Developing a typology of Netflix marathon-watchers: We are not the same, we don’t watch the same.
Karla Hernández Zaldívar 11448725
Master’s Thesis
Graduate School of Communication Entertainment Communication Science
Supervisor: Jessica Taylor Piotrowski, Ph.D. January 31st, 2019
Index Abstract 3 Introduction 4 Theoretical Background 5 Binge-watching or Marathon-watching? 5 Who is Marathon-watching? 7
Motivations to marathon-watch: A Uses & Gratifications Perspective 7
The 2.0 Approach 9 Method 10 Design 10 Participants 11 Materials 11 Procedure 12 Analytic Approach 13 Results 13 Netflix Usage 13
Motivations to Watch Netflix 15
Cluster Analysis 16
Clusters and Viewers’ Motivations 18
Discussion 19
Cluster 1 - Marathon Runners 19
Cluster 2 - The Joggers 20
Cluster 3 - Professional Runners 20
Theoretical and Practical Implications 21
Limitations and Further Research 23
References 26
Non-academic resources 30
APPENDIX A - Demographics 32
Table I: Demographics of the respondents (n=372) 32
Table II: List of participants’ countries of residency 32
APPENDIX B - Netflix Use 33
Table I - Watching habits 33
Table II - Content 33
Table III. Most marathon-watched series (top 20) 33
Table IV. Other activities (not listed in survey) participants do while watching Netflix 34
APPENDIX C - Motivations to watch Netflix 35
Table I - Factor analysis for viewers’ motivations 35
APPENDIX D - Cluster Analysis, Netflix usage and ANOVA table Motivations 36
Table I: Netflix use variables and their significance 36
Table II. Cluster Analysis and ANOVA table Netflix viewers’ Motivations 37
Table III. ANOVA table for viewers’ Motivations 37
Abstract
Netflix users are often labeled as marathon-watchers and non marathon-watchers, and this paper provides information to establish that within the marathon-watchers group, there are different subsets of groups, distinguished by how they consume Netflix, and what type of content they consume. This research studies Netflix marathon-watchers from a motivations perspective, under the Uses and gratification theory. The users are segmented by applying a cluster analysis. The relationship between the clusters and their motivations are furthermore explored with ANOVA analyses. This study surveyed 372 international Netflix users 18 or older and provides
comprehensive information about the factors that construct this relatively new habit of marathon-watching. The results indicate that Netflix marathon-watchers are not a homogeneous group and do not behave the same way in terms of usage nor on the type of content they consume; we were able to create three different clusters of watchers. Academic implications about the findings, and some directions for future research are discussed in this paper.
Keywords: Marathon-watching, User Motivations, Cluster Analysis, Netflix, User, New Technologies, Streaming services.
Introduction
Have you ever stayed all weekend watching the Harry Potter movies? Maybe stayed on the couch for more than four hours savouring the adrenaline of a Super Bowl? Bought those movie tickets on a pre-sale because you wanted to be the first to watch that Star Wars movie as soon as you could? When it comes to be entertained, our lives and habits are shaped towards, of course be entertained, but this industry is also a defining part of our culture.
While entertaining ourselves is the most natural goal for consuming media, it is a also true that by doing so, entertainment media nurtures our political ideologies (Chadwick, 2017),
influences our spaces in medicine (George, Rovniak, & Kraschnewski, 2013), changes the way we purchase and how we do marketing (Upreti, Merikivi, Bragge, & Malo, 2017), how we are educated (Cortés, 2000) and diversifies how we spend our leisure time (Jenner, 2016). All in all, it impacts our entire human experience.
We live in a moment in history in which technological progress not only has affected ways of living, working and consuming media. We’ve gone from reading newspapers, to listening to the radio at home, in the car, then on our phone, to watching television changing channels with a remote, to creating our own digital content, to now choosing how and what to watch at our convenience.
One of the key features of the current entertainment scene, specifically the audiovisual industry, is series watching. No longer do we watch one episode a week with the family on a TV, we do not rent a boxset at the local Blockbuster anymore, but we stream an entire series on our
computer, iPad or smartphone, scan through our favourite Black Mirror episodes, or we catch an episode on demand in the car or a plane.
With the arrival of online streaming services such as Hulu, Amazon Prime and Netflix, media consumption habits have changed notoriously, giving the users a new way to access content with no commercials, on-demand and in a portable way, all on the basis of a monthly subscription
fee. As it is easier to access content, watching habits changed with these streaming services and binge watching is facilitated by them. With these streaming services came the possibility to watch any series at any time and place, creating what now constitutes a widespread phenomenon: Binge-watching, also known as marathon-watching. Today we consume entertainment in ways we never did before.
Theoretical Background Binge-watching or Marathon-watching?
In order to capture the activity of the series watching, two concepts have been used in academia as well as in common life: Binge-watching and marathon-watching. In 2013, the term ‘binge-watch’ appeared on the Oxford Dictionaries Word of the Year list, and since, the concept has become a norm of media consumption for many (Boca, 2017). Recently, academic research has come up with a refined definition of binge-watching, namely, “watching more than two consecutive episodes of the same TV show in one go” (Flayelle, Maurage, & Billieux, 2017) with some scholars expanding this to include the number of episodes, frequency, duration and level of engagement (Fillmore & Jude, 2011). On the other hand, popular culture has also tried to deliver a definition for the concept. Harris Interactive conducted a survey for Netflix in 2013 where they defined watching as watching 2/3 episodes of a series in a row (Spangler, 2013). It seems that
binge-watching involves consuming multiple media in a short amount of time, such as binge-watching the entire first season of Homeland in one weekend.
Although binge-watching is often still a colloquial term, we have seen that there has been a desire (much on the part of media companies such as Netflix) to move away from this negative term (think ‘binge-eating’, ‘binge-drinking’) to a more neutral one. Indeed, using the word “binge” to refer to this watching behaviour puts a stigma on it, when there is not enough evidence to warrant such a perspective. We do not, after all, refer to ‘binge listening’, when a person listens their way through an entire singers’ work or to binge-reading when someone finishes a book in two days.
Why then use this phrase when it comes to viewing habits? In response to this concern, we have seen ‘marathon watching’ slowly pervade the industry and academic space. And here, too, we refer to the concept as ‘marathon-watching’, and in line with the previous academic research (Pittman & Sheehan, 2015), define the concept as consuming two or more episodes of a series in the same viewing.
Research in the marathon-watching behaviour has focused mainly either on users’ motives to marathon-watch or to determine if, or to which point, it is a health issue, but none has tried to establish a more comprehensive overview of this new consuming behaviour (Upreti et.al, 2017). And before we can academically establish whether if this behaviour has effects on the viewer and what these are, we primarily need to understand the boundaries and conditions in which this phenomenon manifests and who are the people that behave like that.
For this paper, it was decided that Netflix, a company that began as a DVD rental company in 1997 and has become a worldwide distribution and production company, was the best fit to develop the topic, since it is the world's leading streaming service with 139 million paid
memberships in over 190 countries (Netflix press site, 2019), against the 100+ million members of Amazon Prime (Vanian, 2018) and 20 millions Hulu subscribers (Welch, 2018). Nowadays,
Netflix's original content is created and transmitted all around the world, which becomes relevant for a research of this essence. Therefore this paper will take a step further into filling this gap, and aims to answer:
RQ: Are there different types of Netflix marathon-watchers and which are these different types of Netflix marathon-watchers?
