Medicalization online: How the media stimulate awareness on medical conditions

(1)

Medicalization online:

How the media stimulate awareness on medical conditions

Submitted in fulfillment of the requirements for the degree of Master of Arts (M.A.) by Mariessa Radermacher

Program: New Media and Digital Culture 1st_{Reader: Dr. Kaspar Beelen}

2nd_{Reader: Prof. Dr. Richard Rogers}

Student no.: 11731990

E-Mail: mariessa.radermacher@gmx.de

Abstract:

This master thesis provides contribution to research on why certain medical conditions become salient among society whilst others do not. A differentiation between rare and non-rare diseases was made in order to see, if only those medically prevalent among society become known. Five mechanisms and characteristics of those diseases well-known could be observed. This research showed, that the legacy media is still a driving force in wielding emphasis on topics considered to be important by society.

Keywords:

(2)

Table of figures

Figure 1: Matrix of diseases covered in this thesis ... 4

Figure 2: Two-step flow of communications (Katz and Lazarsfeld 183)...17

Figure 3: "Overview of Communicative Roles of Opinion Leadership in New Media Environments" (Schäfer and Taddicken 973) ...19

Figure 4: Google Trends HIV/AIDS 01.01.2016-01.01.2018 ...25

Figure 5: merged page-like networks of AIDS Healthcare foundation (green), Act Against AIDS (red) and Greater than AIDS (blue) ...30

Figure 6: AIDS/HIV Facebook post engagement 01.01.2016 - 01.01.2018 ...33

Figure 7: Greater Than AIDS Facebook post engagement count for New York Times articles 01.01.2016 – 01.01.2018 ...34

Figure 8: AIDS Healthcare Foundation Facebook post engagement for New York Times articles 01.01.2016 – 01.01.2018 ...35

Figure 9: AIDS/HIV: Articles per month and year (own representation) ...36

Figure 10: Google Trends ALS 01.01.2016-01.01.2018 ...37

Figure 11: merged page-like network ALS Association (red), ALS Worldwide (blue) and Cedars-Sinai: Amyotrophic Lateral Sclerosis (green) ...41

(4)

Figure 14: Google Trends Endometriosis 01.01.2016-01.01.2017 ...45 Figure 15: merged page-like network of Endometriosis Research Center (blue),

MyEndometriosisTeam (red) and Endometriosis Awareness (green) ...48 Figure 16: Endometriosis: Number of articles per month and year (own representation) ...51 Figure 17: Google Trends Behçet's disease 01.01.2016-01.01.2018 ...51 Figure 18: merged page-like network of American Behcet's Disease Association (blue), Behcet’s Disease Community (red) and Behçet’s Disease: You Are Not Alone (green) ...56 Figure 19: Behçet’s disease Facebook engagement 01.01.2016 – 01.01.2018...57 Figure 20: merged page-like network of Endometriosis (yellow), ALS (green), Behçet's Disease (blue) and AIDS (red) (own representation) ...59 Figure 21:AIDS/HIV Standardized and summed amounts of Facebook posts, The New York Times articles and Google Trends queries per month from 01.01.2016 to 01.01.2018 ...60 Figure 22: ALS Standardized and summed amounts of Facebook posts, The New York Times articles and Google Trends queries per month from 01.01.2016 to 01.01.2018 ...62 Figure 23: Endometriosis Standardized and summed amounts of Facebook posts, The New York Times articles and Google Trends queries per month from 01.01.2016 to 01.01.2018 ..63 Figure 24: Behçet's disease Standardized and summed amounts of Facebook posts, The New York Times articles and Google Trends queries per month from 01.01.2016 to 01.01.2018 ..64 Figure 25: updated Matrix of diseases covered in this thesis ...70

List of tables

Table 1: AIDS/HIV Top 3 categories according to number of nodes for each page-like

network investigated for Endometriosis ...26 Table 2: Top 5 Fan Count ranking with respect to Categories for AIDS Healthcare

Foundation, Act Against AIDS, Greater Than AIDS ...27 Table 3: Top 5 In-Degree Count ranking with respect to Categories for AIDS Healthcare Foundation, Act Against AIDS, Greater Than AIDS ...28 Table 4: Top 5 Betweenness Centrality ranking with respect to Categories for AIDS

Healthcare Foundation, Act Against AIDS, Greater Than AIDS ...29 Table 5: Top 8 Betweenness Centrality with respect to Categories of the merged page-like network for AIDS/HIV ...31 Table 6: merged AIDS network top 5 In-degree counts compared to categories ...32 Table 7: ALS Top 3 categories according to number of nodes for each page-like network investigated ...37

(5)

Table 8: Top 5 Fan Count ranking with respect to Categories for The ALS Association, ALS Worldwide, Cedars-Sinai: Amyotrophic Lateral Sclerosis Program ...38 Table 9: Top 5 In-Degree count ranking with respect to Categories for The ALS Association, ALS Worldwide, Cedars-Sinai: Amyotrophic Lateral Sclerosis Program ...39 Table 10: Top 5 Betweenness Centrality ranking with respect to Categories for The ALS Association, ALS Worldwide, Cedars-Sinai: Amyotrophic Lateral Sclerosis Program ...40 Table 11: Endometriosis Top 3 categories according to number of nodes for each page-like network investigated for Endometriosis ...45 Table 12: Top 5 Fan count with respect to categories for Endometriosis Awareness,

MyEndometriosisTeam, Endometriosis Research Center ...46 Table 13: Top 5 In-Degree count with respect to categories for Endometriosis Awareness, MyEndometriosisTeam, Endometriosis Research Center ...47 Table 14: Top 5 Betweenness Centrality with respect to categories for Endometriosis

Awareness, MyEndometriosisTeam, Endometriosis Research Center ...48 Table 15: Behçet’s disease Top 3 categories according to number of nodes for each page-like network investigated ...52 Table 16: Top 5 Fan Count with respect to Categories for ABDA, Behçet’s Disease: You Are Not Alone, Behçet’s Disease Community ...53 Table 17: Top 5 In-Degree count with respect to Categories for ABDA, Behçet’s Disease: You Are Not Alone, Behçet’s Disease Community ...54 Table 18: Top 5 Betweenness Centrality with respect to Categories for ABDA, Behçet’s Disease: You Are Not Alone, Behçet’s Disease Community ...55 Table 19: All diseases' average number of reactions per post ...58 Table 20: Merged diseases' networks top 3 categories...65

(6)

1. Introduction

Although the developed, western countries managed to reach and maintain a high level of general public health and hygiene (Kraus and Wilms 115), an increasing number of widespread diseases strains on society’s healthcare systems which also has an impact on people’s awareness for modern diseases (Wallenfels n. pag.). Besides an increase of most common and well-known diseases such as cancer or cardiac diseases as a result of lack of exercise (Wallenfels n. pag.), a so-called medicalization has taken place within the last couple of decades (Conrad 3). Medicalization describes the development in which symptoms that were not perceived as medical problems become treated and defined as such (Conrad 4). According to Conrad, the number of problems that become a medical condition increased enormously (3). AIDS/HIV, Endometriosis, Behçet’s disease and Amyotrophic Lateral Sclerosis (short: ALS) have been considered to be genetic disorders, diseases or illnesses for a long time. Nonetheless, medicalization shows that the general interest in health issues and topics must have increased over the past years. Partly as a result of medicalization, the percentage of the gross national product spent on health care has risen considerably, as well as the number of physicians per capita: The amount of US-Dollars spent on health care in the US “increased from 4.5 percent in 1950 to 16 percent in 2006” (Conrad 3); the number of physicians in the US grew from 148 in 1970 to 281 physicians per 100,000 people in 2003 (Kaiser Family Foundation 54). Zimmet and the World Health Organization (WHO) name this trend, that developed societies seem to follow, a chronic disease epidemic (Zimmet 302; 2. Background 2.1 The Global Burden of Chronic n. pag.).

