillespie makes the point that “In any of ‘platform’s’ senses, being raised, level and accessible are ideological features as much as physical ones” (351). This is an important argument; Platforms such as Twitter are recognized as areas which are theoretically extremely accessible, for those with an internet connection, and thus seem like fruitful places to research political engagement (Fridkin & Kenney, 572). Although it should be kept in mind that other scholars have pointed out the limitations of Platform Studies, in the sense that it is necessary to recognize that platforms’ importance can be limited by their specificities (Rieder;

Reaction Chamber; 1), Gillespie’s arguments point to the usefulness of studying platform-specific

content.

1.4 Intro to Twitter as a platform of study

Many academics have argued that Twitter is an important research area, and that Tweets, hashtag networks and linked content are all useful avenues of exploration that add value to the platform as an area of study. Richard Rogers, for example, states that “Twitter has become an emergency communication channel in times of disaster”, a statement which may be hyperbolic if applied literally (the 2016 US election results cannot be compared to the recent Japanese earthquake, for example), but nonetheless “a state of emergency” seems an apt description for those in America who are made more vulnerable by legislation changes enacted by the Trump administration, for example; proposals to repeal the Obamacare health system have worried a large portion of the lower economic classes (Salon). Bernhard Rieder agrees that Twitter is a particularly useful area of research, arguing that the platform “allows for communication and coordination in significant social movements. For this reason, but also due to its relative openness in terms of data collection, Twitter has quickly become a favoured research object for scholars from various fields” (Rieder Reaction

(7)

7

Chamber, 2). Rieder, this time with Erik Borra, goes further on the matter, stating that “The

proliferation of actors involved in the analysis of online data … has led to an epistemological battlefield where different paradigms, methods, styles, and objectives struggle for interpretive agency” (Borra & Rieder, 264). This battlefield metaphor seems suitable when describing the current debates raging in the USA, both in traditional print media and within the “walled gardens” - limited access to information due to ‘friend’ or follower settings (Rogers) - of more recently established social networks (Digital Commerce). Borra & Rieder also warn us as to the problems associated with creating reliable tweet datasets and establishing a sound methodology, issues that will be returned to later.

Bruns and Burgess, when discussing the popularity of social media platforms, state that a “key driver here is the ease with which additional materials (links, photos, video, audio) can also be shared” (Bruns & Burgess, 1). Twitter quite obviously exhibits this feature; the platform easily incorporates media link sharing and thus is an excellent propagator of multimedia virality. However, platform specificities and strict API rules mean that platforms cannot be treated as open books of infinite data, as pointed out most usefully by Rogers (3).

Having established that Twitter can be an important area of research, despite the platform’s limitations and “walled gardens” - limited access to information due to ‘friend’ or follower settings (Rogers) - it is then important to discuss how the platform is utilized for political debates. Bruns and Burgess, having stated that Twitter is used for “ongoing discussion – and instant evaluation – of newsworthy events”, offer insight when they argue “One important aspect of news discussion practices on Twitter is the curating of information related to specific stories” (2). This implies that, like a museum or exhibition, news and information sent and received online has an element of selection, and therefore agency, involved. Such activities lead to conversations being created, as opposed to objectively coming into existence, and emphasize the need for caution when studying Tweets and Twitter users. This brings us to discussions surrounding Framing and Agenda Setting, with emphasis on how these activities are carried out online.

(8)

8

Section 2: Academic Literature Review

2.1 Definition of Framing and Agenda-Setting

Framing and Agenda Setting are contentious areas of study, with many academics offering differing definitions of the terms. Scheufele and Tewksbury usefully summarize the main points of contention in these areas, suggesting that theories of Framing and Agenda Setting first became prevalent during the Mass Media age in mid-20th_{century America and, once established, became}

hugely important terms within the context of political discourse (10-13). Although they summarize, through others, a variety of interpretations of the terms Framing and Agenda Setting, the crucial aspects of each term are clear: “Agenda setting refers to the idea that there is a strong correlation between the emphases that mass media place on certain issues … and the importance attributed to these issues by mass audiences. Framing … is based on the assumption that how an issue is presented in news reports can have an influence on how it is understood by audiences. Framing is often traced back to roots in both psychology and sociology” (11).

Having said this, many academics offer competing definitions of both ‘agenda-setting’ and ‘framing’, definitions that often disagree with one another explicitly. The fields of framing and agenda-setting research are thus highly fractured, with many slight variations of term definitions offered throughout the decades, to the point that the wider view is somewhat obscured by academics’ determination to disagree. Cacciatore makes a useful point when he states that sub-definitions of these terms are crucial, for the purposes of clarity: “Unless these studies are able to conceptually and operationally disentangle (different variations of Framing definitions) … most of the effects they identify are likely confounded and tap different effects models at the same time without being able to disentangle their unique contribution to the criterion available.” (Cacciatore, 14) One thing that does seem clear to academics, however, is that New Media platforms, especially social media platforms such as Twitter, offer many challenges to interpretative definitions of ‘framing’ and ‘agenda-setting’ previously offered by academics, and indeed these platforms may require total redefinition as a result (Cacciatore, 19).

Skogerbø et al. make a useful intervention at this juncture, bridging as they do the gap between mass-media and online framing and agenda-setting attempts. Firstly, they echo claims by Cacciatore that social media platforms require a re-analysis of established theories: “Over the past decades, the continuous tug of war between journalists and their sources for the power to define and

(9)

9 frame news has been described in terms of institutional practices such as Media Logics (Altheide 2013) and Mediatization (Hjarvard 2013, Mazzoleni and Schulz 1999). However, the relationship is currently being renegotiated because of the entrance of digital and social media” (Skogerbø et al., 191).

However, although renegotiations are taking place, and methods of communication have changed, political actors are still said to source power conventionally: “Sourcing from Twitter reinforces the power of the political elites to set the agenda of the news media – they are indeed still ‘leading the dance.’” Indeed, the group’s conclusion that “The agenda-setting hypothesis seems to remain robust and productive in an online environment” is one that allows for continuity between older definitions of agenda-setting and framing formulated during previous decades, and newer explorations and testing of these theories in the 21st_{century (Skogerbø et al., 190-195).}

Therefore, informed by the above cited sources, framing is defined by this thesis as the ways in which individual pieces of specific news content are presented to the public sphere, in this case via Twitter. Discussions of online framing attempts require analysis of the language (tone, word choice) used to introduce each piece of news, as well as the format in which this news is presented (text-only, audio, static image, video or otherwise). Agenda Setting is defined as the wider issues that society – a combination of media, politicians, the public, and those looking from outside political borders – deem to be relevant, and thus refers to the issues which come to mind each time a ‘new’ piece of news is published or presented.

