Discovering new media : uses and gratifications for video on demand and its substitution effect on linear television viewing

Discovering New Media

Uses and Gratifications for Video on Demand

and its Substitution Effect on Linear Television Viewing

Statement of Originality This document is written by Student Vita Stokvis who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

Table of contents. Abstract 4 1. Introduction 5 2. Theory 10 2.1 Television development 10 2.2 Uses and gratifications 13 2.3 Content gratifications 15 2.4 Process gratifications 18 2.5 Social gratifications 21 2.6 Media substitution 23 2.7 Mediation 26 2.8 Theoretical model 26 3. Method 28 3.1 Empirical strategy 28 3.2 Respondents 29 3.3 Measurements 30 3.3.1 Video on Demand 31 3.3.2 Entertainment 31 3.3.3 Content availability 31 3.3.4 Diversion 32 3.3.5 Convenience 32 3.3.6 Social utility 33 3.3.7 Frequency of VOD usage 33 3.3.8 Linear television 34 3.3.9 Control variables 34 3.3 Procedures 35 3.3.1 Survey 35 3.3.2 Statistical analyses 36

4. Results 39 4.1 Principal Component Analysis 39 4.2 Analyses for research question 1 42 4.2.1 Entertainment and higher frequency of VOD usage 45 4.2.2 Content availability and higher frequency of VOD usage 46 4.2.3 Diversion and higher frequency of VOD usage 47 4.2.4 Convenience and higher frequency of VOD usage 47 4.2.5 Social utility and higher frequency of VOD usage 47 4.2.6 Control variables 48 4.3 Analyses for research question 2 48 4.3.1 Frequency of VOD usage and decrease in use of linear television 49 4.3.2 Control variables 50 4.4 Analyses for mediation of frequency of VOD usage. 50 5. Discussion 52 5.1 Significance of findings 52 5.2 Implications 56 5.3 Limitations and future research 57 6. Conclusion 59 Reference list 60 Appendix 1 71 Appendix 2 72 Appendix 3 77 Appendix 4 79 Appendix 5 81

Abstract. This research investigates the uses and gratifications of Video on Demand and its effect on linear television viewing with a cross-sectional survey design. Since Video on Demand is the result of a merge between Internet and television, the most fundamental motives for the use of these two media forms are applied to Video on Demand. Principal component analysis identified five motives for the use of Video on Demand: Entertainment, content availability, diversion, convenience, and social utility. Linear regression analysis on the relation between each motive and frequency of Video on Demand usage pointed out that entertainment, content availability, and diversion lead to a higher frequency of use of Video on Demand. Additionally, it was found mostly young adults in the age of 18-29 (generation Y) watch Video on Demand. Furthermore, the effect of Video on Demand usage on linear television viewing was tested. In accordance to the media substitution theory, the vast majority of the respondents indicated to have lowered their amount of linear television viewing since becoming acquainted with Video on Demand. Also gender (females) and lower education were found to have an effect on the decrease of use of linear television. Video on Demand is determined to be the next major step in the evolution of media content delivery, and further research on Video on Demand in the field of both uses and gratifications and media substitution is encouraged.

1. Introduction. Since the 1940’s, uses and gratification researchers try to find out why people use certain media. The uses and gratifications theory is a widely accepted perspective that provides understanding of consumer behavior, and the determinants of this behavior in terms of media choice and media use (Kang & Atkin, 1999; Kaye, 1998; Lull, 1980). The gratifications are an aspect of satisfaction, and they can be defined as satisfied expectations, needs, and desires (Stafford et al., 2004). Passive offline platforms like cable television, radio, newspapers, magazines, books, and other printed publications are referred to as traditional media (Weiss, 1966). A lot of uses and gratifications research is done for these traditional media. It is found that people listen to the radio to seek companionship (Rubin, 1983; Ruggiero, 2000; Sundar & Limperos, 2013), read for in search of diversion (Payne, 1988), and watch television for entertainment, companionship, and to pass time (Ferguson & Perse, 2000; Kaye 1998; Palmgreen & Rayburn, 1979). The introduction of the Internet caused an explosive growth of what we refer to as new media. New media are all outlets that use digital technology and allow interaction between the user and the medium, like Internet, digital television, and social media channels (Bezjian-Avery, Calder & Dawniacobucci, 1998; Bittman, Rutherford, Brown & Unsworth, 2011; Henten & Tadayoni, 2008; Jenkins, 2006; Stafford & Stafford, 1996; Sundar & Limperos, 2013). All these new media can be used on many different (mobile) devices. According to Dwyer (2010), “one major difference between media industries today and their predecessors is the proliferation of delivery modes and media platforms” (p. 12).