From the perspective of media studies, there emerged certain phenomena among social media in the context of underexplored or newly discovered diseases. Also, long-known diseases suddenly become society’s object of interest. This is evident within the Ice Bucket Challenge, which took place in 2014. Before Chris Kennedy was nominated for the Ice Bucket Challenge and executed it on YouTube (Kennedy n. pag.), this challenge had nothing to do with ALS (Sifferlin n. pag.). Before, everyone chose their own charity to donate to. Kennedy then nominated his wife’s cousin, because her husband was suffering from ALS (Sifferlin n. pag.) and donated 100 US-Dollar to the ALS Foundation (Kennedy n. pag.). When his wife’s cousin posted her own Ice Bucket Challenge video using the hashtags #takingiceforantsenerchiajr and #StrikeOutALS, which were previously used for a non-profit baseball tournament to honor her husband, her video went viral (Sifferlin n. pag.).

(7)

Eventually this resulted in 115 million US-Dollar to be donated to the ALS Foundation (Impact of ALS Ice Bucket Challenge n. pag.). With this example alone, it is interesting to scrutinize why certain actions (on social media) go viral in the context of diseases and why others do not.

In the course of this master thesis, I attempt to identify the mechanisms that drive interest for certain diseases across media. It is interesting how rare diseases such as ALS gain widespread attention, whilst common diseases, such as Endometriosis (Mechsner 478), linger in obscurity and are only occasionally discussed. The focus on mechanisms does not preclude that some events may be of a unique nature and thus, only generate awareness once, whilst examine the respective disease continuously. For this thesis, the following four chronic diseases were selected: AIDS/HIV, Endometriosis, ALS and the Behçet’s disease. The reasons for this choice of diseases will be explained and justified further in section 3.

Proceeding from this, it is crucial to ask who or what ignites a change in public awareness. The main research question of this thesis can be determined as the following: What mechanisms determine or influence these levels of public attention over time? Overall, the research will aim to understand why and how certain chronic illnesses manage to become “famous” and what or who can be made responsible for such trends.

To narrow down the scope, this research concentrates on the United States and the English language only. The United States serve as main focus, because when looking into most liked Facebook pages (also run by private people), initiatives, and organizations, most are located or run by and from the United States. Of course, it was noted that some are only located in the United States but operate internationally. Nonetheless, a general focus on one country only also ensured one constant parameter and constant among all methods and data collected. Accordingly, it is a geographic anchor for the analysis. Another reason that makes the United States particularly interesting to look at, is that by 2000 approximately 45 % of the population suffered from at least one chronic condition (Anderson and Horvath 263). Moreover, 70 % of American adults state to be interested in topics such as health and medicine (Kennedy and Funk 2). Hence, a broad debate and interest for chronic conditions, (genetic) disorders and health topics in general could be expected. Another limitation encountered within this research are the limited options provided by the methods chosen. This means for example, that while gathering data with Netvizz, Facebook’s API changed and thus, certain data had to be omitted.

(8)

In the first section of this master thesis, a definition of the chosen diseases’ characteristics will be given. This means, defining rare, common, salient and non-salient diseases. This helps assigning the four diseases to either a combination of rare and non-salient-, rare and salient-, common and non-salient-, and common and salient type of disease. After this, the diseases themselves will be defined in both a medical and historical sense. Naturally, these definitions encompass a concise overview, leaving out all medical details available.

After this, the theoretical framework follows. I will concentrate on the theories of Agenda-setting in the context of social media, on Web 2.0, on produsage and on opinion leader theory. This is followed by a definition and description of the methods used.

For answering the research question, namely defining what factors and mechanisms determine or influence the different levels of public attention for the chosen diseases, multiple methods were used. This helped identifying both the social media dynamics as well as the legacy media’s level of attention towards the different types of diseases.

In order to get an impression of movements around these illnesses, I will conduct an analysis of the Facebook page-like network and its page-data via the Netvizz tool. From then on, the network and social connections within Facebook will be visualized via Gephi, a visualization tool for humanities. Moreover, the page data will be evaluated with respect to the pages’ most successful status updates and posting activities, for example. The second set of data worked with will be Google trends. The third set of data used is a qualitative analysis of the newspaper attention among The New York Times via The New York Times API. All methods (Google trends, Facebook page-data and the newspaper attention), that show a historical character within this research project, focused on the time frame from January 1 2016 to January 1 2018. Since all four diseases vary in prominence and prevalence, no common time frame could be found encompassing all important dates for every disease respectively. Since the interest of this work lies in new phenomena, the aim was to establish a time frame long enough to be able to make observations that might stretch over a year’s time and are not linked to any of the diseases directly. Consequently, this research can be assigned to both a historical and momentaneous approach towards how diseases become worth knowing within society. This research includes both new media approaches as well as methods dealing with legacy media.

(9)

A coverage of all types of (social) media would have exceeded the scope of this master thesis and thus, this research focuses on Facebook, Google and the legacy media, represented by The New York Times. Eventually, the results will be presented and subsequently interpreted and analyzed.

2. Definition of rare, common, known and unknown diseases

This matrix helps to identify the different types of diseases and what relation they bear to publicity among society. Therefore, the following table shows four quadrants occupied by one disease respectively. The x and y-axes are defined by level of publicity (x-axis) and prevalence of the disease (y-axis) (Figure 1).

Figure 1: Matrix of diseases covered in this thesis

In the following, each partial aspect of the x- and y-axis will be defined and thus, the four investigated diseases classified along this scale.

2.1 Rare disease

According to the European Commission, a disease is considered to be rare when one in 2,000 people is affected by it. Although this number suggests that only few people are affected in the EU, projected on the whole population in all 28 member states, the absolute number of affected people could go up to 246,000 people and the disease is still considered to be rare. While the ratio of 1 in 2,000 people is the most common one, there are other diseases that only affect one person in 100,000+. Considering this, with 5000 to 8000 different rare diseases known so far, approximately 27 to 36 million people are affected by a rare disease in the EU (Policy - European Commission n. pag.).