One way of conceptualizing this is to think of framing as individual acts of information-presentation, acts that are noticeable for how they read, what language they use and what type of media links they share. Agenda-setting is what happens when all these frames, and the messages they imply, are considered as a larger whole, like bricks in a wall. Due to repetition of frames, this wider message is then more likely to come to mind each time a new frame is presented. In this way, a cycle of communication and understanding is established, one determined by traditionally powerful political and journalistic actors, and engaged with by individuals generally and online users of platforms specifically. When considered as a whole, this cycle of Framing and Agenda-setting allows us to engage with content published by political actors and news media sources on platforms like Twitter both on a micro and macroscopic level, without becoming embroiled in overly-pedantic academic re-definitions.

Further, framing of content is unavoidable, insofar that issues need to be explained, and the limits of linguistics force individuals to explain issues in a relatively brief fashion. Therefore, how content is framed each time it is presented for discussion and debate is crucial to how it is originally engaged with by others. This thesis claims that an accumulation of frames, and repetition of particular

(10)

10 stances or viewpoints by an individual or group, over time leads to the ability to better recognize said group or individual’s broader aims and political/social beliefs, otherwise known as their agenda. Thus, the creation and accumulation of frames leads to the setting of wider agendas. Therefore, while researchers often try to ascertain group’s key agendas through analysis and accumulation of individual frames, those creating frames already have an idea of their own agendas, and so analysis of specific frames can be potentially of huge use for researchers and ‘outsiders’. These arguments, although adapted for online research purposes and rephrased for clarity, follow those of Scheufele, Tewksbury and others cited above. Twitter is an ideal platform through which to conduct such framing discussions, due mainly to the 140 characters allowed by the platform per tweet, as well as the ability to include multimedia links with each individual frame.

2.2 Marres’ Issue Networks and Further Research Justification

Although there have been some debates surrounding the suitability of Social Media platforms as sites of political debate, there seems little doubt today that both the established press and the public see such platforms as useful places to find news information and engage in debates surrounding this information (Manjoo). Noortje Marres, when outlining her concepts of Issue Networks, offers useful clarification here:

“During the last decade we have witnessed the proliferation of new Information and Communications Technologies (ICTs) and the exponential growth of Civil Society Organizations (CSOs). The ‘network’ is one of the prime conceptual, practical, and technical sites where these two developments come together. Arguably the most important features of ICTs – of which the internet is a fundamental component, both discursively and logistically – is that they facilitate networked forms of organization.” (Marres, 3)

This claimed organization grounds later arguments in the realm of importance, suggesting that platforms (or ICTs as Marres calls them) are organized and interconnected entities, ones that allow both individual frames and wider agendas to pass through networks of online users and therefore to become relevant to societal debates. Further, Marres claims that “to account for civil society politics in terms of issue networks is to attempt to take seriously the specificity of networks as sites of politics. It is also an attempt to understand civil society politics as a practise in which substantial and technological considerations are closely intertwined” (4). Therefore, an ICT such as Twitter, one favoured by Civil Society Organizations such as Greenpeace and The Guardian, is a useful area of

(11)

11 research because it allows politically-motivated users to engage with the key agendas and frames of the still-strong political elite, in a manner less encumbered by geographical or economic limitations.

Continuing this theme, another of Marres’ key contributions to this thesis lies in three assertions. Firstly, “the issue network proposes that participants in such a network are connected to one another by way of the particular issue with which it is concerned.” This could be rephrased as ‘Content is connection’, or ‘issue instigates all’. This allows for the combined discussion of tweets from all user accounts, even if they do not explicitly tag (through the @ symbol) or engage with other users. Secondly, Marres states that “formatting issues (is) a crucial dimension of the politics of civil society”, a point that runs central to the idea that framing of events online is of importance and requires attention from researchers. Thirdly, she draws attention to “Extended configurations of actors and issues that are marked by antagonism”, thus echoing the political differences between the Trump administration on one side, and groups such as The Guardian and Greenpeace on the other (10-14). It is the disagreements, both explicit and implicit, that take place between these groups that places them within the same network, and even necessitates investigations into their framing and agenda-setting attempts.

2.3 Description of Twitter Specificities

“Our mission: To give everyone the power to create and share ideas and information instantly, without barriers” – Twitter ‘About’ Section

With 313 million unique visitors per month as per June 2016, Twitter is one of the largest Social Media sites in the world (Twitter.com; ‘About’). Its features include; a limit of 140 characters per Tweet; the use of unique usernames or ‘handles’ prefaced with the @ symbol; followers and following features, which can limit the other users that one pays attention to and interacts with; the use of hashtags to coordinate and include users in discussions, and; an allowance for the sharing of external media links within tweets. This combination of characteristics is particular to Twitter, and serves to both order and limit users’ actions and interactions on the platform. The combination of a 140-character limit with the ability to share external media links is important to highlight at this point.

(12)

12

2.4 Description of some wider issues in US Politics and Society

This subsection will briefly describe some recent US political and societal concerns, concerns which were especially prevalent during the 2016 US elections. This exercise will help to guide the methods and findings sections of the paper, and some overlap is expected between the issues prioritized during both the 2016 election and the period studied by this thesis. However, it is highly likely that the following issues will be superseded by more pressing issues during the President’s first 100 days. Therefore, this sub section is considered as a helpful beginning that serves to ground later claims and arguments, but not as a blueprint for events to come.

Firstly, it is important to recognize the specifics of the American political system, which traditionally operates in binary code: if you want to be involved in political decision-making, your opinions fall under either the Democratic or Republican umbrella. In American politics, therefore, both the setting of wider agendas and the framing of individual news events have often traditionally been either ‘red’ or ‘blue’ in colour. Although these oppositions are of course less binary in reality, and should instead be thought of as a spectrum that is infinitely sub-dividable down to each individual’s political beliefs, the two colours are useful indicators of the obvious split in US society and politics (Washington Post).