Together with new media came new areas for uses and gratifications research to find out why people use these new media, and to discover if new media and their technologies can potentially create and satisfy new gratifications (Lichtenstein & Rosenfeld, 1983; Lin, 2004; Palmgreen, Wenner & Rayburn, 1981; Sundar & Limperos, 2013). Existing uses and gratifications like entertainment, companionship, and diversion were found for some new media forms (D’Ambra & Rice, 2001; Ferguson & Perse, 2000; Kang & Atkin, 2000; Sherry, Lucas, Greenberg & Lachlan, 2006). But furthermore, new uses and gratifications, that had not been identified for traditional media, were discovered for new media. Convenience, for instance, was found for the use of Internet for the first time (D’Ambra & Rice, 2001; Kaye & Johnson, 2003). New media can bring new features through technological developments, or through convergence. Media convergence is the merge of media where different technologies integrate into a new medium, device, or platform, and often completely displace traditional media (Bittman et al., 2011; Einav & Carey, 2009; Ruggiero, 2000). It was around the turn of the century that researchers foresaw that Internet and television would become more and more alike, and eventually merge into a new medium (Kaye, 1998; Lin, 2001). Today, Internet and television have indeed blended in together (Dwyer, 2010; Jenkins, 2006; Sundar & Limperos, 2013). As a result, linear television viewing (watching a program on television at the actual broadcasting time-slot) is decreasing in popularity (Cha, 2013; Ferguson & Perse, 2000; Henten & Tadayoni, 2008; Matrix, 2014; Pittman & Sheehan, 2014), because of one immensely popular new medium that has emerged from the convergence between Internet and television: Video On Demand (VOD). Despite the interest, and substantial amount of recent research on uses and gratifications of new

media, research specifically on VOD is very limited. VOD has become a serious threat to traditional television viewing (Pittman & Sheehan, 2014), and it is found to be “the next major step in the evolution of media content delivery” (Yu, Zheng, Zhao & Zheng, 2006, p. 333). Even the television rating measurements include VOD viewers nowadays, and it is consequently important to gain more insight in this form of media consumption which is currently developing and gaining popularity in a very high pace (Henten & Tadayoni, 2008). In this research, VOD refers to all forms of television consumption other than linear television. VOD enables users to watch content wherever and whenever they want (Henten & Tadayoni, 2008). Through the digital television set-top box, viewers can record television programs for later use, pause and resume a program while watching it, and catch up with programs after they have been broadcasted. Several programs also have their own website that offers the possibility to watch full episodes and seasons. VOD has now expanded with the possibility to subscribe to an advertisement-free, paid network like Netflix or HBO. These networks offer remote users a digital library full of video content, and allows users to play and playback any video at any time through internet streaming (Bhadada & Sharma, 2010; Hua, Cai & Sheu, 1998; Cai, 2004). Subscribers can watch television content online and online content on the television, and watch high quality series, films, and documentaries as much as they want, where they want, when they want (Baas, 2015; Henten & Tadayoni, 2008; Matrix, 2014; Netflix, 2015a). VOD content is suited for consumption on multiple (mobile) devices like television, laptops, tablets, and smartphones (Dwyer, 2010).