(10)

In the US, on the other hand, a disease is considered to be rare when it affects less than 200,000 people. A second definition in the US is, that a disease affecting more than 200,000 people, but has no reasonable outlook on recovering the costs for developing or making available a drug from sales in the US, is also considered to be rare (Commissioner n. pag.). Concluding from this, cultural and governmental distinctness define rare diseases differently among states and unions.

2.2 Common or non-rare disease

Unlike for the case of rare diseases, there is no clear definition or statement by governments, states or the EU defining a common disease. No exact numbers are given, or definitions established. Thus, it can only be concluded from the definition of rare diseases, that a common one’s prevalence must be much higher than a rare disease’s. Moreover, sufficient drugs and (long-term) studies must have been conducted in order to (1) ensure the recovery of costs for drugs in the US and (2) to affect at least more than one in 2,000 people in the EU. This definition is not confirmed or all-encompassing but concluded from what applies to the definition rare diseases.

2.3 Salient and non-salient disease

The definitions of salient and non-salient disease are as hard to determine as the one for common diseases. In consequence, a distinction solely between the four diseases covered within this work will be done. This leads to a mere differentiation within the frame of this work, but to no general definition applicable outside of this research. For classifying all four diseases on a range from known to unknown disease, the sum of Facebook likes for the three most popular Facebook pages was calculated. Within this classification, only Facebook pages run in English and containing the English name of the disease were considered. For Behçet’s Disease, the number of likes for the three most popular Facebook pages reaches the total amount of 8,334 likes. Endometriosis, the second least known disease among these four, aggregates to a number of 135,778 likes. Now approaching the rather known diseases, the third in row is AIDS/HIV with 1,483,785 likes in. The best-known disease within this research is ALS with 1,704,173 likes in sum (Appendix 3). Considering this wide range of number of likes, a clear tendency from unknown to known diseases can be drawn. After having set a frame of what to consider a rare, common, known and unkown diseases, the four diseases used within this research will be defined and their media attention over time will be sketched.

(11)

3. Definition of AIDS/HIV, ALS, Endometriosis and Behçet’s disease

A short definition and historical background of each disease respectively will be given in the following paragraphs.

3.1 AIDS/HIV

Historically, HIV crossed from chimpanzees to humans in the 1920’s through consummation of the apes. A so-called SIV (Simian Immunodeficiency Virus) was identified among

chimpanzees that turned out to be almost identical to HIV (Origin of HIV & AIDS n. pag.). Although the transmission of the virus from apes to humans already took place so early, records of people infected only began in the 1970s (History of HIV and AIDS Overview n. pag.). By 1980 scientists assumed that between 100,000 and 300,000 people distributed over five continents suffered from HIV. Until 2013, the number of patients rose to an estimated 35 million people living with HIV. In 2015, The number of people living with HIV is estimated to 1,722,900 in the United States (HIV in the United States: At A Glance n.pag.). For this reason, it can be classified as a common disease with far more than 200,000 people infected.

HIV, in a medical sense, is a virus that attacks the human immune system. It destroys the T-helper cells and by doing so also multiplies itself. Thus, it spreads throughout the whole body. Without treatment, it takes about 10 to 15 years until the immune system is no longer able to defend itself. While HIV stands for human immunodeficiency virus, AIDS is the final outcome of this virus. It stands for acquired immune deficiency syndrome and consists of a set of symptoms as well as the immune system’s inability to defend itself and function properly. Hence, it is also considered to be the last stage of HIV (What Are HIV and AIDS? n. pag.). HIV has been known since the 1980s, although some cases were documented in the 1970s (History of HIV and AIDS Overview n. pag.; Kimball et al. 254). In 2016, approximately 36.7 million people were infected with HIV (Global HIV and AIDS Statistics n. pag.). In the US, currently more than 1.1 million people are living with HIV, whilst 1 in 7 do not know about it (U.S. Statistics n.pag.). Thus, AIDS and HIV can be considered to be a common disease in the United States with more than 200,000 people infected.

3.2 Amyotrophic Lateral Sclerosis

ALS, short for Amyotrophic Lateral Sclerosis, was identified as a particular disease by Jean Martin Charcot in the late 1860s. This disease is more commonly known as Charcot’s disease besides its most common name, ALS or Lou Gehrig’s disease.

(12)

Although the disease has been known for over a century (Amyotrophe Lateralsklerose (ALS) n.pag.), the origin of the name Lou Gehrig’s disease came into play only later, in 1939. He was a famous baseball player for The New York Yankees when he stopped playing during a game, due to not feeling well physically. Only few months after that he was diagnosed with ALS and died two years later. Since then, ALS is more commonly referred to as the Lou Gehrig’s disease (Lou Gehrig and the History of ALS n. pag.).

ALS is an incurable neurodegenerative disorder, but treatable with limits at the moment (What Is ALS n. pag.). Only two treatments are approved in the US so far (What Is ALS n. pag.), but the well-known Ice Bucket Challenge in 2014 fueled the attention for this disease and gathered more than 115 Million US-Dollars of which 89 Million US-Dollars were used for research only (Impact of ALS Ice Bucket Challenge n. pag.). Medically-speaking, ALS’ symptoms are muscle cramps and twitching, weakness in all extremities, difficulties while speaking and swallowing and finally the paralysis of muscles of respiration (What Is ALS n. pag.). The number of people suffering from ALS is estimated up to 450,000 people worldwide (A Life Story Foundation), and more than 20,000 United States citizens at any given time (About ALS n.pag.), due to the fact that it takes 18 months on average to diagnose a person after the first symptoms appeared.

3.3 Endometriosis

The German physician Daniel Shroen described the symptoms which commonly occur in Endometriosis in 1690 (Sutton 3; Gupta et al. 3), but it took another two centuries until the disease was recognized as such. Current estimates believe 7 million women in the United States to be effected by Endometriosis (Gupta et al. 3). The disease was initially diagnosed and discovered by Carl Freiherr von Rokitansky in 1860 (Batt 22; Gupta et al. 3). 25 years later, in 1885, Von Recklinghausen was the first one to name this condition “Endometriosis” (Sutton 5).

From a medical point of view, Endometriosis is a gynecological disease where endometrial tissue occurs outside of the uterus and within the abdomen. According to Rogers et al., Endometriosis affects 10 % of women in the reproductive age (336). Symptoms vary but most common are pelvic pain, long, painful and strong menstrual bleeding, painful urination, painful emptying the bowels, gastrointestinal discomfort and decreased libido (Gupta et al. 3). Another common symptom is infertility. Due to the fact that not all symptoms necessarily need to be present or include others that are not commonly known, the average time until confirmed diagnose takes 7 to 8 years (Gupta et al. 3).

(13)

Besides Endometriosis affecting women themselves, various studies have shown that people concerned show substantial loss in work productivity and reduced quality of life (Nnoaham et al. 370; Reilly et al. 362). Thus, an economic as well as societal burden can be expected among all sufferers of Endometriosis.