However, these red/blue binary definitions and debates surrounding them are called into question when we consider how new forms of media have developed since the turn of the 21st

century. The rise of platforms such as Twitter has coincided with a decline in print media revenue, and thus formerly authoritative mass media sources, ones loyal to the red-blue divide, are now forced to compete with newer sources of information for attention online, with priority given to revenue-generating clicks (Lu & Holcomb). These newer sources of information have not always adhered to the same journalistic standards as older media, and thus the phenomenon of Fake News emerged as a regular talking point over recent years (Posner & Neiwert). The 2016 US election was seen by many as a watershed moment in this regard, with many questioning the impact of algorithms, and whether or not Silicon Valley should be brought to account for the manner in which The Filter Bubble Effect caused users to become ‘surrounded’ by those who shared their political opinions (Borgesius et al., 1). The Oxford English Dictionary went as far as to make ‘Post-Truth’ their word of the year in 2016: “an adjective defined as ‘relating to or denoting circumstances in which objective facts are less influential in shaping public opinion than appeals to emotion and personal belief’”. The famous case of Macedonian news ‘outlets’ offering unsubstantiated articles to willing Republican voters, combined with the publications of those such as Breitbart News, meant that American internet-using voters were

(13)

13 sometimes reading ‘news’ sources that vastly differed from each other in terms of factual content (Kirby). This avenue of discussion led to theories such as the Red Pill/Blue Pill divide, which involved algorithms prioritizing content ‘liked’ by users, and therefore implies that both platforms and users themselves have agency when deciding which ‘facts’ to pay attention to (Borgesius et al., 2).

This being said, social media platforms cannot solely account for the perceived decline in the quality of public debates, or for the disagreements that arose over basic factual content in 2016. Indeed, it has been claimed more recently that Facebook’s algorithm was an easy target that did not cause the damage many originally claimed. (The Verge) Matt Taibbi is perhaps most notable for his scathing critique of his own profession, as well as of the inherent problems in the US political system that allowed a campaign such as Trump’s to gain momentum (despite its inability to engage with social issues beyond seemingly shallow, grandiose claims). His book, written in chronological chapters during the party primaries and presidential election, undermines confidence in the pre-Trump political system as much as it criticizes the “Insane Clown President” himself. Therefore, for the sake of a broader understanding of the less quantifiable aspects of Trump’s success, a success that has led to some questioning America’s role as leader of the Western Hemisphere (Smith), it is necessary to engage in a closer reading of some of Taibbi’s claims and arguments, which are indicative of the issues that defined the build-up to President Trump’s inauguration.

Taibbi’s essential arguments are this: Due to the huge financial backing that influential companies and groups invest in political campaigns, any candidate elected to political office is and has been compromised by corporate interests, often to the detriment of the wider public who it is theoretically their job to represent. Although this is a complicated claim that paints vast swathes of honest individuals with a tarry brush, the key point remains important and is echoed by various others,

The Guardian and The New York Times being notable examples. One of Trump’s key strengths, as the

then-candidate was extremely aware of, was his relative autonomy from big business interests (besides his own). Taibbi spends much time and effort explaining how this was framed by the Trump campaign team to the ‘ordinary’ American voter, who for so long has had to watch as Washington’s political elite seemingly concerned itself only passingly with those outside of its spheres of influence.

This argument makes a lot of sense when trying to understand the harnessing of the wave of popular support Trump rode to power, a wave that consisted primarily of conservative voters and supporters from outside the inner circle of Washington (which Trump famously labelled ‘The Swamp’). This is especially true when they are correlated with the views of Bernie Sanders, whose campaign for the Democratic nomination centred around the same core message as Trump’s: Washington is

(14)

self-14 involved and over-reliant on corporate funding. Although this framing of events is the point of departure for Sanders and Trump, it is nonetheless an important similarity (Taibbi 58).

These, however, are just some of the key issues surrounding Trump’s election. If it is accepted that Hillary Clinton did not engage with or convince voters as she should have, then we must ask: what did Trump do that was so appealing? The answer, although layered and complicated perhaps beyond full comprehension, lies partly in his lack of outside funding, and therefore perceived lack of reliance on others. This was framed to voters as a huge positive, with polls suggesting that Trump would represent a “new beginning” for previously disenfranchised rural Americans in particular. The answer also lies partly in Trump’s relationship with the US media and their inability to originally recognize the threat he posed to their profession. Trump regularly framed Media and news outlets as “Fake”, “Bad” or “Liars”, frames that undermined any negative reports of Trump and his team in the eyes of some voters. Most obviously, it lies in Trump’s ability to create frames that resonated with various sections of US society.

The 2016 election, and the primaries before it, were notable for their lack of in-depth political debating (Patterson). Instead of speaking extensively about social reforms, or outlining detailed election promises, candidates were dragged into debates about hand-size, unreleased tax returns and fearmongering about dangerous minorities. It has been well-documented that Trump’s Fleisch-Kincaid scores (readability tests designed to indicate how difficult a passage in English is to understand) were the lowest for any presidential candidate, and that the Republican primary debates were notable for their lack of socially relevant debate (Sedivy). Although difficult to prove, it is tempting to state that this was - and is - an ephemeral period in US politics, with real issues being side-lined in favour of consumable, bite-sized quotables, which proliferated across all media, both social and otherwise (Strauss). Such quotables often spread virally on Social Media platforms such as Twitter, with recent studies suggesting the more serious aspects of ‘Meme Culture’ and its effects on the views of US voters (PepetheFrogFaith).

This all implies that New Media platforms such as Twitter are hugely important disseminators of framed ideas, which in turn can affect both the political beliefs and voting habits of individuals. This is recognized by and reflected in the anxiety and regret exhibited by liberal media outlets during and directly after the 2016 election results were announced. Various media outlets, such as The Guardian,

The Huffington Post, The New York Times, and others, all published opinion pieces that agreed on one

point: Trump’s election was an unmitigated disaster, one that they themselves had contributed to by refusing to recognize the dangers of applying old mass media journalistic standards and expectations to behaviours surrounding these newer, faster technologies.