This research tries to contribute to the knowledge of consumer consumption behavior of new media, by investigating the user motives for watching VOD. RQ1: Which motives lead to a higher frequency of Video on Demand usage? However, Rubin (1981) posits that the search of user motives alone is not sufficient, and it is accordingly very important to also take the actual effects of media use into account. Therefore, this research will additionally investigate the consequences of VOD on linear television, based on the media substitution theory. According to this theory, VOD users will spend less time on watching linear television since they watch VOD, because they perceive VOD as a better substitute for linear television (Cai, 2004; Kang & Atkin, 1999; Neuman, 1988). Although this theory may seem logical at first, research findings are not yet consistent. As most studies have pointed out that linear television is indeed suffering from the use of online media (Cha, 2013; Kayany & Yelsma, 2000; Kaye, 1998; Kaye & Johnson, 2003; Lin, 1998; Vitalari, Venkatesh & Gronhaug, 1985), others did not find substitutable relationships between traditional and new media (Grotta & Newsome, 1982; Robinson, Barth & Kohut, 1997). In order to expand insights in consequences of new media use, the second research question will try to gain more understanding and consistency in the media substitution theory regarding linear television. RQ2: What are the consequences of the use of Video on Demand on linear television viewing?

In the following sections, first, a literature review will be given which will elaborate more on both uses and gratifications theory and media substitution theory. Additionally, several hypotheses will be formulated in order to investigate the research questions. Furthermore, an illustration of the conducted research methodology will be given in the method section. After this, the research results will be presented, followed by a discussion and a conclusion, in which the research questions will be answered, and implications for future research will be given.

2. Theory. This chapter will provide a review of the literature on the uses and gratifications theory and the media substitution theory. To start, an overview of the development from traditional television towards VOD is presented. Following, the uses and gratifications theory is elaborated on. Since this theory is not yet applied widely on VOD, the most relevant research findings in the field of Internet and television will be discussed and applied to VOD, which will result in the formation of several hypotheses. Subsequently, the media substitution theory will be outlined and applied to VOD, resulting in the last hypothesis. This chapter will end with the presentation of the theoretical model. 2.1 Television development The history of television development can be roughly divided into three periods. The first period, dating from the mid 1950’s to the early 1980’s, was the era with scarce offerings by only a small number of networks that reached the mass audience. The second period, following until the late 1990’s, was all about channel and network expansions, branding strategies, and more quality television offerings (Jenner, 2014). It was within these periods that the videocassette recorder (VCR) - a separate device that enables to record television programs on a videocassette for later playback - entered the consumer market and gained great popularity (Henke & Donohue, 1989). As the VCR made time-shifted viewing possible, this indicates that consumers back then were already interested in displacing the actual broadcasting schedule (Lin, 2001). Furthermore, around the 1990’s, the Internet gained public face as its number of active users grew rapidly (Brady & Elkner, 2011), and Internet

“became a major vehicle of mass communication in the late 1990’s” (Sundar & Bellur, 2011, p. 485). Devices in living rooms got technologically improved, like the shift from VCR to DVD (a Digital Video Disc with the same functions as a VCR, but in the size of a CD), from cable television to digital television, and from desktop computer to laptop. Especially due to the emergence of the Internet, the computer and television were expected sooner or later to technologically unite into sophisticated new media forms with many applications (Berniker, 1995; Ebersole, 2000; Ha & James, 1998; Jenkins 2006; Kaye, 1998; Lichtenstein & Rosenfeld, 1983; Newhagen & Rafaeli, 1996; Ruggiero, 2000). This led to the third period, dating from the late 1990’s to the present. This period is all about the increase of digital distribution platforms and audience fragmentations (Jenner, 2014). Digital television caused a big growth in the number of television channels, as regional, international, intercontinental, and all sorts of niche channels could easily be received everywhere (Ha & James, 1998). Sundar and Limperos (2013) state that media convergence is typical for this day and age as online content overlaps many traditional media forms. As a result, there has recently been an explosive growth of new interactive media and it was in this third period that the predicted convergence between television and Internet was realized, resulting in VOD (Cha & Chan-Olmsted, 2012; Dwyer, 2010; Einav & Carey, 2009; Jenkins, 2006; Kaye, 1998; Lin, 2001; Matrix, 2014; Sundar & Limperos, 2013). Nowadays, people can watch television content online and they can browse the web on their television, and so the Internet is becoming more like the television and vice versa. Digital television together with the Internet made time shifted viewing an integrated part of the television viewing experience. Through time shifted viewing, users can play, pause,