3.4 Behçet’s disease

Originally, Behçet’s disease was reported by a Turkish dermatologist in 1937. In modern times it is considered to be an inflammatory disease. Except for inflamed eyes, which can cause blindness, all other symptoms are episodic and regress by themselves. Besides the name Behçet’s disease, this disorder is also called “Silk Road disease” due to the fact, that a lot of patients reside along the ancient Silk Road (Ishigatsubo 2). Thus, a positive family history can be implied within the occurrence of this disease. Although 420 in 100,000 people suffer from Behçet’s disease in northern Turkey, the disorder is considered to be rare due to the low frequency among northern Europe and the US (fewer than 1 in 100,000 people) (Behçet Disease n. pag.; Calamia et al. 602). The reason why Behçet’s disease was chosen as a representative for a rare and unknown disease is, because a chronic, and incurable disease was needed in order to fit into the general common criteria chosen for the other three diseases investigated within this research.

4. Theoretical Framework

The theoretical framework for this thesis will now be given, in order to later support and explain the findings made through the generated datasets. Therefore, web 2.0 and its affordances, the Agenda-setting theory in the context of new media and the opinion leader theory will be covered.

4.1 Web 2.0

In order to set a frame in which the following Agenda-setting and opinion leadership theory will range in, a definition of Web 2.0 follows. The term Web 2.0 was initially coined by Eric Knorr in the CIO magazine in 2003 in which he cited Scott Dietzen, describing the new emergence of an interactive web “a universal, standards-based integration platform” (90). Cormode and Krishnamurthy consider the websites that were highly popular in 2008, such as Facebook and YouTube, as part of the new Web 2.0 (n. pag.). They defined technological, structural and sociological axes to differentiate between Web 1.0 and Web 2.0 (n. pag.).

(14)

Moreover, Web 2.0 has a strong social factor (personal profiles, connectivity afforded by friendships) and stimulates the creation of user-generated content (Cormode and Krishnamurthy, n.pag.; Kilian et al. 4). Although it is referred to Web 2.0 and the “old” web or Web 1.0, there is no all-encompassing definition or categorization of these two types of web. Rather, Web 2.0 evolved from what was different in contrast to Web 1.0 (Cormode and Krishnamurthy n. pag.). Whilst Cormode and Krishnamurthy promote the term Web 2.0 (n.pag.), Kilian et al. consider it to be lacking of a definite definition (Kilian et al. 4).

Additionally, they criticize that the new central elements of Web 2.0, such as interactivity, connectivity and customer integration, delineate in highly distinct applications and services which refutes a common definition (Kilian et al. 5). With Web 2.0 being in full utilization, van Dijck and Nieborg added that the majority of people (52 %) is inactive on the web whilst only 13 % are actual creators of the user-generated content found online (Van Dijck and Nieborg 860). Therefore, the advantage of the participatory Web 2.0 has to be questioned. In relation to this research, it is thus important to keep in mind, that only a small percentage of users represent the actual creators, whilst most remain passive or inactive. Moreover, the collective thought and equal creativity among all users seems to be idealized, although there is evidence that entertainment, career and family rate rank higher as motivation factors than the mere communal effort as a driving force (Van Dijck and Nieborg 862). In the context of this thesis, this implicates that topics usable for entertainment are easier and more often engaged with online, whilst it could be expected that diseases, such as the ones investigated in this research, might not be of interest for the broad mass that wants to be amused by the internet’s content. Of course, it needs to be noted at this point, that the ALS Ice Bucket Challenge for example turned a call for donations into an entertaining social media trend.

To that end, both types of content covered on social media might be intertwined with each other for the sake of utilizing users’ motivation to participate with the request to support social or health campaigns. On the other hand, Van Dijck and Nieborg state that a common interest in topics, products or brands is also another reason for participation. This shows, that people interested in the same disease, be it because they are affected themselves or know friends or relatives affected, are motivated to contribute to the online discussion. A further drive of visiting user-generated content websites is the user’s exposure to “(viral) forms of social media (‘friends’ networks) or by plain marketing mechanisms” (Van Dijck and Nieborg 862). This supports the observation that could be made during the ALS Ice Bucket Challenge.

(15)

Once a famous person triggered the participation in the Ice Bucket Challenge and nominated other people to take part in it, it went viral and enrolled an exponentially increasing number of users over time (Analysing the Ice Bucket Challenge: A Force for Good in Charity Marketing? n.pag.). In the following, the characteristics and affordances of the Web 2.0 will be identified.

There are prototypes, such as MySpace and Facebook, which are clearly considered to be part of Web 2.0. Others, which were founded during the Web 1.0 era, for example, added features and layers over time and thus, managed to establish themselves within Web 2.0 over time. An example used in this context was Amazon, since it was founded in the 1990s but gradually managed to add features such as user reviews, ratings and the implementation of user profiles, although the social aspect of friendships and connectivity to one another is scarcely used. As already mentioned above, the term Web 2.0 was coined in 2004, whilst the first prototypes of websites belonging to it only emerged shortly before that (late 2003 and early 2004). This implies, that pages created after 2004 and containing features such as user profiles, connectivity among users, providing space for user-generated content and allowing third-parties to enhance the platform or website through an API are considered to be part of Web 2.0 (Cormode and Krishnamurthy n. pag.). Wherefore, without Web 2.0, the phenomena of Agenda-setting and opinion leadership online, which will be defined in the following two chapters, would not have come to existence. This is due to the fact that only when the broad mass began using the Web 2.0 as a means of communication, distribution and connectivity, such theories expanded onto the web and its affordances.

4.2 Agenda-setting theory

According to Soroka, as well as Dearing and Rogers there are three types agendas: the media agenda, public agenda and policy agenda (Soroka 7; Dearing and Rogers ix). By establishing these three agendas, Dearing and Rogers propose “that Agenda-setting is best understood as a process of interaction among three types of agendas” (Dearing and Rogers ix). Furthermore, Metzger, extending this triplet of agendas, states that the increasing social connectivity resulted in a new type of agenda, the so-called blogosphere (568). She states that this might not necessarily constitute a new form of agenda, but rather a subtype of the public agenda, which, in turn complicates the Agenda-setting processes known so far (Metzger 568). Different opinions occur around the question if the blogosphere is influenced by the traditional media, if they set their own agenda or might even influence the traditional media’s Agenda-setting (B. Lee et al. 69; Delwiche, n.pag.). This is central to the research question addressed within this

(16)

If they are, it will be crucial to identify them and in which direction the spheres influence each other. Therefore, this concept will be elaborated on further.

The original notion of Agenda-setting theory covers the perception of media in a way that it does not decide what people should think, but rather what to think about (Cohen 13). Shortly after that, McCombs and Shaw ran a study on Chapel Hill voters who were asked what they think were the key concepts of the presidential campaign in 1968.