(15)

15 Trump’s election also corresponds with global fears about the rise of populism and emotionally-driven politics. Dutch journalist Rob Wijnberg wrote a convincing piece in the wake of Trump’s election, in which he outlines the need to totally rethink how media outlets function in the 21st_{century, and specifically how they create economic value. Such pieces reflect the wider}

international community’s realization that social media platforms have hugely impacted the way that news, facts, and debates are presented to publics. Therefore, these arguments also reflect Mass Media’s anxieties surrounding frame building and agenda-setting processes currently occurring in the Western world. Such concerns allow us to view our research questions as worthwhile investigations; if New Media has the power to help realign journalistic standards, through content-heavy framing practices and the ready accessibility of ‘facts’ online, then an investigation of how and if some of these framing practices are carried through to 2017 on Twitter has potential to add worthwhile analyses to the conversation.

2.5 Description of Users and Pre-Research Expectations

2.5.0 Introduction to Users

It is now time to briefly introduce the groups studied in this paper, to gain preliminary insight into the issues each group may prioritize on Twitter during the first 100 days of Donald Trump’s presidency. As the research question and its justifications at the beginning of this paper makes clear, the Twitter accounts of three groups will be analysed; President Donald Trump, Greenpeace USA, and The Guardian USA. More specifically, the following Twitter handles will be investigated:

• @RealDonaldTrump • @POTUS

• @GuardianUS • @greenpeaceusa

Choosing both the President’s official and personal Twitter accounts allows for investigation into the contrasts between how governmental authorities frame events, and how the individual in question does so. It is no secret that such prolific tweeting from the leader of America is unprecedented and perhaps unique amongst global leaders (Garun), and so it would be remiss not to discuss the ways in which each compare with the other. This is seen as how two parallel-yet-separated

(16)

16 elements of a President’s frame-making arsenal work together, to set agendas for public discourse. Further, much contemporary news analysis has highlighted the idiosyncratic nature of Donald Trump’s tweeting habits, and so contrasts are expected not just in content of posts, but also in the language and focus of each post. In other words, although Trump’s private and POTUS accounts are expected to contain the same agenda-setting attempts, insofar that they will focus on the same broader issues from the political perspective of Trump and his allies, the specific frames composed to carry out this agenda-setting is expected to contrast significantly between accounts.

Choosing The Guardian and Greenpeace were relatively straight-forward choices, insofar that both groups position themselves as opponents of Trump and his supporters’ political and social values. They both also post regularly to Social Media platforms, have significant follower bases, and are concerned with political developments, thus conforming to Marres’ assertions regarding the importance of “networked forms of organization” (3). The Guardian is a fair representation of the established left-leaning media, with all the labels and biases such a phrase naturally implies. Greenpeace has a long history of opposing perceived social injustices, something indicated in the name of the organization and a brief glance at the history of actions undertaken by its members (Greenpeace Website). The USA-specific accounts of both Greenpeace and The Guardian were chosen in order that as many tweets and posts as possible would be relevant to the conversation.

Inclusion of the POTUS’ first 100 days in our research question(s) came about for multiple reasons: Historically, the first 100 days of each US President’s time in office is viewed as critical. It is when election promises are prioritized and platforms are laid for the remainder of a President’s term in office (Watkins). This dates back to Franklin D Roosevelt, who used his first 100 days to pass sweeping legislative reform to combat the economic depression of the late 1920s and early 30s. Since this time, the 100 days has become a key feature of each presidency (Harvard Business Review).

Trump’s 100 days are no exception to this rule, and indeed may perhaps come under more scrutiny due to the seemingly unique nature of his election campaign. Directly after his inauguration, the new US President began to implement a series of controversial changes, such as changing the White House’s website content and rhetorically committing to following up on his pre-election promises to “build a wall”. These initial actions have caused fierce debate amongst American citizens, as well as in the international media (Taylor). If Trump’s first 100 days culminate on 29 April 2017, this allows for a Master’s thesis paper that closely analyses the framing characteristics of social media content published during this time.

(17)

17

2.5.1 Greenpeace description and original expectations

Greenpeace USA offer a lot of information on their website, with the ‘About’ page a useful point of reference:

“Our mission: Greenpeace is the leading independent campaigning organization that uses peaceful protest and creative communication to expose global environmental problems and promote solutions that are essential to a green and peaceful future.” – Greenpeace ‘About’ page.

“We ‘bear witness’ to environmental destruction in a peaceful, non-violent manner. We use non-violent confrontation to raise the level and quality of public debate. In exposing threats to the environment and finding solutions we have no permanent allies or adversaries. We ensure our financial independence from political or commercial interests.” - Annie Leonard, Greenpeace USA Executive Director, quote taken from Greenpeace ‘About’ page

“Our fight to save the planet has grown more serious — the threat of global warming, destruction of ancient forests, deterioration of our oceans, and the threat of a nuclear disaster loom large.” – Greenpeace ‘About’ page

Regarding agendas the Greenpeace USA Twitter account will attempt to promote; quotes like the ones above suggest that Greenpeace is true to its name, prioritizing issues such as environmental protection, climate conservation, limiting pollution, and other similar agendas. Their claim to “creative communication” is of interest here; later sections of this paper will discuss this creativity as seen through the groups’ Twitter account. Indeed, a glance at said account’s ‘Bio’ section offers further important details: “Fighting for a greener, healthier world, no matter what forces stand in our way. #resistoften #lovetrumpshate”. Such references to ‘resistance’ and using the president’s name as a pun suggest that this group will oppose vociferously any attempts by the POTUS to alter environmental legislation. It is not expected that the group will engage with other aspects of the President’s decision making, aspects such as Foreign Policy decisions, unless they directly engage with and affect Greenpeace’s environmental concerns.