rewind, or fast forward content at any moment (Sundar & Limperos, 2013). At first, catch up television became available on the broadcaster’s web sites. This way, viewers did not need to set devices in advance to record a program, but complete episodes and seasons of television programs could be viewed online. But as Internet and television have converged, catch up television also became available directly on the television. This way, viewers could watch any television program they want, when they want, where they want; television could be viewed on demand. The era of media convergence has inevitably led to new business opportunities (Dwyer, 2010). Netflix, for example, is a company that started as a DVD rental service, but developed itself into one of the biggest and most well know online video streaming networks of the world today (Pittman & Sheehan, 2015). So besides catching up on programs from the broadcasting schedule and watching online, content can also be directly streamed from the Internet. Viewers need to subscribe to a VOD network like Netflix or HBO and pay a monthly fee. On these networks, not only regular movie, television, and documentary content is offered, but also movies, series, and documentaries that are produced exclusively for the VOD network. Especially series are proven to be tremendously popular amongst VOD viewers (Jenner, 2014; Matrix, 2014; Pittman & Sheehan, 2015). VOD content can be consumed both on television and on a device with an Internet connection, like a laptop, tablet, or smartphone. VOD offers the possibility to watch television over the Internet, with additional functional unique characteristics as superior content, technical benefits, convenience and efficiency (Cai, 2004; Ferguson & Perse, 2000; Haridakis & Hanson, 2009; Kaye, 1998; Lin, 2001; Matrix, 2014). The development of VOD opportunities and the emergence of specific VOD networks has fundamentally changed

television consumption, and we may be entering a new, fourth, period in television history (Jenkins, 2006; Jenner, 2014; Pittman & Sheehan, 2015). 2.2 Uses and gratifications The uses and gratifications theory is a widely accepted theoretical perspective for studying media consumption, as it explores why people become involved with certain types of media, and the consequences of that media use; the gratifications they obtain from it (Kang & Atkin, 1999; Rubin, 1984; Ruggiero, 2000; Stafford et al., 2004). Early theories on media consumption, like the hypodermic needle theory and the agenda setting theory, stated that the media have power over the consumers (Shaw, 1979). In Jenner’s (2014) first stages of television development, audiences sat together in front of the television and the broadcasters determined where they would look at. The old television consumers, and traditional media users in general, were passive collectives, and consequently, theories encompassed the effect of media on people (Jenkins, 2006; Kang & Atkin, 1999; Metzger & Flanagin, 2002; Rubin, 1983). Contrary to these theories, the uses and gratifications theory questions not about what media do to people, but what the people do to the media (Klapper, 1963; Stafford et al., 2004). The theory takes a user-level view and focusses on the individual audience member’s motivation for media consumption (Haridakis & Hanson, 2009; Rubin, 2009; Ruggiero, 2000; Stafford et al., 2004). Uses and Gratifications emphasizes the audience member’s potential for initiative and activity in their choice for media use. From passive collectives, new media users have become active individuals (Kang & Atkin, 1999; Stafford & Stafford, 1996; Rubin, 2009). People differ in their motives to use certain media, but the

other way around, “media may be differentiated by their ability to satisfy some needs better than others” as well (Lichtenstein & Rosenfeld, 1983, p. 97). The user is in the driver’s seat and consumers now have the power to regulate how the media are integrated in their lives (Jenkins, 2006; Sundar & Limperos, 2013). Audience activity is a central aspect of the uses and gratifications theory (Hawkins, Reynolds & Pingree, 1991; Katz et al., 1974; Levy & Windahl, 1984; Palmgreen, Wenner & Rosengren, 1985; Papacharissi & Rubin, 2000; Rubin, 2009). Several studies found that individuals purposefully select and consume media content to obtain gratifications (Kang & Atkin, 1999; Rubin, 1984, Rubin, 2009). Especially new media make individuals active users, as their input and responses are required for use (Blumler, 1979; Fredin & David, 1998; Stafford & Stafford, 1996). This is in line with the findings of Metzger and Flanagin (2002), whose results demonstrated that people were passive users of traditional media, and “new technologies are being used to achieve more active goals in comparison to traditional media” (p. 347). In the case of VOD, users can actively and purposefully browse the digital libraries and select from endless content options to watch. In contrast to traditional television consumption, where much less input was required since viewers could only choose to watch a selection of linear television channels. When considering uses and gratifications research altogether, the past decades pointed out that the gratifications for different kinds of media can be divided into three categories: (1) content gratifications, regarding the actual content, (2) process gratifications, referring to the use of the medium itself, and (3) social gratifications, regarding the social aspects of the content and media use (Cutler & Danowski, 1980; Katz et al., 1974; Stafford & Stafford, 1996; Stafford et al., 2004; Sundar & Limperos, 2013).