In a second step, they compared these statements with the actual content dealt with during the presidential campaign (McCombs and Shaw 177). What could be seen from the initial study on Agenda-setting was that the political status is reflected imperfectly by the individual who rather uses a composite of media but barely consumes all media and all news available (185). This means, that every individual interviewee gave a different perspective on how the politicians positioned themselves during the campaign, due to different sources of information consulted. Further, McCombs and Shaw identified the salience of affect which took place especially among people highly interested in politics (186). This means, that people feeling affectionate towards a certain topic, person or object covered by the media, block their openness towards new information and facts about this topic. This phenomenon was also observed among people barely interested in politics, but primarily among those highly interested in it (187). For the interpretation of the data relevant to this research, this implies a certain reticence towards topics which people already have an opinion about. With these phenomena in mind, McCombs and Shaw also discovered that recipients recognize the importance by the amount of media coverage of a certain topic (176). In practice, this means that recipients consider a topic, which is covered thoroughly by the media, to be important and thus, might raise their interest in the topic as well. In the context of this research, this implies that once the media covered one of the diseases thoroughly, an increased interest should be identifiable.

Today the Agenda-setting theory also applies to developments like new, online and social media. This also resulted in a critical engagement with the notion of power of the media within Agenda-setting. Several researchers have developed different notions of Agenda-setting, such as inter-media and reversed Agenda-setting, and how it behaves in contact with social media and big data.

(17)

This extensive research for over more than three decades resulted in an identification of certain factors that influence Agenda-setting. For example, inter-media Agenda-setting is one of these factors. Inter-media Agenda-setting refers to the influence mass media agendas have on each other (Lopez-Escobar et al.225; Golan 326). Lee et al. for instance, identified an inter-media influence between traditional media and blogs (B. Lee et al. 68). They “examined the inter-media relationship between newspapers and Internet bulletin boards” (Lee et al. 68) in 2000 during the general election in South Korea. Both the first and second level of Agenda-setting were examined. Second level Agenda-setting in this sense means, that the media, on the one hand, have the power to decide on topics and information covered in the media. On the other hand, they also manage to set emphases within the news agenda and thus, influence what is considered to be important (Weaver et al. 191). Their results demonstrated on the first level of Agenda-setting, that the newspaper influenced peoples’ opinions who read the Internet bulletin board. Further, Lee et al. found out that on the second level of Agenda-setting, the Internet bulletin boards influence the newspaper coverage (Lee et al. 68).

This implies, that the Internet serves as a “source that influences the traditional news media in terms of intermedia Agenda-setting” (Lee et al. 68). Delwiche though discovered, that the blog agenda does not necessarily coincide with the agenda of traditional media (Delwiche, n.pag.). Within this context, Metzger identified, at least in some instances, an inversion of media-to-public Agenda-setting taking place (568). This means, that the media-to-public agenda gets to set the agenda for the mainstream, instead of traditional media setting the agenda (Metzger 568). This finding also matches with the reverse Agenda-setting defined later on. Accordingly, Chaffee and Metzger found out that “the interactive or two-way communication capacity and overall increased information flow associated with new technologies may give more power to people whose agendas would not normally be reported in the major mass media” (375).

For this research, this implies a shift from the traditional news media towards the individuals and collectives outside of the traditional media institutions, who are able to affect the mainstream media’s agenda. Moreover, they state that, due to this development, recipients are not only able to set their own agendas but to “influence others’ issue agendas by helping them locate and contact people who care about similar issues” (Chaffee and Metzger 375). Thus, a certain inter-personal Agenda-setting can be identified among the newest developments due to the connectedness and communities afforded by the Internet. Needless to say, this also applies

(18)

The connectedness and ability for users to create their own content and their own agenda, enables the individual users to influence the mass media in terms of what they want to be informed about.

Shoemaker and Reese concentrated on the five-tier conceptual model which describes five factors that influence the media agenda (Shoemaker and Reese 8). Rather than talking about how media institutions themselves influence each other (Lopez-Escobar et al.; Golan; B. Lee et al. 68), Shoemaker and Reese talk about factors like “journalists, media routines, organizational factors, social institutions, and cultural or ideological considerations” (Russell Neuman et al. 195) and call this the hierarchical model (Shoemaker and Reese 9). Therefore, the latter authors mentioned factors applicable on the internal, organizational level of media institutions whilst the formerly mentioned authors focus on the interdependence of media institutions themselves. Another important point to mention is, that Russell Neuman et al. discovered in their study that the public agenda in the way it is reflected in social media is not taking up a fixed position in relation to the news agenda which is provided by the traditional news media (210). Meaning, that the social media talk comparatively more about social or public order issues than the legacy media. In the following, reverse Agenda-setting and the interdependence of traditional and new media will be covered.

According to Kim and Lee, who reintroduced the term reverse Agenda-setting, they generated a new discussion around the notion or idea of setting. Before that, reverse Agenda-setting meant the mere phenomenon of journalists reacting to perceived or real public interest and thus resulting in the public agenda to be influential on the media agenda (McCombs 132). With Kim and Lee reintroducing reverse Agenda-setting, they try to broaden the notion and emphasize the growing importance of online and social media and their influence on the legacy media agenda (Kim and Lee 183).

In the course of their research, they introduced a so-called second-level Agenda-setting. As already mentioned above, this means that the media do not only have the power to set topics and issues covered by the news agenda, but can also set certain foci and emphasize certain attributes more than others (Weaver et al. 191). It is important to note here, that second-level Agenda-setting is not the same as framing. Framing encompasses not only the foci and emphases the media set but also how users interpret and consume the way news are conveyed (e.g. word choice, structure of the article, device used etc.) (Lecheler and de Vreese 186).

(19)

For the recipient, second-level Agenda-setting eases the process of classifying the information just received (Weaver et al. 191). This is because recipients then only need to rely on what the media emphasize rather than executing their own classification among the news they read.

While talking about Agenda-setting and the so-called second-level Agenda-setting (or: framing), one must notice that, at this point, the notion of opinion leadership starts to tie into what has been said before. The next chapter covers the notion of opinion leadership and thus, will not be covered thoroughly at this point. Without giving away too much of the following paragraph, it can be said that through opinion leaders (or the mass media) certain spotlights can be set for the recipients in order to influence the light in which certain topics are casted. Thus, a clear distinction is no longer possible and the borders between these two theories become blurry. Chong, for example, states that framing is “the essence of public opinion formation” (870). Withal, he states that public opinion leaders generate common frames for people to eliminate, filter and further inform themselves on topics they are interested in (Chong 870). Russell Neumann et al., at this point, found a certain independence to be evident between social and traditional media in the context of issue framing (210). The significance is that although they might influence one another (inter-media Agenda-setting), their topics covered and issues framed are relatively independent from traditional media, journalism or certain spokespersons (Russell Neuman et al. 211).

In the context of diseases in particular, Merscheim discovered already in 1978 and 1984 that the media tend to cover severe diseases, such as cancer, disproportionately (Merscheim 107; Merscheim 93). Moreover, according to Baumann et al., a coverage of well-known diseases in contrast to those which are in need of explanation is common among the media. Implying, although a well-known disease might not be as prevalent as others (such as anorexia), it is more likely to be covered by the media than diseases about which is known only little (Baumann et al. 442). Although the notion of Agenda-setting originated within political studies, several other research projects showed that Agenda-setting is also present in medical agendas, such as skin and breast cancer, only to name a few (Dixon et al. 178; Ogata Jones et al. 107). For this research, this means that an Agenda-setting process can be expected for other medical conditions alike. Overall, medical journalism follows the same rules as journalism in general.