As entertainment shapes and it is shaped by our culture, consuming behaviours play a major role in how we spend our time and how we demand entertainment. This paper will allow us to better understand marathon-watching and who is behaving like this and it will enable us to outline some motivations that lead a person to consume/disregard content in a specific way.
Who is Marathon-watching?
Although consensus about the definition of marathon-watching is still an ongoing matter, academia is currently focusing on the number of episodes of a series in one sitting. On a deeper thought, as there are different types of content to watch, and also different situations, as one person can watch 2 episodes of a sit-com with a duration of 20 minutes each, and another can watch 2 episodes of a drama with a duration of 50 minutes each, probably motivated by different reasons. In the end they would spend different amounts of viewing time, although they both would be
marathon-watching. Who are these people that engage in marathon-watching and what distinguishes from one another?
This paper will review users’ marathon-watching habits, not only divided into marathon and non-marathon-watchers, but rather into different types of marathon-watchers as a heterogeneous group. If we look at this habit in a more comprehensive way in terms of duration, length, the place where they do it, with whom, if they do any other activity at the same time and how often, we might be able to find interesting and new conclusions. Along with demographic information, we will look deeper into the traits that characterise the people that (don’t) marathon-watch and try to define who they are, as it has been proven that some characteristics influence the gratifications sought by media users (Greenberg & Hnilo, 1996). This leads to our first hypothesis:
H1: There are different clusters of Netflix marathon-watchers to be found based on their Netflix usage.
After all, it seems imperative to outline a marathon-watcher typology to learn how to differentiate users from one another, and this paper is here to provide such information, or at least take a step into that direction.
Motivations to marathon-watch: A Uses & Gratifications Perspective
The classical way to consume TV forces the viewer to schedule their viewings upfront, because of the program scheduling, the competition and the industry (Pittman & Sheehan, 2015).
Nowadays, new technologies have enabled users to decide when, where and how to watch their content, changing the ways they select and engage with it (Jenner, 2016). Users now show a proactive orientation towards certain content, actively defining their preferences, by choosing to spend their leisure time watching and also what to watch (Rust & Alpert, 1984). But not only the technology is to blame for people marathon-watching, but the question of why they do it, plays a role to better understand this new habit.
Gratifications have been proven to be a crucial matter to understand media usage behaviour (Tukachinsky & Eyal, 2018) and understanding motivations will help to identify what users to watch. One of the most powerful theories that can help answering these inquiries is the Uses and Gratifications (U&G) theory (Katz et. al, 1973), as humans are believed to be individuals that respond and act according to their own needs and the ones provoked by contextual factors (Lerner, 1987). Its’ basic premise is that individuals seek out media that fulfils their needs and leads to ultimate motivations (Lariscy, Tinkham, & Sweetser, 2011).
This theory is able to provide a perspective that allows media use to be better understood and to dissect the relation between individuals and the context in which they live in, by factoring users’ psyche into the gratifications they seek for consuming media (Palmgreen, Wenner, & Rosengren, 1985). It is also known that motivations to consume media are not universal and vary across media types (Greenberg & Hnilo, 1996), genres (Abelman, 1987) and cultures (Tokinoya, 1996). Palmgreen & Rayburn (1978) recognised that there is also a social factor that plays a part in media use, not as motivation but as a predictor, meaning that external circumstances such as labour hours, family situation and content availability could determine more about the media use than personal motivations.
From a classical perspective on media consumption, users consume it by choosing the medium and content that best satisfies their needs, but they do so under restrictions of the timing and medium (Ruggiero 2000). Opposed to that, this new technological scheme in which viewers
have control over their entire viewing experience, how is it that they feel motivated to view? With all this freedom and choices, what really motivates a person to watch and engage so deeply that they turn their viewing into a marathon-watching behaviour?
The 2.0 Approach
In previous studies about motivations TV has been one of the most researched medium. A paper by Rubin (1981) showed that watchers usually used TV for passing time, companionship, arousal, content, relaxation, information, escape, entertainment and social interaction, while a study by Conway & Rubin (1991) outlined passing time, escape, entertainment, relaxation, information and status as motivations to watch TV.
In previous research, the Uses and Gratification theory has proven to be effective while researching emerging mediums, and to provide insight in the key drivers of marathon-watching behaviours (Sung, Kang, & Lee, 2018). As the Internet emerged, research by Stafford and Gonier (2004) were able to segment general internet users based on the gratifications they seek when browsing online, encountering that users mainly use the Internet as a source of information, communication, and socialising.
Research has also been able to translate the U&G theory to the new era, and has lead to the identification of five predictors of online services use: Permanent access, information,
entertainment, pass time and status (Papacharissi & Rubin, 2000) and online gaming research provides us with an escape motivation (Yee, 2006). Also, Feeney (2014) established that people marathon-watch as a reward (e.g. after a hard day of work) as an experience they plan ahead and look forward to which creates a pleasant experience for them.
Only very recently a more comprehensive light has been shed on this topic to put all the factors together. Pittman & Sheehan (2015) looked at the marathon-watching situation from an U&G perspective by conducting a survey on 262 TV self-proclaimed marathon-watchers, finding three motivations for them: Relaxation, engagement and hedonism. They also found that, for some
watchers, program quality (aesthetics) and the social aspect are the main motivators to do so and that heavy marathon-watchers (defined by them as watching an entire season or a series in one or two days) value all of the above.
In addition, Tukachinsky & Eyal (2018) looked into the personality antecedents and the psychological experiences of viewers relative to the content they consume. Their results revealed that the viewers’ gratifications are enjoyment and involvement.
In a study about motivations to marathon-watch, Steiner & Xu (2018) propose that motives to watch can be used to expand the U&G theory. By conducting semi-structured interviews, they found that viewers have as motives to marathon-watch: Catching up, relaxation, sense of
completion, cultural inclusion, and improved viewing experience.
Since marathon-watching allows the user to watch series straight through, it is only comprehensible to think that viewers get more absorbed in the story by not having to wait for a week or two to find out how the story evolves (Sung, et.al 2018), setting immersion as a possible motivation as well.
All of this leads to our second research hypothesis:
H2: The different clusters of Netflix marathon-watchers will have different motivations to watch Netflix.
By applying this theory within a new perspective and incorporating new online media motivations, this paper strives to provide a more comprehensive insight in the factors that enable marathon-watching.
Method Design
A cross sectional online survey was implemented among Netflix users aged 18 years and older, recruited from personal social media (Facebook and WhatsApp) as a convenience sample.
Using social networks is a good way to step into different groups of people that fit a particular population (Biernacki & Waldorf, 1981).
Participants
Initially 427 individuals responded the survey; after treating missing variables with a list-wise deletion, we retained 372 responses, with a mean age of 35.36 years. Of all the participants, 85.2% were female and 14.8% male. 267 users reported to be married or in committed relationships and 184 to have children. 47.3% work/study (out of home) less than 40 hours per week, while 27.4% work/study more 40 hours, and the rest (25.3%) exactly 40 hours (See appendix A, Table I).
The 372 respondents reside in 30 different countries, in which Mexico (n=121), the United States of America (n=104), The Netherlands (n=62) and Canada (n=23) are the most frequent countries of residence (see appendix A, Table II). From all the respondents, 53.49% said they were the primary holders of the Netflix account they use.