2.5.2 The Guardian

description and original expectations

In contrast to @Greenpeaceusa, @GuardianUS is expected to take a much wider view of events in the United States, and therefore should report on almost every detail of the new President’s

(18)

18 time in office. As a branch of the wider Guardian Media network, @GuardianUS shares the aims and values of its affiliates, except with a focus on events in America. Although not among the studied Twitter handles for this paper, the Bio of @Guardian is instructive at this point: “The need for independent journalism has never been greater.” Such a claim, when combined with the Bio of @GuardianUS - “The @Guardian's US coverage, conversations and reporters. Share stories with us securely and confidentially at https://securedrop.theguardian.com/” – suggest that this news outlet feels its role to be of utmost importance in the current US political climate. It can be inferred from these Bios, and from a glance at The Guardian’s website, that the Trump Administration is deemed to be in opposition to the priorities held by The Guardian and its journalists. Although these values are not explicitly stated on their website, the organization goes as far as to call itself “the world’s leading liberal voice” in the ‘About’ section of its US website. This implies it will have far more in common with Greenpeace than with President Trump and those who support him, an implication that becomes obvious having read even a small number of @GuardianUS tweets.

2.5.3 President Trump description

and original expectations

Donald J. Trump was the Republican nomination for President in the 2016 US Presidential Elections. Notable for his brazen personality, wide business portfolio, and lack of previous political experience, Trump defeated Democratic candidate Hilary Clinton and was thus elected the 45th

President of the United States, taking office in early 2017 (Donaldjtrump.com). Having gained wide coverage for his use of Twitter, through the handle @RealDonaldTrump, the new president became associated with the handle @POTUS once he came to office in January 2017. Both these handles therefore theoretically represent the same individual, with an expectation that @POTUS will concern itself with less personal issues and opinions than those presented by @RealDonaldTrump in the past.

Of course, it would be of huge surprise should either @POTUS or @RealDonaldTrump not subscribe wholly to the ideologies, priorities and rhetorical devices of President Trump and his wider administration. It is expected that both these users will Tweet frames that forward the agendas of Trump, his family, the Republican Party, and those associated with it. It is expected that the platform will be used to offer positive accounts of legislative changes and events participated in by the President over the first 100 days of his time in office. The other side of this coin should also be mentioned here; it is expected that both the President’s accounts will use the platform to negatively

(19)

19 frame the actions of others, and to criticize those who oppose the President, whether that be political opponents such as Hilary Clinton, outgoing President Barack Obama, or those news outlets who choose not to praise the actions of the new administration. The exact nature of these frames remains to be seen, but concepts such as ‘Fake News’ and the labelling of ‘Liars’ are expected to carry over from the 2016 Election, considering the President’s much-discussed idiosyncratic Tweeting habits.

2.5.4 Summary of Pre-Research Expectations

Therefore, this paper expects two vociferously ‘pro-Trump’ groups (@RDT and @POTUS), and two defiantly ‘anti-Trump’ groups (@Greenpeaceusa and @GuardianUS) to emerge. Although the tweets of @RDT and @POTUS are expected to concur and overlap noticeably, both in style and content, the other two groups are not expected to share many priorities beyond being both ‘Anti-Trump’ and environmentally aware. This is because of @Greenpeaceusa’s relatively narrow focus on environmental issues, which contrast with @GuardianUS’ wider reporting obligations, which include climate and environmental concerns. These expectations will be discussed and compared with the actual findings in later sections of this thesis.

Finally it is clarified here that, during the following sections, shortened names will be used for each user. @RealDonaldTrump will be called @RDT, @POTUS will remain @POTUS, @Greenpeaceusa will be shortened to @Green, while @GuardianUS will be referred to as @Guard.

(20)

20

Section 3: Methodology

3.1 Key research features to guide Methods

This section will describe and justify the methods used to obtain data for the later Findings and Discussion sections of this paper. Before these methods are detailed, a brief explanation of some key research features will serve to ground and justify future explained methods. These key features are:

1. A focus on individual tweets as ‘standalone’ topics of research. If each post is considered as a piece of framed information, one that is intended to correspond to wider social and political agendas, it is necessary to analyse the methods used to create specific frames. As well as being a crucial discussion aspect in its own right, analyses of individual tweets allows for a discussion that eventually moves from micro to macro and back again. Again, the ‘bricks in a wall’ analogy is useful here. It is expected that each user, due to the different wider agenda it subscribes to, will frame events in a different manner to other users. Twitter’s platform informs this research choice - the combination of 140 characters with an ability to link external media offers an ideal setting to study the framing of issues by politically or socially motivated users. Similarly, it is expected that frames created by @POTUS, @Guard and @Green will conform with set ‘guidelines’ that are pre-determined by company or White House strategists. It is expected that @RDT, although perhaps conforming to an accidental style or template, is much more likely to offer frames that contain less external links, more personal opinions and to have an overall less ‘polished’ effect.

2. Examination of framing and re-framing attempts as part of a wider prioritization of certain off and online events. It is expected that President Trump’s accounts will proactively engage with concrete events in Twitter discussions, providing original tweets that positively frame his administration’s decisions, decisions that will be then contested by the other pages in this study. This cycle of framing and re-framing may perhaps obfuscate the portrayal of actual events in America by these users, with certain events prioritized and re-emphasized ahead of others. For example, should the White House sign controversial legislation, it is expected that social media will be used to appease concerned voters, but will be used by Trump’s opponents to highlight his perceived flaws and general unsuitability to lead. These legislation debates might conceivably overshadow the framing of visits or other decisions made by the Trump administration during the same time. This requires both knowledge of offline events during

(21)

21 the 100 days, as well as analysis of individual frames, and so can be said to be more macroscopic than the previous key feature detailed above.

3. The above two key research features will be supplemented by analyses of hashtags, retweets of other users, and other methods of communication creation that the Twitter platform allows for. These analyses will be conducted as supplements to the two points explained above, due to the necessary limits imposed by the research brief for this thesis and research question explained in previous sections. Therefore, although hashtag and link frequency investigations, for example, will be conducted according to the limits imposed by the tools used to gather Tweet data, conversations of such will not be prioritized unless they offer new conversational material that enriches discussions central to the research question.

All three of the above points are included here so that an audience may be reminded of the specific priorities of this paper before the following sections occur. In short, this paper is intent on discussing different tweet-creation methods, with the purpose of discussing how these methods contribute to the establishment of users’ wider priorities during President Trump’s first 100 days in office.

3.2 Creation of Manual List of Concrete Events

To best understand the main issues of the period, it was decided to establish a list of the main topics of discussion during each of these 100 days. Therefore, a list of each day’s developments, both in and outside the White House, was created, with the intention of comparing offline and online actions during the period with frames provided by the four studied users.