(20)

Everything that covers so-called news factors, such as newness, surprise, negative headlines, geographical and social closeness to the reader, dismay, prominence, and emotionalization, is also interesting for medical journalism (Rossmann 304). This does not only influence what topics are covered but also how news are conveyed. This results in prominent actors or sensational stories to overshadow and warp a disease’s state of affairs (Rossmann 304).

Particularly in the case of AIDS, an event-related coverage of this disease could be observed. From the 1980s onwards the media coverage about AIDS continually increased up to the point where in 1987 two German newspapers, FAZ and die Welt, dedicated at least one article to AIDS in more than every second edition (Boes 188). Another important event that caused another increase of media attention was the death of the actor Rock Hudson due to AIDS in 1985. Again, this resulted in an immense increase of articles about AIDS shortly after his death, even if case reports about his death were spared (Maurer and Reinemann 207).

In the case of Endometriosis, the exact opposite could be observed. According to 91 women asked in a survey, all suffering from Endometriosis, the media do not cover Endometriosis sufficiently (Steinberger 303). Moreover, when these women were asked what reasons were responsible for this state of affairs, 45 % believed that Endometriosis appears to be too unspectacular and unknown, so that important gatekeepers might not classify Endometriosis as newsworthy enough. Another 25 women believe, that the reason for this is that the disease affects women only and another 23 think, that period pains are considered to be normal among society and thus, do not demand any media coverage. Another argument that 12 women mentioned, was the non-existing lobby for media coverage and the fact that those affected barely go to the public with their personal stories and experiences (Steinberger 303). At this point, it must be noted that these surveys only represent subjective estimations about media coverage of Endometriosis. This concludes, that no objective source could be found which investigated the coverage of Endometriosis in the media. Nonetheless, the only available source was used in order to draw at least a subjective estimation on how Endometriosis is covered in the media.

In comparison to Endometriosis and AIDS, ALS only received great online and media attention once through the ALS Ice Bucket Challenge which took place in 2014 (1. Introduction 1; Xu 210).

(21)

In combination with other social media phenomena, Xu concludes from these observations, that nowadays journalists are obliged to connect and interact within online social communities in order to know what happens and to “re-establish themselves as trusted information providers” (210). Literature dealing with the media coverage of the Behçet’s disease could not be found. This supports the assumption that the Behçet’s disease is barely known and thus, is scarcely mentioned in the media. Nonetheless, “a direct influence of the mass media over Twitter [was] not observed, but rather a two-step information process” (Aruguete 51) could be identified, such as introduced by Lazarsfeld. The two-step flow of information will be defined in the following paragraph.

McCombs and Coleman conclude, the rise of the Internet has not eliminated the Agenda-setting influence completely (504). This is because most news distribution channels, particularly online news sites, show high redundancy in what topics are covered in the agenda (McCombs 91). For the theory of Agenda-setting in the context of new media, this implies the possibility that no great change in the conception of Agenda-setting needs to be introduced (Coleman and McCombs 505). Of course, it cannot be said precisely in how far the issue agendas of old and new media differ, but as McCombs and Coleman inferred, non-traditional media are as important to the younger generation as traditional media are to the older generation (Coleman and McCombs 505). In practice this means, that due to the affordances of web 2.0 an almost instant distribution and circulation of content has been enabled. Moreover, interactivity, interconnectivity and the dynamic role of users also becoming producers supports the discontinued factor of time lag among Agenda-setting influences (Aruguete 39).

4.3 Opinion leader theory

Just as in the case for Agenda-setting, the concept of opinion leadership was initially investigated during a political campaign in the US. Lazarsfeld et al. asked themselves, what factors influenced the voters during the political campaign of 1940 (1). They investigated the influence of the press, radio and personal contacts among other factors on the voter’s choice (5). Those who were most engaged with the election campaign were identified as so-called opinion leaders (49). These were identified “by asking people to whom they turn for advice on the issue at hand” (49) and then investigated the communication between adviser and the one who sought advice. It seemed that opinion leaders rather relied on media than on personal relationships when informing themselves about a certain topic (151). As an outcome of this research, Lazarsfeld et al. introduced the two-step flow of communications (151).

(22)

Accordingly, this model implies “that ideas often flow from radio and print to the opinion leaders and from them to the less active sections of the population” (151). The reason for that is that the influence of legacy media was expected to be relatively high, whilst it could be proven that the factors most influential on less active voters were friends and family (51). The following figure clarifies what is meant by the two-step flow of communications (fig. 2).

Figure 2: Two-step flow of communications (Katz and Lazarsfeld 183)

Therefore, the opinion leaders function as a mediator between less interested people and the mass media (Lazarsfeld et al. 151). At this point, it is essential to say that this research does not investigate the people influenced, but rather opinion leaders and if, for example, Facebook pages may function as such. It can thus be said that the focus of this theory in context of this work lies in the relationship between mass media and opinion leaders in case such a relationship can be observed or identified at all.

In order to establish a touching point with the methods used in the following, I want to refer to a suggested method for measuring opinion leadership by Coleman et al. (Coleman et al. 114). In the 1960s, Coleman et al. asked doctors to name three or four other doctors who they maintain most social contact with, name three or four other doctors who they share information about medication with, and who they would contact in case they needed advice on a certain drug (114). With this information, they created a so-called sociogram.

(23)

Although it cannot be compared one-to-one with the page like network identified through Netvizz, there are similarities among the construction, interpretation, and usage of both visualizing means for interdependencies of a certain network.

Another touching point of opinion leadership and the methods applied in this research is the measurement of opinion leadership among social media. The rapid growth of social media and the notion of produsage, through which users become producers of content themselves in various spheres of the web (Bruns 2), has led to an increasing number of online social networks that serve as “dominant channel[s] of information exchange” (Zhang et al. 12). This means, that users generate their own content through blogs for example, whilst sharing already existing content simultaneously (Bruns 2). In that sense, this ties together the notion of Web 2.0 which was elaborated on above.

Within the theory of opinion leadership in the sphere of social media, Zhang et al. identified the crowd to occupy a decisive role within trend setting (12). Moreover, according to Zhang et al., opinion leaders are able to raise local awareness, but only the broad mass manages to create an equally broad reaction (Zhang et al. 12), also observable during the ALS Ice Bucket Challenge which took place in 2014.

In order to not only focus on the historical roots of the theory of opinion leadership, a current view on the developments of opinion leadership within social and new media will be established. Therefore, it is important to note at this point that Merton stated in 1949, that opinion leaders vary in volatility in the scope of issues they are highly informed and knowledgeable about (189). This means, that opinion leaders might hold leadership for one topic only (monomorphic) or various topics (polymorphic) (Merton 213). By the same token, their scope differs from locals to cosmopolitans, meaning that their scope of influence varies in size and topic relevance to either the large society or local issues (Merton 189).