Materials
We developed a self-reported scale to measure the users’ motivations to watch Netflix, based on U&G scale by Palmgreen and Rayburn (1979), which supports that gratifications are good predictors of (recurrent) media use. This research was the first to provide measurements by using the U&G theory for TV consumption. More recently, this scale has been used to investigate different angles of media consumption in other papers (Leung, 2007; Barton, 2009; Whiting, & Williams, 2013).
The survey instrument included three major sections: (1) Demographics, (2) Netflix usage, and (3) Motivations to watch Netflix. The motivations scale had 15 (5-points Likert) scale items.
As a filter question, participants were asked if they owned, or had access to, a Netflix account and had used it in the last three months. A set of demographic questions were included in the survey to observe if these factors play a role in their media consumption.
Given that this study applied the definition for marathon-watching as “watching two or more episodes of the same TV series in one sitting” by Pittman and Sheehan (2015), participants were asked how many episodes of the same series they watch on a regular watching day to identify marathon-watchers. To have a more comprehensive understanding of the habit, we asked the length of the viewing, how many days per week they watch any content on Netflix and how long on average. We also asked the participants questions about how they watch Netflix, when they watch series, and also about their most watched series and genre preferences, as we wanted to know the type of content they consume.
Items for motivations to watch TV were adopted from a Palmgreen and Rayburn (1979) scale for gratifications sought: Entertainment, relaxation, to pass time, learning about things, to forget and companionship. Items on new technologies and their influence on the viewing
experience were included (aesthetic experience, sense of completion and catching up), based on a study by Steiner & Xu (2018), to reflect on how viewing technology itself could be a new
motivator.
For all the motivation items, participants were asked to indicate in a slider how much each reason to watch applied to them, 0 (left) being "It does not apply to me at all" and 5 (right) being "It applies definitely to me”. Descriptives for the measures are provided in the results section and the survey can be consulted in Appendix E.
Procedure
After a pilot was ran with eight Netflix users, the data collection took place in November 2018. Participants were invited to fill the survey via social media and it was administered online, unsupervised and open to users all over the world, no personal data recorded. After reading a brief description of the survey, participants were asked to accept an informed consent form and indicate their voluntary participation in the study. The respondents did not get any reward in exchange for answering the survey.
Analytic Approach
We computed descriptives and frequencies for all the demographics and Netflix usage items, to see how the sample was distributed. To define Netflix watching motivations, we applied a factor analysis to the motivation items. A factor analysis is a statistical method that reduces a number of variables to a smaller set of variables (Costello & Osborne, 2005). The extraction method was a Principal Component Analysis, and a varimax rotation was used. An internal consistency test for the factors was done (Cronbach’s alpha), which indicates the degree to which several items measure the same concept (Gliem & Gliem, 2003).
The Netflix usage variables were used to define if there are different types of (non) marathon-watchers by doing a cluster analysis, a statistical method that allows to group a set of cases into different sub-sets, so these cases end up having significant factors in common. A K-mean approach was used to create the right number of clusters in the most efficient way, as the number of clusters must be determined first. K-means cluster groups are based on a minimum distance from the cases to the centrums of the variables. In principle, when there are several cases with several attributes and the researcher wants to classify the cases based on those attributes, this type of analysis is recommended (Teknomo, K., 2006).
A series of custom tables were created to compare percentages of members in a specific variable, per cluster. Finally, an ANOVA analysis twas done o be able to compare the means of each cluster in the different usage, content and motivations variables.
Results Netflix Usage
On the question of how many episodes of the same series watched in the same sitting, 13.4% (n=50) of the participants only watch one episode per sitting, while 37.1% (n=138) watch two; 31.7% (n=118) indicated that they watch three, and 8.9% (n=33) indicated that they watch
four and exactly the same amount chose five or more. At a glance, we can already separate marathon-watchers (n=322) from non-marathon-watchers (n=50).
In terms of hours spent watching Netflix, 15 respondents (4%) indicated that they watch for less than an hour, 181 (48.7%) one to two hours, 120 (32.3%) two to three, 35 (9.4%) three to four, and 21 (5.6%) indicated that they watch for more than five hours. Regarding the length of the episodes, 209 (56.2%) participants indicated the episodes they watch last less than an hour, 152 (40.9%) one and 10 (2.7%) more than an hour.
Among all participants, 156 (41.9%) were found to watch Netflix five or more days a week, followed by 61 (16.4%) that watch three days, 60 (16.1%) that do it two days, 53 (14.2%) four days, and 42 (11.3%) who only do it once a week. The analysis also revealed that 97% (n=361) of the participants watch Netflix at their own house, while only 2.7% (n=10) do it at someone else’s; one respondent indicated to watch in daily public transport. Moreover, 220 (59.1%) participants indicated that they watch Netflix by themselves, while 152 (40.9%) said they do it with someone else (See appendix B, Table I).
Participants indicated their preferred genres to watch (they were allowed to choose all that apply to them): Comedies 70.2%, Drama 55.6%, Romance 51.3%, Action and Adventure 46%, Sci-Fi and Fantasy 45.7%, Docu-series 38.4%, Thrillers 26.3%, Horror 16.4% and Non-Scripted 12.9%. The recently marathon-watched series by the respondents support these findings, being all comedies and dramas (e.g. House of Cards and House of Flowers, Friends and How I Met Your Mother). For an overview of the series, see appendix B, Tables II and III.
Moreover, when asked if they did any other activities at the same time while watching Netflix, 103 (27.7%) respondents indicated that they do, while 198 (53.2%) chose “Sometimes” and 71 (19.1%) said they did not. From those who indicated that they do other activities, 28.2% (n=105) indicated they second-screen (use computer, cellphone, or other device), 15.9% (n=59) said they cook, while 12.4% (n=46) indicated they chat either in person or by phone, 9.1% (n=34) said they
work/study and 2.2% (n=8) indicated reading. The rest (8.2%) chose “other” as an option (see appendix B, Table IV).
Motivations to Watch Netflix
A factor analysis was conducted to define the motivations to watch Netflix from the items in the survey. As there was only one item measuring Entertainment (Factor loading = .746) and one item measuring Relaxation (Factor loading = .464), those were considered as motivations by
themselves and were not included in the factor analysis. The two items that were expected to load in the variable Pass Time loaded into one factor, plus one more item (“I watch Netflix when there is no one else to talk to or be with”); however, when performing the reliability test (α = .640) it resulted into an unreliable scale. Deleting the third item augmented the reliability to .656, still not enough, so deleting the last two items made theoretical sense to refer to Pass time with only the first item.
When performing the factor analysis, the two items for the motivation Learning about things loaded into one factor as expected. A reliability analysis was conducted resulting on a Cronbach’s alpha of .613 which was not reliable enough, reason why the item with the highest loading into the PCA was kept and the other item was deleted for the motivation of Learning about things.
A factor analysis confirmed that the two items that were expected to load together into one motivation named To forget were indeed measuring the same thing, but unexpectedly, four other items fell in the same factor (“I watch Netflix because I feel satisfied when I finish watching it”, “I watch Netflix because it makes me feel like I have achieved something when I finish it”, “I watch Netflix because it makes me feel less lonely when there is no one else around”, “I watch Netflix to have a beautiful experience for my senses”). As for the reliability, this items showed a Cronbach’s alpha of .803, which means it is a reliable scale. We computed a new variable with the means of all the items, named Get Away, which was what all the items theoretically had in common.