A variety of news outlets extensively documented Trump’s first 100 days, including; The

Guardian, USA Today, Fox News, CNN, ABC, and others. Therefore, there is no shortage of reported

sources to aid in the construction of a timeline of noteworthy or concrete events during the period 20th_{January 2017 until 29}th_{April 2017. The following list of concrete or salient events was constructed,}

using online media sources The Guardian, and Fox News. To avoid any pre-existing journalistic biases, no one source was relied exclusively upon when creating this list. Although this cannot be called an ‘exhaustive’ list of all noteworthy events that occurred in and around the Trump Administration during the aforementioned dates, it is an extensive list and thus functions as a useful indicator of the most important events that occurred during each group period. Although attempts have been made to explain background details, some aspects of this list lack context, and may thus be lacking in

(22)

22 information for a reader unused to the US political circuit. This is unavoidable, as a thorough ‘beginners’ explanation of each individual mentioned would distract from the point of this list – to summarize all important ‘concrete’ events that may have been framed by the studied users on Twitter. This list can then function as a base from which to investigate how frames offered on Twitter relate to and interact with ongoing events in US politics during Trump’s first 100 days. The sources for this list are fully referenced in the Bibliography section of this thesis. It is noticeable that these 100 days are divided into ten groups of ten days in the following list. Explanations for this choice will occur directly after the list, and will then be linked to the further methods described in the following sections.

*Note – As the specific contents of the ‘Concrete Events’ list are not useful at this exact point in the paper, and to create space for a larger Discussion section, the list in its entirety can be seen in the appendices to this paper. Specifically, Appendix 1 provides the entire list of Concrete Events.

3.3 Creation of 10-day groups

Having established the list of salient or concrete events that could potentially be framed online by @RDT, @POTUS, @Green and @Guard, the entire 100-day period was divided into ten groups of ten days - labelled in chronological order and referenced from here on as Group 1, Group 2, and so on. This is considered an original method for this thesis, with no academic papers offered as justification. The benefits of this division were multiple: it limited the volume of tweet data studied at one time during the research phase, thus allowing for a clearer division of labour. This meant that a focus on daily events became easier, and so frames were prioritized. This division also meant that more regular ‘stops’ occurred, resulting in a discussion section that remained more focussed on the research question over the entire period studied. It also allowed for easier comparisons between user handles’ frames of specific events, insofar that each Group period was associated with less concrete events than the 100 days as a whole. Although dangers exist in such a fabrication of time-groups, such dangers are judged to be limited as long as the fabrication of these Group periods is kept in mind and inevitable overlapping of content is recognized as research progresses. For the exact date period of each ten-day group, see Figure 1.

(23)

23

Figure 1: Date range of each ten-day group period

Once each of the ten groups has been broadly defined according to the key or ‘concrete’ events that took place during its time, the next step is to analyse the Twitter content of all four users, using the DMI-TCAT tool.

3.4 Specific Methods of Capture

It is now necessary to describe the processes carried out to obtain data from Twitter for these discussions. The Digital Methods Initiative was crucial in this regard. Self-described as “one of Europe's leading Internet Studies research groups”, DMI is “comprised of new media researchers and PhD candidates, it designs methods and tools for repurposing online devices and platforms … for research into social and political issues” (DMI Website). Specifically, the DMI-TCAT tool was utilized for data accumulation for future discussions. This section will explain the methodologies behind the research carried out into all four mentioned Twitter handles, with the aim of establishing a clear guideline as to how the findings discussed in later chapters came about.

The creators of the Digital Methods Initiative Twitter Capture and Analysis Toolset (DMI-TCAT) state that the tool “allows users to retrieve and collect tweets from Twitter and to analyse them in various ways” (DMI Website). These tweets are not open to the public and require a password and username to access them, in line with Twitter’s terms of agreement. Having contacted DMI through my university, and having explained my thesis as well as my need for tweet data, I was provided with access to a dataset that consisted of all tweets from the four previously mentioned Twitter handles, dating from 1st_{January 2017 and continuing to the present day. This date-range incorporates each of}

Trump’s first 100 days and thus can be said to be an almost exhaustive tweet dataset. The term ‘almost’ is used here with good reason, as will be explained in the “Methodology Problematics” subsection.

(24)

24 In total, 9101 tweets were collected that fall within the 100 day-range. This full dataset was exported as .csv files and opened using Microsoft Excel Spreadsheet Software, as per the recommendations of Erik Borra in his TCAT explanation paper and corresponding YouTube explanation videos.

3.5 ‘Masterlist’ Creation

In order that all data could be referenced without hesitation, and to ensure full confidence in the dataset, a full selection of tweets from all Twitter handles (@RDT, @POTUS, @Greenpeaceusa, @GuardianUS) was downloaded using the TCAT tool and saved on both hard drive and Google Drive. This entire set was labelled ‘masterlist’ and was neither edited nor altered once the dataset had been created and downloaded. All research explorations subsequent to this took a copy of this ‘masterlist’ as their starting point. This copy was pasted into a new spreadsheet each time, and so the ‘masterlist’ remained unchanged throughout.

Having copied all tweet data from said ‘masterlist’ onto a new Excel spreadsheet, a few steps needed to be taken before further research investigations could be carried out. Firstly, all tweets from @RDT were separated from the rest, and copied into a new sheet. This was done by ordering all tweets according to their Twitter handle, then copy-pasting all 507 @RDT tweets into a new page. This list was saved as filename ‘@RDT All Tweets’. Then, this @RDT data was put into chronological order, from earliest to latest. This established a chronological list of all tweets from account @RDT, which could be then analysed, and later separated into Groups 1-10, again via copy-pasting tweets falling within each group into new Excel sheets. Once this was completed, the base was ready and research investigations could begin. The methodologies previously described for capturing and ordering the tweets of @RDT were replicated to capture data from @POTUS, @Guard, and @Green.

Having created four individual groups of tweets, which were further subdivided into ten subgroups, I was then reluctant to subdivide any more, in case the application of too many chronological borders affected in any way the framing discussion. Next, all tweets were manually perused, with the intention of creating Tweet categories that could direct and organize both the Findings and Discussion sections of this thesis. This category creation involved the use of precedents set by previous researchers, a thorough description of which can be found in the following subsection.