Applying this observation onto the new media environment, it is evident that this effect amplified through social media particularly. In the context of new media, Schäfer and Taddicken identified four communicative roles of opinion leadership. The fourth represents a new role which has only come into play due to new media (973; fig. 3).

(24)

The so-called mediatized opinion leaders give advice whilst using all channels of communications that had been present before the rise of new media but also rely heavily on mass media in order to retrieve necessary information (fig. 3). Looking at mediatized and former opinion leaders, it becomes evident that both share a small, homogenous group of similar opinions shared within them. Albeit the difference between mediatized and non-mediatized opinion leaders is that the former possess a large circle in which they share their ideas (973; fig. 3).

Figure 3: "Overview of Communicative Roles of Opinion Leadership in New Media Environments" (Schäfer and Taddicken 973)

In the context of this research, this means that opinion leaders still range in spheres where they find similar opinions but manage to reach an increased number of people due to social media. Therefore, this could indicate that experts or the initiators of the Facebook pages in the field of each disease might only manage to become salient within their sphere of interest but not manage to step outside of this circle. Consequently, a quite homogenous picture of other Facebook pages within the networks could be expected. Speaking in terms of social networks, which will be investigated through page-like networks generated by Netvizz, it can be said that “social network centrality is positively correlated with both opinion leadership and consumer susceptibility” (S. H. Lee et al. 72).

(25)

By the same token, nodes high in betweenness centrality are more influential over others but are prone to be influenced by others (S. H. Lee et al. 72; Mutschke 264). They state, that only in-degree centrality positively affected the nodes’ influence (S. H. Lee et al. 72). In the context of Facebook networks investigated within this research, it is questionable if this is correct. This is due to the fact, that the networks investigated show the post activity as a separate criterion and thus, it needs to be proven if degree count and post activity correlate with each other.

Despite all that, for this research, this means that nodes high in betweenness centrality and a high degree of in-links are the ones important to the respective discourse on Facebook. Another operationalization S. H. Lee et al. investigated was the out-degree centrality and betweenness centrality functioning as proneness to influence (75). Thus, out-degree counts indicate how likely a node is influenced, whilst the number of in-degree counts is an indicator for functioning as an influencer (S. H. Lee et al. 75). In 1999, Burt already observed that opinion leaders actually function as opinion brokers, since they “transmit information across the social boundaries between status groups” (47). Further, they are strongly connected to other groups and thus, manage to stand out compared to those opinion leaders only active within their own network (48). Since the two-step flow of communications has already been introduced, it will now be applied onto Burt’s understanding of it by introducing his terminology in the context of networks. Burt calls the strength of a relationship between two nodes or networks “cohesion” (39) and their similarity of relationships “equivalence” (39). Together, they encourage contagion of topics, opinions, and trends (Burt 39). If contagion occurs through cohesion, it is due to socializing communication. In this sense, it means that it occurs through mutual likes between nodes in the investigated network. If it occurs due to equivalence, it is through competition for mutual nodes’ opinions within their common network (39). Thus, Burt considers the well-known two-step flow of communications a composition of contagion by cohesion through opinion leaders (that way information flows into the network) and contagion by equivalence which “triggers adoptions within” (47) networks. Opinion leadership, in that sense, describes influence performed via strong edges between barely equivalent (similar) people (Burt 46). For this research, this implies that opinion leaders serve as transmitters between networks rather than within them. Moreover, they incent discourse about certain topics between weakly equivalent people (Burt 46), but nonetheless remain within their sphere of interests and topics and thus, are monomorphic (Merton 213).

(26)

After having established the theoretical framework for this research, in the following the methods will be described and explained in terms of how they help answering the research question(s) of this thesis.

5. Method

Among all methods, this research focused on the English language and the English terms for the diseases respectively. This decision was founded on the notion of the inner English-speaking circle according to Kachru (356) as well as the expectation that United States citizens have a growing interest in medical topics dealt with in the media (Anderson and Horvath 263; Kennedy and Funk 2). Thus, all websites, Facebook pages or accounts taken into account within this research are written and maintained in English and within the United States. A consideration of all websites, Facebook pages and accounts in other languages would have gone beyond the scope of this research. Additionally, it makes more sense to investigate different languages and regions separately due to cultural and societal discrepancies.

5.1 Google trends

Google trends gives back the number of queries for a certain term in a distinct time frame and either worldwide or for a selected country. Thus, it helps pinpointing certain events that took place online and therefore, resulted in an increased number of search queries for each disease respectively. For Google trends, the English term for each disease was chosen to be searched from January 1 2016 up to January 1 2018. Then, the data was downloaded for the United States and each disease respectively, resulting in four data sets. From this point, peaks along the timeline from 2016 to 2018 can be inferred to important dates or viral media occurrences. This indicates a resonance from actual events and online trends towards the usage of Google, in order to inform oneself about the media coverage of such phenomena. In combination with the following methods, this tool helps to identify key dates through which other important viral trends among Facebook users or the legacy media can be derived.

5.2 Netvizz tool

With the Netvizz tool, only the momentary state of a Facebook page can be given. This tool is considered to be a data extractor for the Facebook platform (Rieder n. pag.). The tool is provided by the Digital Methods Initiative of the University of Amsterdam (Rogers 7). Facebook was chosen to be investigated as one perspective of social media, because the majority of Americans uses Facebook most often according to a study conducted in 2018 (Smith and Anderson 1). For this research, the three most popular public Facebook pages were identified through the number of followers.

(27)

The English terms for each disease respectively were used as search terms on the Netvizz search function, resulting in one search query for “Endometriosis”, “Amyotrophic Lateral Sclerosis”, “AIDS” and “HIV” separately and “Behçet’s disease”.

From then on, the type of Facebook site that should be crawled was set on “pages” since they are publicly accessible and show number of likes and connected websites openly. Again, within the selection of Facebook pages, only those located and run in or from the United States were taken into account.

Therefore, the “about” page or their referring webpage were scanned for the location. In case of the Endometriosis Awareness page, it was ensured that they are located in the US by asking them directly where exactly they run their Facebook page from. The Netvizz tool was then used to gather the page-like network of the three most popular public Facebook pages, their page data from every disease respectively. Therefore, the page ID was identified via http://lookup-id.com/. Then, the page ID was put into the search bar of Netvizz for the page like network and the depth was set on 1. This means, that the Facebook page’s depth of the like network is shown to the extent of all pages the initial Facebook page liked and all likes that the initial page received back (Rieder n. pag.). This was repeated for all 12 Facebook pages selected respectively (three per disease). The page data was scraped for a time frame of two years, namely from January 1 2016 to January 1 2018, in order to coincide with the Google trends’ and New York Times API’s data chronologically.

A short overview of the findings that could be made and the reason for choosing the respective method for this research will be given. With the resulting page like networks, one can identify, if the three selected pages are either directly connected or via a third page which they both like. Moreover, network centrality and connectivity were identified. Another interesting finding could be to see what pages the initial page liked, in order to pinpoint their position within the network. By liking other pages, the initial Facebook page positions itself within a network of actors that might be either homogenous or scattered from governmental institutions, over celebrities towards CEOs, companies or Nonprofit Organizations. From the positioning within this network, it can be concluded if the initial Facebook page tries to raise awareness for the respective disease rather through fellow disease pages or through connecting with pages normally positioned outside of the context of diseases. Moreover, their Betweenness Centrality,

(28)

Proceeding from these findings, it could be interesting to look at the strength of the distinct connections, meaning to identify strong or weak interactions between the actors.