2.3 Content gratifications Content gratifications are obtained from the content that is carried out by a medium (Kaye, 1998; Stafford & Stafford, 1996; Stafford et al., 2004). Stafford et al. (2004) found that people reach out to the media in general in order to be entertained, and motion content - despite its carrier - is found to be watched mainly because viewers find it entertaining (Cha & Chan-Olmsted, 2012). Studies on user evaluations of the Internet found that people use the Internet and online media mostly for entertainment (D’Ambra & Rice 2001; Kayany & Yelsma 2000). Ferguson and Perse (2000) studied the Internet as a functional alternative to television, and found that the most frequent use of Internet was for entertainment purposes, and also fulfilled the need for users to unwind. Entertainment is, in many uses and gratifications studies, found to be one of the most dominant uses for the computer, and gratifications of the Internet (Cai, 2004; Cha & Chan-Olmsted, 2012; D’Ambra & Rice, 2001; Ferguson & Perse, 2000; Kaye, 1998; Papacharissi & Rubin, 2000; Ruggiero, 2000). Additionally, Haridakis and Hanson (2009) in their study for motives of watching YouTube video’s, found that these videos were mostly watched and shared because people were entertained by them. In addition to Internet, several television studies have also pointed out that viewers mainly watch television in order to be entertained (Cha & Chan-Olmsted, 2012; Kaye, 1998; Lichtenstein & Rosenfeld, 1983; Lin, 2001; Palmgreen & Rayburn, 1979; Rubin, 1981). People are amused, experience joy and excitement, and unwind from watching television, and entertainment can be considered as one of the main gratifications for television watching (Ferguson & Perse, 2000; Heeter & Greenberg, 1985; Kang & Atkin, 1999; Levy &

Windahl, 1984). Since VOD is an equivalent of television watching, it is most likely that these results will account for the use of VOD as well.Due to the large offerings and easy accessibility of VOD, this platform offers the viewer many options to have a pleasant rest and to be entertained. Furthermore, when looking at the VOD subscription networks, these networks are financially independent of advertisers because of the subscription fees charged to users. This results in enormous budgets, which are invested in the production of high quality content programs (Klarer, 2014). Millions are spent on star casts and profound characters, high level storylines are invented, and surprising plot twists are added, which results in high quality content and increasingly challenges the viewer on an intellectual level (Baas, 2015; Matrix, 2014). Together with the increasing popularity of VOD in the last couple of years, the content has become of superior quality. VOD is changing what, how, and when viewers watch content. Users watch more, and in larger doses at a time (Matrix, 2014; Pittman & Sheehan, 2015). Contrary to linear television broadcasting, VOD networks release all episodes of a series at once (Klarer, 2014). “When Netflix released all fifteen episodes of a new season of Arrested Development in the summer of 2013, reports showed that approximately 10% of viewers made it through the entire season within twenty-four hours” (Matrix, 2014, p. 119). The content in VOD libraries nowadays is of such good quality with such complex storylines, that it is crucial not to miss one episode as viewers will lack too much information to comprehend the next episode. Market research conducted by Netflix in 2014 found that 61% of its subscribers watched multiple episodes back-to-back regularly. The phenomenon of watching two or more episodes in a row is called binge watching (Jenner, 2014; Klarer,