(25)

25

3.6 Creation of Tweet Categories

In their 2012 work ‘Information, Community, and Action: How Non-profit Organizations Use

Social Media’, Saxton and Lovejoy analyse the tweets of 100 American Non-Governmental

Organizations (NGOs), with the aim of gaining insights into how such organizations use Twitter to engage with stakeholders and the public. Having categorized NGO tweets into three types - Information Tweets, Community tweets, and Action Tweets - they conclude that “the adoption of social media appears to have engendered new paradigms of public engagement” for these NGOs (337-353).

These categorizations were persuasive and thus used as a basis for investigations conducted in this thesis. They applied most obviously to the tweets of @Greenpeaceusa, an NGO with many of the same characteristics as those described by Saxton and Lovejoy. Therefore, in line with this model, I categorized @Greenpeaceusa’s tweets according to three categories: Information, Community, and Action.

- Information Tweets are published to inform followers and do not ask for engagement beyond consuming the tweeted content. Such content includes any information that can be provided to users with the help of 140 characters and external links.

- Community Tweets ask for online input, ask questions of followers or directly engage with followers through tags (@), and thus foster conversation and/or community spirit specifically on the Twitter platform.

- Action tweets call for some type of involvement outside of the Twitter Platform, for example to download a template, sign a petition or to attend a rally offline. These actions can take place on different websites, or offline.

I then added a fourth category, ‘RT’. This category contains any tweet prefaced with the letters RT, which means that they are individual tweets created by other users and repeated by Greenpeace. This was done because ‘RT’ was found to be extremely common, making up 83 out of the top 100 most engaged-with @Green tweets. These are seen to be important because here Greenpeace functions as a carrier of others’ frames as well as a creator of original ones. @Green’s tweets were then sorted according to these four categories, the findings and wider discussion of which can be found in later sections.

(26)

26 Once the results from the above categories were analysed, this process was deemed worthwhile, and so Tweets from the other three handles were also broken down into categories. It became immediately apparent that, due to the unique nature of each of the three other users, new categories were needed. Due to a lack of obviously applicable academic guidance on this matter, the following categories were created:

@RealDonaldTrump: Community, Foreign Policy, Internal Affairs, Media & Opposition, Other

- ‘Community’ Tweets are intended to directly address the President’s followers and voters. Often featuring generalities and broad claims, these tweets regularly mention; Speeches, Rallies, Presidential ceremonies, Congratulations to assorted groups, “Thanks” and other such messages.

- ‘Foreign Policy’ tweets are concerned with American foreign policy. Tweets in this category reference foreign nations, other world leaders, and comments on events outside of the United States. Tweets regarding foreign visitors to the White House were put in this category also. - ‘Internal Affairs’ tweets are those that referenced issues within the United States. This

includes statements on policy, legislative actions, comments on movements and developments within the country. Tweets concerning the proposed Mexican Border Wall and the infamous travel ban from various Muslim countries were also included here, as both these issues involve limiting immigrants and their actions within the USA.

- ‘Media & Opposition’ tweets were those that took aim at Trump’s opponents. This was a wide group, including; The New York Times, CNN, and other media outlets; Members of the Democratic Party; Former President Barack Obama; and many others. These tweets were almost always condemning or pessimistic in tone.

- ‘Other’ tweets were those that did not fall comfortably into any of the above categories. The rarity of such tweets meant that creating specific sub-groups felt pedantic.

@POTUS: Plans & Positivity, Manufactured Quotes & Speeches, Media & Opposition, Visitors, Retweets, Other

- ‘Plans & Positivity’ tweets functioned similarly to Community Tweets in the @RDT and @Greenpeace categories, insofar that they addressed followers/voters and dealt with generalities. Declarations of administrative appointments and promises to “Make America Great Again” also fell into this category.

- ‘Manufactured Quotes & Speeches’ tweets were those that on the surface seemed connected to President Trump, but did not match with the tweeting style established by @RDT, thus

(27)

27 suggesting that they were either polished by members of White House staff, or not seen by Trump at all. Tweets announcing planned appearances or offering media links to live speeches were also included here. Any tweet that did not offer anything except for polished advertisement was included here.

- ‘Media & Opposition’ tweets were those dealing with Trump’s opposition, including; Media Outlets, Political opponents, and various others. The conceptual basis for this category was the same as the @RDT category of the same title.

- ‘Visitors’ tweets were those detailing visits by foreign representatives to The White House. This included foreign elected leaders, royalty, and assorted diplomats.

- ‘Retweets’ were those tweets prefaced with ‘RT’. Like the @greenpeaceusa category of the same name, these tweets were noticeable due to their high frequency and so needed a category of their own.

- ‘Other’ Tweets were those that did not comfortably fall into those already described above.

@GuardianUS: Administration, Climate, Foreign Policy, Internal Affairs, Media Problematics, Societal Concerns, Russia, Anti-Trump, Other

- ‘Administration’ tweets were those that dealt with members of Trump’s team, cabinet, advisors, and others. Some of the individuals referenced were: Daughter and advisor Ivanka Trump, advisor Steve Bannon, Supreme Court Judge Neil Gorsuch, Press Officer Sean Spicer, and various others.

- ‘Climate’ tweets were those concerned with the environment, the impact of climate change and other similar concerns.

- ‘Foreign Policy’ tweets were those concerned with affairs outside of the United States, but still of potential concern to the United States. Examples include events in the Middle East, China, and North Korea.

- ‘Internal Affairs’ tweets were those that concerned themselves directly with governmental and administrative actions within the United States, and thus mirrored quite closely the @RDT and @POTUS categories of the same name.

- ‘Media Problematics’ tweets were those that dealt specifically with unreliable media reporting. Although there were few tweets that comfortably fell within this category, it was important to include in order to more directly engage with some of @RDT’s wider “Fake News” claims.

- ‘Societal Concerns’ tweets were those that dealt with issues outside of the immediate political decision-making sphere in the USA. Examples include features on homelessness within the

(28)

28 country, and ‘state-of-the-nation’ style references. This category was created to allow a sharper definition of ‘Internal Affairs’ tweets across user-handles.