The factor analysis showed that the following items loaded into one factor, although they were not expected to load together: “I watch Netflix to give me things to talk about with other
people”, “I watch Netflix because of the quality of the technology I watch it in (e.g. 4K smart TVs)”, “I watch Netflix because is easy to watch it in my smartphone or tablet”, ”I watch Netflix because I want to keep up with the story” and “I watch Netflix to catch-up with stories I haven't watched for a while or that I left half-way”. The reliability test showed that this was not a reliable scale, as the Cronbach’s alpha was only .706, but erasing any of the variables would not increase it, so we kept this scale as is with modest reliability. Theoretically, these items are measuring the use of the technology and how the easiness to access the content enables engagement, so it was decided to keep the five items as Catching up, since all of the items are about easily keeping up with the content and experience of watching.
Six motivations were created: Entertainment, Relaxation, To Pass Time, Learning about things, Get Away and Catching up. An overview of the factor loadings, reliability scores, Eigen Values and variance explained can be found in Appendix C, Table I.
Cluster Analysis
Different tests were done with 2, 3, 4 and 5 clusters, and the ideal scenario was found at 3 clusters, when after ten iterations, the clusters stopped changing.
The totality of the 372 cases were included in the cluster analysis and the following variables were not considered for the cluster analysis because they were just used as inclusion criteria for the survey (Do you have (access to) a Netflix account?, Have you used it in the last three months?), and because they were nominal (Please name your favourite series to watch more than two episodes in a row, Please name the last series that you watched two or more episodes in a row) or dichotomous variables (Are you the primary holder of the account?).
Dummy variables were computed based on the cluster membership for each case of the data as Cluster 1 (n=64), Cluster 2 (n=107) and Cluster 3 (n=201). Although Cluster 3 seems more packed than the others, the size of the three clusters is acceptable to be labeled as different groups,
and although clusters above 100 individuals are recommended, less than 100 cases could just mean that we have an unrepresented cluster in our sample (Wedel & Kamakura, 2000).
An ANOVA table was formulated to distinguish which of the Netflix usage variables are statistically significant at a 95% sig. (p=0.05) level. Episodes per sitting, hours per sitting and days a week were the three significant variables. A table with the clustering variables and their
significance level can be found in the appendices (see appendix D, Table I).
Cluster 1: In this cluster, 34.4% of the members watch three episodes of the same series on a single sitting, while 62.3% watch four or more episodes. 57.8% of them watch on average of two to three hours and 32.8% watch from three or more than four hours. Moreover, 70.4% of the members indicated they behave like this two to three days a week, while 21.9% behave like this only one day a week. The rest, who end up being minority, do it four days a week. This cluster scored the lowest on the Horror genre and the highest (54.7%) on the Action and Adventure genre.
Cluster 2: Half of this group (50.5%) can be defined as on the edge of marathon-watchers since they indicated to watch two episodes of the same series in the same sitting, while 24.3% watches only one, and 25.2% three episodes. Also, 68.2% of the members of this cluster watch between one to two hours and 24.3% do it for two to three; a few more (6.5%) watch less than an hour. They indicated that mostly, they behave like this two or three days a week (71%) and 26.2% once a week, the minority (2.8%) indicated to watch four days a week. Content-wise, this cluster significantly scored the lowest on Thriller and Non-scripted genres, but scored in the middle in all the other genres.
Cluster 3: The members of this group are high marathon-watchers, with 47.3% indicating that they watch three or more episodes per sitting, 40.8% watching two episodes and only 11.9% one episode. This cluster stands out because its members watch more often, with 22.4% indicating that they watch Netflix four days a week, 25.4% five days, and 52.2% >5 days a week. When they behave like this, 50.7% do it for one to two hours, 28.4% two to three and 11.4% watch three to
four or more hours. This is the group that spends more time watching regularly, although they watch a similar amount of episodes to cluster 2. This cluster is the one in which, compared to the other clusters, members consume more of the following genres: Horror with 20.9%, Non-scripted with 16.4%, Comedies with 72.6/ and Docu-series with 39.8%. For more differences, see Figure 1.
Figure 1. Clusters demographic and basic Netflix use differences
The fact that we were able to create three different clusters corroborates our hypothesis that there are different types of marathon watchers, and our null hypothesis is rejected.
Clusters and Viewers’ Motivations
To cross the motivations to watch Netflix with the clusters, an ANOVA test was used; An ANOVA test is used to analyse the variance of means between groups, by testing the mean scores of the clusters and establishing if they differ statistically significant from each other (Green, 1973). Along with the ANOVA, a Bonferroni test was used to see which clusters differ from each other significantly and as it turns out, at a 95% significance level, only two motivations resulted
statistically significant between groups: Entertainment (M=3.31, SD=.788, p=.017) and Catching Cluster 1 17.2% /sample Cluster 2 28.8%/sample Cluster 3 54% /sample Relationship status Married/Committed 68.8% Single 31.2% Married/Committed 70.1% Single 29.9% Married/Committed 73.6% Single 26.4 Children Yes 39.1% No 60.9% Yes 48.6% No 51.4% Yes 53.2% No 46.8%
Study/Work < 40 hours p/week: 34.4% > 40 hours p/week: 65.6% < 40 hours p/week: 43% > 40 hours p/week: 57% < 40 hours p/week: 53.7% > 40 hours p/week: 46.3% Watch Netflix Accompanied 37.5% By themselves 62.5% Accompanied 41.1% By themselves 58.9% Accompanied 41.8% By themselves 58.2% Other activities while watching Yes 28.1% Sometimes 56.3% No 15.6% Yes 18.7% Sometimes 52.3% No 29% Yes 32.3% Sometimes 52.7%% No 14.9%
up (M=2.01, SD=1.20097, p=.000), while the others seem to be much more similar between groups (See Appendix D, Table II).
For the Entertainment motivation (p=.017, 95% CI [-.4750, -.0250]), Clusters 2 (M=4.33) and 3 (M=4.56) differ significantly from each other (p=0.24), meaning that while Cluster 2 scored high in Entertainment as a motivation to watch Netflix, Cluster 3 scored even higher. Although this two clusters are motivated by entertainment, Cluster 3 clearly has this motivation as the highest one and is also the one that watches more episodes more days a week. Furthermore, for the Catching up motivation, Cluster 2 (M=1.68) differs from Cluster 1 (M=2.42, p=.000, 95% CI [-1.1875, -.2918]) and Cluster 3 (M=2.05, p=.030, 95% CI [-.7037, -.0254]) significantly. This group in the most neutral content-wise from all the groups. An overview of all the motivations scores per cluster can be found in Appendix D, Table III.
With all these results, we can confidently say that we found support for our second
Hypothesis, as different clusters of Netflix marathon-watchers indeed have different motivations to watch. Although our findings just proved differences between some clusters and some motivations, we reject the null hypothesis.
Discussion
Our results shed light on a new path in terms of how we look at the marathon-watching phenomenon. We are now able to say that there are different types of marathon-watchers, and not only non-marathon-watchers and marathon-watchers. Within our discoveries, we have enough information to name and refer to this groups by their differences to each other, so we created the following groups of marathon-watchers:
Cluster 1 - Marathon Runners
Average episodes, more hours per sitting, less days per week. This cluster can sprint a marathon, they are used to exercise the habit hard in each training session, but don’t do it that often. This group occasionally run marathons and they really commit to it because they are prepared, in
shape, but cannot do it so often. As marathon runners, they tend to be more of a lonely wolf type, because as indicated before, this is the group that watches by themselves, instead of accompanied, more in comparison to the other clusters. This group are marathon watchers that run long in every sitting, but they don’t do it as often, with most of them only behaving like this two or three days a week, which is less than half of the week. Being the group that watches less days a week on average, this life situation could be influencing how they spend their time overall.