Eventually, the page data for each Facebook page respectively was scraped for the given time frame of January 1 2016 to January 1 2018. Results most interesting among those datasets are the number of posts published over time, their engagement, shares and reactions. Through this, similarities and differences of activeness on social media can be observed in contrast to The New York Time’s coverage of the disease and Google trends. Thus, this method is particularly interesting in comparison to other methods used within this research.

5.3 Gephi

Gephi is a visualization tool for humanities. For this research project, it was used to visualize the page like network of the three most liked Facebook pages identified for each disease. Therefore, the dataset scraped via Netvizz was loaded into Gephi. The algorithm Force Atlas 2 was used on every visualization created. The particular settings that were used for each figure shown in the following course of this thesis, can be found in the appendix (appendix 3).

5.4 Newspaper attention

Via The New York Times API, an exhaustive collection of articles, which mentioned one of the four diseases respectively, could be compiled. With The New York Times API, it is possible to query The New York Times newspaper database. The API returns a snippet is provided so that a short overview of the article’s content is given. This helps the researcher to identify the scope and importance of the article queried. For this research, the time frame was set from January 1 2016 to January 1 2018 in the code. Then, each disease was crawled with the following search terms: “AIDS” and “HIV” separately, “Endometriosis”, “A.L.S.”, and “Behçet’s disease”, since this was the common spelling in The New York Times. From the snippet preview, it is possible to deduce the broad context of the article as well as the date on which it was published. The New York Times here, functions substitutionary for the traditional and legacy media in the United States. Although, depending on the criteria focused on, either USA Today or the Wall Street Journal are considered to be the biggest North American newspapers, The New York Times still remains the most influential one (Hanson 432).

For this research, it is more important to look at the most influential institutions rather than the biggest ones, in order to identify indicators of media attention.

(29)

This data helps to identify the extent and peak dates on which The New York Times reports about the diseases and in which context. Proceeding from this, assumptions can be made in connection with the data compiled through the other methods. Combining these datasets with the findings made on Facebook or Google Trends, I expected the agenda covered to be similar among both new and legacy media. This means, that in case a peak is observable in the Google trends dataset, a higher number of articles by The New York Times compared to other moments in time should be expected. Of course, this should also apply the other way around, according to the setting theory introduced above. Since it is not yet clear if an intermedia Agenda-setting exists or not, this is a mere assumption based on Lee et al.’s research (68). In order to retrieve data from The New York Times, The New York Times API was used. This is a Developer key which allows access through a Python script. With this script, all articles that mention the queried keyword were retrieved. From the look at the abstract, the date and time could be derived. The Python script can be found in the appendix (appendix 3).

6. Results

Proceeding from what has been said so far, the results will now be presented. Each disease will be examined individually from various angles. Since some data only becomes interesting in comparison to each other, certain groups of data will be compared for every disease respectively. These combinations occur within the page-like network and page data results.

In order to identify the most central and important nodes within one network, specifically those nodes most central (Betweenness Centrality and In-degree count), level of publicity (fan count), post activity (activeness on Facebook) and categories (sphere of interest) were taken into account. The networks consist of those pages which either like or are liked by the initial, investigated Facebook page, resulting in a page-like network generated for every disease’s top 3 Facebook pages in terms of fan count. The following criteria were chosen to be compared:

• In-degree and categories • Fan count and categories

• Betweenness Centrality and categories (with respect to every network)

For identifying the most prevalent spheres of interest in the network, the following criteria were compared:

• Category ranking according to prevalence

(30)

These groups ensure that all relevant facts and figures are extracted from the data retrieved and can later, during the discussion, put into context and a bigger picture. A detailed analysis of all data retrieved for this research, can be found in the appendix (appendix 3). All data presented in the following, were selected due to their importance for answering the research question.

6.1 Results for AIDS/HIV

6.1.1 AIDS/HIV Google Trends results

Beginning with the AIDS/HIV dataset, the result most striking is that there are peaks on the Google Trends on dates that reveal to be important for society, such as remembrance days or calls-for-action (Figure 4). The following results for Google Trends for each disease will be represented in the same manner1_.

Figure 4: Google Trends HIV/AIDS 01.01.2016-01.01.2018

Looking at this figure, the peaks observed among HIV and AIDS search queries within Google, took place on April 24 2016 (100), November 26 and 27 2016 and 2017 and December 25 2016. The lowest point took place on August 13 2017, showing a value of 28. What is interesting when looking at these peaks, it becomes evident that important dates, for example the World AIDS Day, cause awareness and therefore, queries increase.

1 _{The x-axis represents the time frame covered for this Google Trends query. It is divided into monthly} sections from January 3 2016 onwards up until December 3 2017. It is important to note, that Google normalizes the results by a maximum of 100 and derives all other queries from this number. The absolute number of queries cannot be identified, but rather their ratio. The blue line represents the weekly number of queries whilst the orange one shows the monthly average. Only the highest and lowest points show labels, in order to emphasize those numbers important for this research.

(31)

This day takes place December 1 and is celebrated yearly (About World AIDS Day n.pag.). Along the week before that awareness campaigns begin to take place (Clark n.pag.) which are visible on Google trends.

6.1.2 AIDS/HIV Page-like network results

Observations that are striking among all three page-like networks (with their three dominant categories emphasized), is that all three contain the category Nonprofit Organization among them. This result is supported by the fact that Nonprofit Organizations rank highest among category ranking according to the quantity of nodes for all three page-like networks (Table 1). These observations are important to point out in terms of the research question, since the pages’ spheres of interests will become interesting when looking at their ability to attain salience among the media.

Category ranking

AIDS Healthcare Foundation

Percentage Act Against

AIDS Percentage Than AIDS Greater Percentage

No. 1 Nonprofit Organization 26.4 % Nonprofit Organization 23.95 % Nonprofit Organization 24.69 % No. 2 Media/News 13.6 % Media/News 16.39 % Health 15.62 %

No. 3 Community 8.8 % Health 13.03 % Organization 11.59 %

Table 1: AIDS/HIV Top 3 categories according to number of nodes for each page-like network investigated for Endometriosis

Although Nonprofit Organizations represent the most common category of Facebook pages (Table 2), it is evident that the pages with highest fan count are not Nonprofit Organizations. Rather, Media/News pages occupy the most popular and salient pages among this network (Table 2). This shows, that AIDS/HIV pages might try to gain attention or salience through those who would be able to report on them. Undoubtedly, their true motivation behind liking such pages cannot be concluded merely through these facts and figures. Nonetheless, this finding is interesting since they might seek for media attention by liking some of the biggest Media/News institutions in terms of fan counts.

Medicalization online: How the media stimulate awareness on medical conditions