2014; Pittman & Sheehan, 2015). Especially generation Y, young adults in the age between 18 and 29, are most likely to binge watch (Matrix, 2014). The outstanding aesthetic quality of the program content is addicting and was found to be an important motive for binge watching on Netflix (Pittman & Sheehan, 2015). Entertainment is found to be an important motive for the use of both Internet and television. In his examination of television viewing motives in 1981, Rubin already found a significant positive association between viewing for entertainment and the quantity of television viewing. As the big budgets used to go solely to Hollywood, they are invested in television production now too, resulting in very high quality content (Klarer, 2014). Binge watching is a popular phenomenon nowadays, because content is of such high quality and it is so easily accessible (Jenner, 2014; Matrix, 2014). Therefore, this research proposes the content as a new motive for the use of VOD. VOD is surpassing cable television, and traditional television viewing behavior has changed since we have access to VOD (Pittman & Sheehan, 2015). It is therefore hypothesized that users who reach out to VOD for entertainment purposes will have a higher frequency of VOD use. And furthermore, it is hypothesized that users who reach out to VOD for its content in a high degree will use VOD more often H1: Identification with entertainment motives for watching VOD will have a positive effect on the frequency of VOD usage. H2: Identification with content motives for watching VOD will have a positive effect on the frequency of VOD usage.

2.4 Process gratifications Rubin (1983) identified two types of television viewers; one watching for the content, and the other watching for the process. So besides the content, the user experience with the medium itself can also provide in specific gratifications. Lichtenstein and Rosenfeld (1983) were the first to propose that medium-specific gratifications are not predicted by innate needs or perceptions of use, but are triggered by features experienced while using a particular medium. These gratifications, obtained from the actual use of the medium, are referred to as process gratifications (Metzger & Flanagin, 2002; Stafford et al., 2004). Especially new media are suited to provide in process gratifications due to their convenient features (Sundar & Limperos, 2013). Viewers already became acquainted with displacing linear television viewing in the late 1970’s through VCR, and its technical benefits for time shifted viewing was found an important motive for using VCR (Henke & Donohue, 1989; Rubin & Rubin, 1989). Through interactivity features as time-shifted viewing, also know as zipping (Stafford & Stafford, 1996), consumers can actively and dynamically manage their own VOD consumption (Jenkins, 2006; Sundar & Limperos, 2013). Stafford and Stafford (1996) found that zipping was used for commercial avoidance, and to deal with interruptions when viewers are involved in a program. VOD tackles both commercials and interruptions, by offering a dynamic commercial-free platform, and the ability to shift in time (Sundar & Limperos, 2013). Not having to wait for the actual time slot is found to be one of the three main reasons for users to watch VOD, together with disliking commercial breaks, and the quality of the content (Matrix, 2014). The active new media consumer does not want to wait

another week for the next episode - or a few months for the next season after a bloodcurdling cliff hanger, and he or she does not want to be limited by linear television (Jenner, 2014, Matrix, 2014; Pittman & Sheehan, 2015). With VOD, users do not have to. VOD platforms release all episodes of a series at once in response to audience preferences (Klarer, 2014). The new media consumer, with generation Y in front, is active and wants to control their own viewing behavior (Jenkins, 2006; Sundar & Limperos, 2013). Traditional television schedules are bypassed, and viewers can fast forward through commercial breaks or even watch on advertisement free VOD networks. VOD’s big advantage over linear television is the user’s enormous amount of choice, since the content of VOD is not tied to a linear broadcasting schedule. VOD offers over hundreds of millions of hours of series, documentaries, television programs, and movies in any possible genre (HBO, 2015; Netflix, 2015a). In the VOD libraries, users can easily move through the medium towards the desired content (Sundar, 2008; Sundar & Limperos, 2013). Haridakis and Hanson (2009) found that this navigability was experienced as novel and very pleasant for YouTube users. As new features are found able to provide gratifications, mobility as a process gratification was first discovered by Wei and Lo (2006) to be obtained from the use of cell phones, and later on, found to be obtained from all mobile devices (Kim, Sundar & Park, 2011). Mobility has actually contributed to the development of VOD because it was experienced as such a strong and favourable feature (Henten & Tadayoni, 2008). VOD is no longer medium-specific and has no boundaries (Matrix, 2016). When watching content on VOD platforms, on the program’s own website, or when being subscribed to a VOD network, the only requirement for using it is an internet connection. This means that VOD