Cluster 2 - The Joggers
Average episodes, more hours, average days a week. They represent 28.8% of the sample. This is the starter cluster, the viewers that cannot run a marathon, but definitely can exercise the habit. Its’ members don’t actually run, but jog, as they keep watching constantly but not that often, and not that much in one sitting. They tend to maintain in shape and keep up with the content they consume. This particular cluster has the highest amount of people indicating that they don’t do other things while watching, and in line with this profile, it could indicate more commitment to enjoy the content, since they don’t have the opportunity to watch that often and not for that long. This group watches less episodes per sitting and only do so mostly two or three days per week, but they watch for a shorter amount of time, as they are slow runners. Mostly motivates by Entertainment, and given their content choices, this is the most conventional type of watcher, with no highlighted preferences on any genre.
Cluster 3 - Professional Runners
More episodes, more hours, more days a week. This cluster are the biggest consumers, they are the professional runners, the ones that not only run but also are training the habit quite often in long sessions. These people, therefore, have the opportunity to explore more content, and so they are the group that watches the most varied mix of genres. They study/work the less amount of hours in comparison to the other clusters, so they overall have more leisure time. In comparison with the other clusters, a determinant factor could be that they are able to watch Netflix more often than the
other clusters, starting to prove our point that life situations can influence watching habits.
Accordingly, the members of this group appear to be the most distracted, since the majority of their members indicated to do (sometimes) other things while watching (85%). In line with their profile, since they marathon-watch truly often, we can think of the possibility that they don’t pay so much attention to the content as much because they have the opportunity to watch Netflix almost every day.
As our results reflect, most of the people (n=322) in this study are indeed
marathon-watchers, and the ones that are not were still classified into a cluster that tends to watch on a regular basis. By looking at the overview of all mean scores per motivation, it is clear that most viewers are, in a way, demanding, since they scored high (more than 3) in almost all of them, with the exception of Catching up. This might also be a hint for us, researchers, to look further into hidden or non-researched motivations up until now, to be more creative and try to find out what is new.
In conclusion, by conducting this cluster analysis, we realised that not only viewers’ watch Netflix differently in terms of episodes, days and length, but they also tend to differ in the type of content they watch and what motivates the to feed their habits. These findings are vital to keep pointing further research into the right direction and finally understand the underlying factors behind marathon-watching and so we can start looking into the effects (if) that it has on the viewers’ psyche, media habits, and after, maybe we can start talking about health.
Theoretical and Practical Implications
Previous research has emphasised how to measure a marathon-watcher and has tried to establish whether watching is an addiction-like activity, but has treated Netflix marathon-viewers as one homogeneous sample. With this study, by looking at some demographic information, plus motivations to watch, we were able to take a step further in proving that marathon-watchers are indeed not an homogeneous, but rather an heterogeneous sample that differ from one another in terms of usage, but also content and motivations.
Based on these results, by acknowledging that Netflix-watching is not an activity that happens completely isolated from other activities, we realise that the circumstances that surround the habit can determine how it happens, meaning that when a person goes to the movies or is watching a show at a regular broadcast TV, it is less likely that they second-screen, read or talk on the phone because the circumstances are not adequate and they don’t have control over the viewing, risking missing out on the content. Netflix watching, on the other hand, happens with technology that allows the viewer to control the speed of watching (pause, rewind, forward), and enables other activities to be tangled into the watching experience.
There are some recognisable patterns about Netflix watching not being much of a social activity in terms of companionship, since a large amount of participants indicated watching by themselves, in line with other studies, such as one by MarketCast (2013), which indicated that marathon-watching is not a social activity, given that 56% of the viewers do it by themselves and 98% do it at their own home. This could imply that owning the entire watching experience, as in being in control of it, could be an important factor for the behaviour.
As our findings support our first hypothesis, we assume that there are different types of Netflix marathon-watchers. This study was able to provide an interesting snapshot of the type of viewers that watch Netflix and we can already see how they differ from users of other type of traditional media such as TV, regarding accessibility and hours spent watching the same series. Because this continuous watching schema was not possible with the TV, plus Netflix has no commercials, we could be looking at a more engaging way to deliver content with the streaming platforms (Pittman & Eanes, 2015). One thing to notice about Netflix use is that once an episode of a series is finished, and as long as there are more to stream, the platform will automatically cue the next episode for the viewer, and although there is an opt-out possibility, it is well known that Netflix makes it quite easy to marathon-watch.
Marathon-watching is indeed a mix of traditional TV watching and internet use, therefore this result in unique motivations that we are still exploring, and whilst this study did not have the clearest result on that matter, it for sure will be a guide for further research on the topic.
Nonetheless, this motivations seem to fit still within the Uses an Gratification theory, but step-in on uncharted territory as we approach to a more defined concept of marathon-watching, realising that is broader than we thought at first (Pittman & Sheehan, 2015).
Furthermore, not only academia is benefitted by studies like this, but for companies such as Netflix, input like the results here presented should be valuable as they can and must learn more about the consumers. Their platform works on algorithms as the viewers watch, but they miss out on the social factor, they don’t know why the viewers are choosing that series at that time, if they are choosing themselves, or if they just got caught-up in the story. What is true about this
algorithms is that their complexity has become so accurate with content suggestions, that it only makes it easier for a viewer to marathon-watch. It takes more effort to opt-out of a sitting (Pittman & Sheehan, 2015).
Watching Netflix can happen differently for each individual and although two people could be both marathon-watchers, one can do it once a week, while other can do it five times a week but for a smaller amount of time. What these results show in an overview, is that segmenting the Netflix watchers is a necessity to properly understand the viewers’ motivations and needs, but also the medium itself. This viewers are diverse and complex, and therefore should not be segmented by episodes watched anymore.
Limitations and Further Research
Studying motivations to watch Netflix has been investigated before, but not often crossed with Netflix usage nor demographics, and this first attempt to do so was an assertive step in the academic scene, but we must recognise that sample-wise, there are some issues that can be done better in future research.
As our sample was smaller than 500 respondents, Cluster 1 ended up having less than 100 members. Our sample consisted on more than 80% females participants and a more mixed sample could benefit future research to determine if gender can influence the marathon-watching habit. Moreover, 1/3 of the participants reside in Mexico, and 1/3 in the U.S.A, while the rest is spread all over the world. Since our results already hint that demographics can influence marathon-watching and knowing that the Netflix catalogue in each region is different, further research might benefit with a sample in specific regions or countries.
By looking only at series marathon-watching, we left out content such as movies, which we consider important to look into, as it is possible to marathon-watch, for example, the star Wars saga. How would that work? That is a questions for further research as we were no able to provide the answer given time and resources limitations. Our results also showed that some genres (Comedy and Drama) are more subject to marathon-watching, so more studies on the influence of the content is needed.
By using a K-means cluster, we determined that there are indeed different types of marathon-watchers, but with this analysis it is not possible to know which attribute contributes more to the grouping since it is assumed that each attribute has the same weight while creating the clusters. To overcome this issue we used a large amount of data, allowing the cases to have more distance to each others’ centers.
Moreover, the scale that we created to measure the motivations for people to watch Netflix was a first attempt to include several factors for different types of mediums (TV and internet). It seems like many of our scales were two items or less, which lead to instability, issue that we contrasted by leaving one item per motivation. To develop more reliable scales, this paper provides ideas on what works and what doesn’t.