- ‘Russia’ Tweets were those concerned with alleged Russian interference in the 2016 elections and investigations to possible links between the Trump administration and Russia. Tweets referencing Russian politics and society were also included here. References to former National Security Adviser Mike Flynn were included here also, due to the ongoing investigations by intelligence agencies into his alleged connections to Russia.

- ‘Anti-Trump’ tweets concerned themselves with President Trump specifically, the political decisions enacted by him and claims made by him online and in the media. These tweets included both political and ad hominem attacks, by Guardian writers and those from various sections of American society.

- ‘Other’ tweets were those that did not fall comfortably into any of the above categories, but still concerned themselves with various themes connected to this thesis.

3.7 Methodology Problematics

At this point it is important to note some specifics and problematics regarding the @GuardianUS account. Firstly, although 6158 Tweets were collected from this account, many were unconnected entirely from this thesis, because they concerned themselves with sporting issues, other nations’ environmental concerns or otherwise. Therefore, the original 6158 tweets were analysed and reduced to 2196 US-centric tweets, so that statistics and figures could be as accurate as possible. These 2196 tweets were then assigned to the categories described above.

Secondly, the DMI-TCAT tool did not collect all tweets from @Guard for the full 100-day period. Tweets collected and discussed began on the date 15th_{February 2017, which was day 27 of}

Donald Trump’s first 100 days. Therefore, any discussions about the period 20/01/17 – 14/02/17 will not be able to include @Guard. Although this is far from ideal, the volume and characteristics of the tweets available after 15/02/17 means that a discussion that recognizes its limitations can still incorporate the incomplete @Guard dataset.

(29)

29

3.8 Methodology Conclusion

This methodology was carried out with the intention that a reliable dataset could be created, one that was both ‘mappable’ and divided into groups small enough to identify intricate details, but nonetheless offered enough research opportunities to make the subgroups worthwhile and the analysis of framing attempts as open as possible.

(30)

30

Section 4: Findings

4.1 Findings Introduction

This section will present the original findings from the methods described in the previous section. Once all notable findings are presented, the following section will offer further discussion based on these findings. This section will mention a series of figures and statistics, with the intention of providing a base from which a meaningful discussion section can be created.

It is noted here that, although many research investigations were carried out, some were more fruitful than others. This findings section will outline the main investigations carried out into the datasets acquired via the DMI-TCAT tool, findings which can later contribute usefully to the Discussion section of this paper. The ‘Categories’ findings are of particular importance to later discussions, with the figures mentioned in sections 3.2 – 3.6 offering multiple opportunities for analysing the framing methods of the four users studied in this paper.

4.2 Basic Stats Overview

To begin investigations, a Twitter Basic Stats Overview was conducted, based on information visualizations offered by the DMI-TCAT tool:

1. @GuardianUS:

(31)

31 As explained earlier, the dataset acquired for user @GuardianUS using the DMI-TCAT tool did not capture data for the entire 100-day set. Data for this user begins on 15/02/17, day 27 of the 100 days. From this day onwards, a total of 6158 tweets were published by @GuardianUS.

The Guardian USA tweeted 6158 times during the 84 days for which data was available, on average just over 73 times per day. This is the largest tweet volume over the four studied accounts. Figure 2 offers the following information:

- Highest daily Tweets: 137 Tweets on 27/02/17. - Lowest daily Tweets: 45 Tweets on 02/04/17. - Total Followers: 193,544

- 6016 Tweets of 6158 contained links to outside media sources, 97.7% of total tweets.

2. @Greenpeaceusa:

Figure 3: Overview of @Green’s overall tweet stats for 100-day period, courtesy of DMI-TCAT

Greenpeace USA tweeted 1934 times during the 100 days, on average just under 20 times per day. This is the second-largest tweet volume over the four studied accounts. Figure 3 offers the following information:

(32)

32 - Lowest daily Tweets: 5 Tweets on both 04/03/17 and 23/04/17.

- Total followers: 205,174

3. @RealDonaldTrump

Figure 4: Overview of @RDT’s overall tweet stats for 100-day period, courtesy of DMI-TCAT

@RealDonaldTrump tweeted 509 times during the 100 days, on average just over 5 times per day. This is the second-lowest tweet volume over the four studied accounts. Figure 4 offers the following information:

- Highest daily Tweets: 13 Tweets on 20/01/17. - Lowest daily Tweets: 0 Tweets on 15/04/17. - Total Followers: 32,676,700

(33)

33 4. @POTUS:

Figure 5: Overview of @POTUS’s overall tweet stats for 100-day period, courtesy of DMI-TCAT

@POTUS tweeted 500 times during the 100 days, on average 5 times per day. This is the lowest tweet volume over the four studied accounts. Figure 5 offers the following information:

- Highest daily Tweets: 39 Tweets on 01/03/17. - Lowest daily Tweets: 0 Tweets on multiple occasions. - Total Followers: 18,923,299

4.3 Category Findings

4.3.0 Summary of Created Categories

Next, as explained in the Methodology section, the tweets from each Twitter account were categorized according to their overarching message, with the intention of investigating how Tweets from each account attempt to frame issues across the first 100 days of Donald J. Trump’s presidency. Having originally modelled our methods on those of Saxton and Lovejoy, who analysed NGO tweets

(34)

34 by dividing them into three categories, these categories were then expanded according to the specific nature of each account studied in this dataset. As explained in detail earlier, these categories were:

- @RealDonaldTrump: Community, Foreign Policy, Internal Affairs, Media & Opposition, Other - @POTUS: Plans and Positivity, Echoed Quotes or Speeches, Visitors, Retweeted Content,

Media & Opposition, Other

- @greenpeaceusa: Information, Community, Action, Retweet

- @guardianUS: Administration, Environment, Foreign Affairs, Internal Affairs, Media Problematics, Societal Concerns, Russia, Anti-Trump, Other

Once each tweet from each account was assigned to a category, this allowed for the following original findings.

4.3.1 Category Findings @RealDonaldTrump

@RDT tweets a total of 504 times across the 100-day period. These 504 Tweets were categorized into more specific sub-categories. The breakdown of each group can be seen in Figures 6 and 7:

Figure 6: Breakdown of @RDT’s Tweet categories over 100 days

Framing Trump: How do The Trump Administration, The Guardian and Greenpeace USA frame issues on Twitter during the first 100 days of the Trump Presidency?