content can be enjoyed on the computer, the television, and on numerous mobile devices like laptops, smartphones, and tablets, at any possible location in or out of home - which makes VOD consumption highly mobile. Accordingly, it is proposed that the navigability and mobility will also contribute to the convenience of the process of using VOD. The distinction between content and process gratifications is also illustrated in the research of Rubin (1981, 1983), where a negative relation between content and process viewing motivations was found. Research respondents indicated that they often watch television just to pass time, despite of the program content. Additionally, more uses and gratifications studies have found that television is often viewed to pass time (Cha & Chan-Olmsted, 2012; Katz, Haas & Gurevitch, 1973; Lichtenstein & Rosenfeld, 1983; Palmgreen & Rayburn, 1979). In the search for understanding television and Internet use, both media have also been frequently found to be ideal for users who want to forget about school, work, problems, or other things (Rubin, 1981; Ferguson & Perse, 2000; Ruggiero, 2000). According to Ferguson and Perse (2000), the Internet is more and more competing with television as a way to pass time. The process of using VOD shares more resemblance with the Internet, as more user input is required than for traditional television viewing. Broadcasted television offers the viewer content immediately after the television has been switched on. The viewer can easily switch channels in search for other content, but the offerings are limited. When watching VOD, users can navigate through endless content offerings and make an active choice (Dwyer, 2010). However, the big advantage of new media is that their processes are equipped with intelligent technology. VOD is responsive to its users which makes the viewing experience very gratifying (Fredin & David, 1998; Sundar & Limperos, 2013). The devices on which users consume VOD are wirelessly

connected (Dwyer, 2010). VOD records users’ viewing behavior, and uses these data to make personalized suggestions for content that the user might want to watch, and VOD even automatically suggests to play the next episode of a series when the current one is finished. Altogether, the usage of VOD on itself and its associates processes are highly convenient, and perfectly suited for users as a means of diversion. Accordingly, it is hypothesized that high identification with motives for watching VOD for diversion, together with its convenient features will lead to more frequent use of VOD. H3: Identification with diversion motives for watching VOD will have a positive effect on the frequency of VOD usage. H4: Identification with convenience motives for watching VOD will have a positive effect on the frequency of VOD usage. 2.5 Social gratifications The social context of media use is also an important motive in uses and gratifications studies. Television and Internet are both a valuable source of social interaction. The uses and gratifications theory posits that media use is also largely shaped by their particular social characteristics (Haridakis & Hanson, 2009; Katz et al., 1973; Lull, 1980). Stafford et al. (2004) were the first to name social gratifications as a separate category, next to process gratifications and content gratifications. Television watching has traditionally been an activity central to family life (Lull, 1980). Watching television is often found to be a favourable social activity to do with family, friends, or others. It even helps to maintain friendship and family solidarity (Katz et

al., 1973; Klapper, 1963; Maccoby, 1951; Matrix, 2014; Rubin, 1981), as it can grant companionship when viewed together (Cha & Chan-Olmsted, 2012; Greenberg, 1974; Haridakis & Hanson, 2009; Ruggiero, 2000). Interestingly, Lull (1980) found that television is even able to function as actual companionship for people who are alone. The background noise of a television can make people feel less lonely, and function as substitute companionship - mostly by individuals who have limited opportunities for face to face social contacts in real life (Katz et al., 1974; Lull, 1980; Rubin, 1981). Besides the process of watching, the content on Internet and television also stimulates social interaction. It can provide a conversation topic, and people are found to consciously seek exposure to content in order to make small talk (Kaye, 1998; Levy & Windahl, 1984; Lull, 1980; Matrix, 2014; Palmgreen & Rayburn, 1979). Webster (1985) posited that living rooms are often the decor of a conversation about whether or not to watch television, and if so, what to watch? Additionally, Ferguson and Perse (2000) found that people browse the web in order to spark conversations, and research by Hampton and Wellman (2003) pointed out that Internet use had a positive relation to social interaction. So both television and Internet are a valuable source of social interaction. The media use itself and the stories, characters and themes of the available content can serve for social utility. Mass media are used by individuals to connect (Cai, 2004; Cha & Chan-Olmsted, 2012; Ferguson & Perse, 2000; Hawkins et al., 1991; Lull, 1980; Papacharissi & Rubin, 2000; Sundar & Limperos, 2013; Webster, 1985). As multiple people may not always be available to watch content on television in the linear broadcasting scheme together, VOD is perfectly suited for co-viewing because it can be viewed where- and whenever it suits users (Baas, 2015). This way, more people are able