Furthermore, because this study sampled people from all over the world, it is difficult to determine whether the results could be generalised, since Netflix is present in more than 190 cities,
and they constantly generate new content, even with new characteristics as their more recent “create you own story” interactive series/movie content Black Mirror: Bandersnatch, which leave current research a bit behind, but with a clear path to stay tuned and creating innovative ways to learn more about marathon-watching. Interactive content will change the future of this line of research, because not only it enables the viewer to get involved in the story, but also requires of them much more attention to the content as they play a role in the story.
Conclusion
Overall, we found unexpected but enlightening results, and we were able to conclude that there are indeed different types of marathon watchers, but it is still needed to determine in which and how many ways they differ in wider research. This paper reinforced a precedent by referring to the activity as marathon-watching and not binge-watching. By doing so, we took away the stigma that comes with it and the research had a new, refreshing perspective. This papers’ biggest
contribution is that it took a step further into looking at the marathon-watching phenomenon from a more comprehensive manner that it is not only based on the number of episodes consumed, but also on the viewers themselves.
Research on this topic might hint us about how engagement factors are beginning to take over the viewers’ because of the difference that content makes in their daily lives, and along with aesthetics provided by new technologies, marathon-watching is being re-coded by how much escapism content provides (Smith, 2014).
This research can lead the way to look into this topic widely as a a social, technological and creative matter. Our advancements in the entertainment field leaves some questions that are left unanswered, but nonetheless made: Should we keep relying on theories that were developed in a completely different entertainment-scenario? Should we focus on the consumer at the same level that we do on the content? Who is defining who, research defines industry or industry defines research? Are we in the middle of a research black mirror reality?
References
Abelman, R. (1987). Religious television uses and gratifications. Journal of Broadcasting & Electronic Media, 31, 293–307.
Barton, K. M. (2009). Reality television programming and diverging gratifications: The influence of content on gratifications obtained. Journal of Broadcasting & Electronic Media, 53(3), 460-476.
Biernacki, P., & Waldorf, D., (1981). Snowball sampling: Problems and techniques of chain referral sampling. Sociological methods & research. 10, 2, pp 141-163.
Boca, I. P. (2017). Binge-watchers. Behavior patterns and emotions. Studia Universitatis Babes- Bolyai - Ephemerides. Retrieved January 15, 2019, from https://www.ceeol.com/search/ article-detail?id=711927
Chadwick, A. (2017). The hybrid media system: Politics and power. Oxford University Press. Conway, J. C., & Rubin, A. M. (1991). Psychological predictors of television viewing motivation.
Communication Research, 18(4), 443–463. doi:10.1177/009365091018004001 Cortés, C. E. (2000). The Children Are Watching: How the Media Teach about Diversity.
Multicultural Education Series. Teachers College Press, 1234 Amsterdam Avenue, New York, NY.
Costello, A. B., & Osborne, J. W. (2005). Best practices in exploratory factor analysis: Four
recommendations for getting the most from your analysis. Practical assessment, research & evaluation, 10(7), 1-9.
Fillmore, M. T., & Jude, R. (2011). Defining “binge” drinking as five drinks per occasion or drinking to a. 08% BAC: which is more sensitive to risk?. The American Journal on Addictions, 20(5), 468-475.
Flayelle, M., Maurage, P., & Billieux, J. (2017). Toward a qualitative understanding of binge- watching behaviors: A focus group approach. Journal of behavioral addictions, 6(4),
457-471.
George, D. R., Rovniak, L. S., & Kraschnewski, J. L. (2013). Dangers and opportunities for social media in medicine. Clinical obstetrics and gynecology, 56(3).
Gliem, J. A., & Gliem, R. R. (2003). Calculating, interpreting, and reporting Cronbach’s alpha reliability coefficient for Likert-type scales. Midwest Research-to-Practice Conference in Adult, Continuing, and Community Education.
Green, P. E. (1973). On the analysis of interactions in marketing research data. Journal of Marketing Research, 410-420.
Greenberg, B. S., & Hnilo, L. R. (1996). Demographic differences in media gratifications. Journal of Behavioral and Social Sciences, 1, 97-114.
Jenner, M. (2016). Is this TVIV? On Netflix, TVIII and binge-watching. New media & society, 18(2), 257-273.
Katz, E., Blumler, J., & Gurevitch, M. (1973). Uses and Gratifications Research. The Public Opinion Quarterly, 37(4), 509-523.
LaRose, R., & Eastin, M. S. (2004). A social cognitive theory of Internet uses and gratifications: Toward a new model of media att endance. Journal of Broadcasting & Electronic Media, 27 48(3), 358-377.
Leung, D., Law, R., Van Hoof, H., & Buhalis, D. (2013). Social media in tourism and hospitality: A literature review. Journal of travel & tourism marketing, 30(1-2), 3-22.
Palmgreen, P. and Rayburn, J. (1979), “Uses and gratifications and exposure to public television”, Communication Research, Vol. 6 No. 2, pp. 155-180
Papacharissi, Z., & Rubin, A. M. (2000). Predictors of Internet use. Journal of Broadcasting & Electronic Media, 44(2), 175–196. doi:10.1207/s15506878jobem4402_2
Pierce-Grove, R. (2016). Just one more: How journalists frame binge watching. First Monday, 22(1).
Pittman, M., & Sheehan, K. (2015). Sprinting a media marathon: Uses and gratifications of binge- watching television through Netflix. First Monday, 20(10).
Rubin, A. M. (1981). An examination of television viewing motivations. Communication Research- An International Quarterly, 8(2), 141–165. doi:10.1177/009365028100800201
Ruggiero, T. (2000), Uses and gratifications theory in the 21st century, Mass Communication & Society, Vol. 3 No. 1, pp. 3-37.
Sherry, J., Greenberg, B., Lucas, S., & Lachlan, K. (2006). Video game uses and gratifications as predictors of use and game preference. In P. Vorderer & J. Bryant (Eds.), Playing computer games: Motives, responses and consequences. Mahwah, NJ: Lawrence Erlbaum Associates. Song, I., Larose, R., Eastin, M. S., & Lin, C. A. (2004). Internet gratifications and Internet
addiction: On the uses and abuses of new media. CyberPsychology & Behavior, 7(4), 384-394.
Stafford, T. F., & Gonier, D. (2004). What Americans like about being online. Communications of the ACM, 47(11), 107-112.
Steiner, E., & Xu, K. (2018). Binge-watching motivates change: Uses and gratifications of streaming video viewers challenge traditional TV research. Convergence,
1354856517750365.
Sung, Y. H., Kang, E. Y., & Lee, W. N. (2018). Why Do We Indulge? Exploring Motivations for Binge Watching. Journal of Broadcasting & Electronic Media, 62(3), 408-426.
Tokinoya, H. (1996). Part IV: New Studies in Media Gratifications: A Typological Study of Media Gratifications Theory in Japan (< Special Issue> Uses and Gratifications from the Mass Media: New Perspectives). 行動科学研究, 1996(1), 115-137.
Tukachinsky, R., & Eyal, K. (2018). The Psychology of Marathon Television Viewing: Antecedents and Viewer Involvement. Mass Communication and Society, 21(3), 275-295.
TV consumption: Binge Watching and Marathon Watching. ICIS 2017 Proceedings. Weaver Lariscy, R., Tinkham, S. F., & Sweetser, K. D. (2011). Kids these days: Examining
differences in political uses and gratifications, Internet political participation, political information efficacy, and cynicism on the basis of age. American Behavioral Scientist, 55(6), 749-764.
Wedel M, Kamakura WA (2000) Market segmentation: conceptual and methodological foundations, 2nd edn. Kluwer Academic, Boston, NE.
Whiting, A., & Williams, D. (2013). Why people use social media: a uses and gratifications approach. Qualitative Market Research: An International Journal, 16(4), 362-369.
Yee, N. (2006). Motivations for play in online games. CyberPsychology & Behavior, 9(6), 772–775. doi:10.1089/cpb.2006.9.772
Zillmann, D. (1988). Mood management through communication choices. American Behavioral Scientist, 31(3), 327-340.
Non-academic resources
About Netflix. (n.d.). Retrieved January 19, 2019, from https://media.netflix.com/en/about-netflix Feeney, N. (2018, March 28). When, Exactly, Does Watching a Lot of Netflix Become a 'Binge'?
Retrieved October 7, 2018, from https://www.theatlantic.com/entertainment/archive/ 2014/02/when-exactly-does-watching-a-lot-of-netflix-become-a-binge/283844/
Graser, M. (2013, March 13). 10 Insights from Studies of Binge Watchers. Retrieved October 9, 2018, from https://variety.com/2013/digital/news/10-insights-from-studies-of-binge- watchers-1200004807/
Lynch, J. (2018, August 28). Guy Pearce says Netflix hates the term 'binge-watching' now, and told him not to use it in interviews. Retrieved September 26, 2018, from https://
pearce-2018-8/?international=true&r=US
MarketCast. (2013). MarketCast study finds TV “binge-viewing” creates a more engaged
viewer for future seasons and not a bingeing habit. Retrieved from http://www.prweb.com/ releases/2013/3/prweb10513066.htm
Netflix Technology Blog. (2015, December 01). Caching Content for Holiday Streaming – Netflix TechBlog – Medium. Retrieved October 21, 2018, from https://medium.com/netflix- techblog/caching-content-for-holiday-streaming-be3792f1d77c
Ramsay D. (2013) Confessions of a binge watcher. CST Online. Available at: http://cstonline.tv/ confessions-of-a-binge-watcher (accessed 17 December 2018).
Smith, C. (2014). “The Netflix effect: How binge watching is changing television,” TechRadar (16 January), at http://www.techradar.com/us/news/internet/the-netflix-effect-how-binge- watching-is-changing-television-1215808, Retrie4ved December 28, 2018.
Spangler, T. (2013, December 14). Netflix Survey: Binge-Watching Is Not Weird or Unusual. Retrieved November 27, 2018, from https://variety.com/2013/digital/news/netflix-survey- binge-watching-is-not-weird-or-unusual-1200952292/
Vanian, J. (2018, April 19). Amazon Has Over 100 Million Prime Members. Retrieved from http:// fortune.com/2018/04/18/amazon-prime-members-millions/
Welch, C. (2018, May 02). Hulu passes 20 million US subscribers, says offline downloads are coming. Retrieved January 19, 2019, from https://www.theverge.com/2018/5/2/17309336/ hulu-20-million-subscribers-announced-offline-downloads-new-feature
APPENDIX A - Demographics
Table I: Demographics of the respondents (n=372)
Table II: List of participants’ countries of residency
Country of residency Frequency Cumulative Percent
Andorra 1 .3 Argentina 1 .5 Australia 6 2.2 Austria 2 2.7 Brazil 1 3 Canada 23 9.1 Colombia 2 9.7 Costa Rica 1 9.9 Denmark 1 10.2 Finland 1 10.5 France 5 11.8 Germany 10 14.5 Guatemala 1 14.8 Ireland 1 15.1 Italy 4 16.1 Japan 1 16.4 Malta 1 16.7 Mexico 121 9.2 Netherlands 62 65.9 New Zealand 1 66.1 Poland 1 66.4 Portugal 1 66.7 Romania 1 66.9 Saudi Arabia 1 67.2 Spain 4 68.3 Sweden 3 69.1 Switzerland 1 69.4 UAE 1 69.6 UK 9 72 USA 104 100 Total 372 100 Characteristics
Gender Age Relationship status Children Study/Work hours
Male 55 18-28 92 Married/relationship 267 Yes 184 Less than 40 176
Female 317 29-38 140 Single 105 No 188 40 or more 196
39-48 118
APPENDIX B - Netflix Use Table I - Watching habits
Table II - Content
Table III. Most marathon-watched series (top 20)
*Narcos 24
Friends 23
The Haunting of Hill House 21
House of Cards 15
The Chilling Adventures of Sabrina 14
*House of Flowers 14
Money Heist 12
The Bodyguard 11
Elite 10
How I Met Your Mother 9
Suits 9
The Sinner 8
Cable Girls 7
Grey's Anatomy 6
Orange is the New Black 6
The Crown 6
The Good Place 6
Riverdale 5 Stranger Things 4 *Club de cuervos 4 *El Chapo 3 Watching description Episodes per
sitting Own decision what to watch Days per week Length of watching (hours) Place to watch
One 50 Yes 295 One 42 < one 15 Own house 3
6 1
Two 138 Sometime
s 67 Two 60 One to two 181 Someone else’s house 10
Three 118 No 10 Three 61 Two to three 120 Public transport 1
Four 33 Four 53 Three to four 35
> five 33 Five 51 > four 21
> five 105
Genres Action and
Adventure Comedies Thrillers Docu-series Drama Horror ScriptedNon- Romance Sci-Fi
171 261 98 143 207 61 48 191 170
*This content might not be available in all locations because Netflix catalogue changes depending on location.
Table IV. Other activities (not listed in survey) participants do while watching Netflix
Cleaning, house chores 6
Cooking 1 Crocheting 2 Eating 2 Exercise 1 Folding clothes 1 Ironing 3 Knitting 2 Laundry 2 Make-up 1
Organising and cleaning 1
Play with my kids 2
Playing 1
Playing with kids 1
APPENDIX C - Motivations to watch Netflix Table I - Factor analysis for viewers’ motivations
Motivation factor and items LoadingsFactor Eigen Value explained (%)Variance Reliability coefficient Entertainment
* To be entertained
Relaxation
Because it helps me to relax
To Pass Time 1.332 11.4% .640
Because it helps me to pass time .738
When I’m bored .803
When there is no one else to talk to or be with .493
Learning about things 1.201 11.36% .613
To learn about people, places and things .812 To keep up about current issues and events .758
Get Away 4.993 20.17% .803
It helps me to forget my problems .791
I want to get away from things .815
I feel satisfied when I finish watching it .568 It makes me feel like I have achieved something
when I finish it .666
It makes me feel less lonely when there is no
one else around .658
To have a beautiful experience for my senses .486
Catching Up 1.754 15.06% .706
Because of the quality of the technology I watch
it in (e.g. 4K smart TVs) .540
Is easy to watch it in my smartphone or tablet .575 I want to keep up with the story .797 To catch-up with stories I haven't watched for a
while or that I left half-way. .789
To give me things to talk about with other
people .544
APPENDIX D - Cluster Analysis, Netflix usage and ANOVA table Motivations Table I: Netflix use variables and their significance
Variable Mean square F Sig.
With which gender you identify yourself with? .369 2.952 .053
Are you married or in a committed relationship? .625 3.113 .046
Do you have children? .278 1.110 .331
How many hours do you work/study outside home per week?
4.222 6.113 .002
When you choose content on Netflix, do you decide what to
watch yourself? (No one else decides what you are watching) 2.579 3.989 .019 How many days per week do you watch any content on Netflix? 449.128 673.537 .000 When you watch any content on Netflix, on average, for how
long do you watch? 48.642 83.699 .000
On a regular watching day, how many episodes of the same
series do you watch? 61.484 69.150 .000
When you watch Netflix, do you usually do it by yourself or
accompanied? 1.119 4.624 .010
When you watch Netflix, do you do any other activities at the
same time? 1.464 3.210 .041
What kind of content (genre) do you usually watch? Action and
Adventure (e.g. Daredevil) .526 1.990 .138
What kind of content (genre) do you usually watch? Comedies
(e.g. How I met your mother) .074 .353 .703
What kind of content (genre) do you usually watch? Thriller (e.g.
Manhunt: Unabomber) 1.097 5.783 .003
What kind of content (genre) do you usually watch? Docu-series
(e.g. Making a murderer) .029 .120 .887
What kind of content (genre) do you usually watch? Drama (e.g.
House of Cards) .601 2.448 .088
What kind of content (genre) do you usually watch? Horror (e.g.
The Haunting of Hill house) .890 6.671 .001
What kind of content (genre) do you usually watch? Non-scripted
(e.g. Queer eye) .435 3.924 .021
What kind of content (genre) do you usually watch? Romance
(e.g. LOVE) .111 .443 .642
What kind of content (genre) do you usually watch? Sci-fi and
Fantasy (e.g. Black Mirror or Sens8) .568 2.299 .102
*The F tests should be used only for descriptive purposes because the clusters have been chosen to maximize the differences among cases in different clusters.
Table II. Cluster Analysis and ANOVA table Netflix viewers’ Motivations
Table III. ANOVA table for viewers’ Motivations Cluster analysis and viewers’ motivations
Cluster Entertainment Relaxation To Pass Time about thingsLearning Get Away Catching Up
1 Mean 4.37 3.31 2.48 2.09 1.29 2.42 N 64 64 64 64 64 64 Std. Deviation .82616 1.37869 1.72739 1.68767 .88254 1.19263 2 Mean 4.33 3.11 2.28 2.14 1.18 1.68 N 107 107 107 107 107 107 Std. Deviation .81003 1.57413 1.69235 1.77747 1.15135 1.18065 3 Mean 4.56 3.23 2.68 2.44 1.28 2.05 N 201 201 201 201 201 201 Std. Deviation .75185 1.49541 1.81381 1.69037 1.10881 1.17286 Total Mean 3.31 3.21 2.53 2.29 1.25 2.01 N 372 372 372 372 372 372 Std. Deviation .78852 1.49541 1.76876 1.71804 1.08451 1.20097
Clusters * viewers’ motivations
Sum of squares Df Mean Square F Sig.
Entertainment * Clusters Between Groups 5.069 2 2.534 4.145 .017 Within groups 225.606 369 611 Total 230.675 371 Relaxation * Clusters Between Groups 1.768 2 .884 .394 .675 Within groups 827.877 369 2.244 Total 829.645 371 To Pas Time * Clusterrs Between Groups 11.121 2 5.561 1.785 .169 Within groups 1149.553 369 3.115 Total 1160.675 371 Learning about things * Clusters Between Groups 9.254 2 4.627 1.573 .209 Within groups 1085.807 369 2.943 Total 1095.062 371
Get away * Clusters Between Groups .880 2 .440 .373 .689 Within groups 435.475 369 1.180 Total 436.355 371 Catching up * Clusters Between Groups 22.618 2 11.309 8.143 .000 Within groups 512.486 369 1.389 Total 535.105 371
APPENDIX E - Survey questions
Thank you for agreeing to participate in this project being carried out under the auspices of the ASCoR research institute, which forms part of the University of Amsterdam. ASCoR conducts scientific research into media and communications in society.
The title of the research project for which we are requesting your assistance is ''Understanding Netflix Users''. The objective of the research is to learn more about how people use Netflix.
In order to participate in this project: - you must be 18 years or older; - be able to understand English; and - be a current Netflix user.
For this project, you will complete an online survey. It should take approximately 10 minutes to complete.
As this research is being carried out under the responsibility of the ASCoR, University of Amsterdam, we can guarantee that:
1) Your anonymity will be safeguarded, and that your personal information will not be passed on to third parties under any conditions, unless you first give your express permission for this.
2) You can refuse to participate in the research or cut short your participation without having to give a reason for doing so. You also have up to 24 hours after participating to withdraw your permission to allow your answers or data to be used in the research.
3) Participating in the research will not entail your being subjected to any appreciable risk or discomfort, the researchers will not deliberately mislead you, and you will not be exposed to any explicitly offensive material.
For more information about the research and the invitation to participate, you are welcome to contact the project supervisor, Dr Jessica Piotrowski (j.piotrowski@uva.nl), or the student leader, Karla Hernández Zaldívar (karla.hernandezzaldivar@student.uva.nl) at any time.
Should you have any complaints or comments about the course of the research and the procedures it involves as a consequence of your participation in this research, you can contact the designated member of the Ethics Committee representing ASCoR, at the following address: ASCoR secretariat, Ethics Committee, University of Amsterdam, Nieuwe Achtergracht 166, 1018 WV Amsterdam; 020-525 3680; ascor-secr-fmg@uva.nl. Any complaints or comments will be treated in the strictest confidence.
We hope that we have provided you with sufficient information.
If you would like to participate in this study, please click on the arrow below and make sure that the English version is turned on in the settings, in order to see this survey properly.
We would like to take this opportunity to thank you in advance for your assistance with this research, which we greatly appreciate!
Kind regards,
Informed consent for participation in the study 'Netflix watchers, who are they?''
I hereby declare that I have been informed in a clear manner about the nature and method of the research, as described in the email invitation for this study.
I agree, fully and voluntarily, to participate in this research study. With this, I retain the right to withdraw my consent, without having to give a reason for doing so. I am aware that I may halt my participation in the experiment at any time.
If my research results are used in scientific publications or are made public in another way, this will be done such a way that my anonymity is completely safeguarded. My personal data will not be passed on to third parties without my express permission.
If I wish to receive more information about the research, either now or in future, I can contact
karla.hernandezzaldivar@student.uva.nl. Should I have any complaints about this research, I can contact the designated member of the Ethics Committee representing the ASCoR, at the following address: ASCoR secretariat, Ethics Committee, University of Amsterdam, Postbus 15793, 1001 NG Amsterdam; 020- 525 3680; ascor-secr-fmg@uva.nl.
I consent, begin the study
I do not consent, I do not wish to participate in the study - Skip To: End of Survey If “I hereby dec... = I do not consent, I do not wish to participate in the study”
Demographics What is your age?
With which gender you identify yourself with Female
Male
Other (specify)
I would rather not to say
In which country do you currently reside? ▼ Afghanistan (1) ... Zimbabwe (1357) Are you married or in a committed relationship
Yes No
Do you have children Yes
No
How many hours do you work/study outside home per week? Less than 40 hours per week
40 hours per week
More than 40 hours per week
We are interested on people that has used Netflix at least once in the last three months. Please answer the following questions to know your situation.
Do you have (access to) a Netflix account Yes
No
Have you used it in the last three months Yes