to see content, and consequently more people will be able to participate in a social interaction. With friends and family, and also at work in so called ‘water-cooler conversations’, a term that describes “the phenomenon of people talking about TV shows at work after watching them the previous evening” (Einav & Carey, 2009, p.128). With roughly 140 million Netflix and HBO subscribers worldwide (Netflix, 2015b; Popper, 2015), it is very likely to be able to discuss the intense characters, storylines, and plot twists of content that is produced especially for VOD networks with others as well. Research even showed that the fear of missing out in social conversations is found to be related to watching Netflix, and young adults actually follow series to secure their social position within groups (Matrix, 2014). This leads to the fifth hypothesis: H5: Identification with social utility motives for watching VOD will have a positive effect on the frequency of VOD usage. 2.6 Media substitution In response to the new media user, VOD created a more appealing media form for watching programs with all its advantages compared to traditional television viewing. The media substitution theory states that if a new medium or technology is viewed as more desirable than the old one, users will reduce the time they spend on that old medium, or even replace it completely (Cai, 2004; Kang & Atkin, 1999; Kayany & Yelsma, 2000; Kaye & Johnson, 2003; Lin, 2001). The likelihood for replacement increases when the new medium is perceived as more convenient, less costly, or has superior content (Lin, 2001). Multiple studies have found results conform this theory. Almost all big media were negatively influenced since the television found its way into the living room. At the expense

of television viewing, less time was spent on reading newspapers, magazines, books, cinemas, and radio listening (Belson, 1958; Cai, 2004; Coffin, 1948; Koolstra & van der Voort, 1996; Maccoby, 1951). After its introduction, the computer also started to compete more and more with other media, including television. It is discovered that the computer is perceived as a functional equivalent of the television, and additionally satisfies the same needs. Consequently, computer owners spend less time watching television (Cai, 2004; Ebersole, 2000; Ferguson & Perse 2000; Rogers, 1985; Vitalari et al., 1985). Furthermore, the Internet is being used for the same motives as television, and caused people to spend more time on the web and less time on watching television (Cha, 2013; Kayany & Yelsma, 2000; Kaye, 1998; Kaye & Johnson, 2003; Morris & Ogan, 1996). The increasing popularity of complete television series sold in DVD boxes at the beginning of this century had a negative influence on television viewing as well (Jenner, 2014). But besides the television, also the printed newspapers and magazines have suffered from the use of online media (James, Wotring, and Forrest, 1995; Lin, 2004). Cai (2004) states that “the computer has evolved from a single-task machine to a multidimensional medium. It has taken on more and more of the functions that traditional media possess. Computer users can watch videos, listen to the radio, read newspapers and magazines, etc., all on the same medium” (p. 29). In several studies respondents have indicated to spend less time on watching content on the television, as they watch it online more and more (Cha, 2013; Kaye, 1998; Vitalari et al., 1985). Matrix (2014) reported that “nearly eight out of ten American adults who have Internet access watch television on demand” (p. 125). Linear television watching is thus suffering from VOD.

Delivery modes like radio, newspapers, and television are technologies that get improved over time, resulting in new media (Jenkins, 2006). VOD is such a new medium, with improved technology. And according to Bryant and Fondren (2009), “concerns about displacement effects become most evident at the introduction of a new medium” (p. 505). VOD enable users to make their own television schedule (Jenner, 2014), with the possibility to consume the content on multiple mobile devices, next to the television and computer. Features like these make VOD more convenient than linear television. Furthermore, most programs can be watched on demand for free, and viewers can get around the fee for programs that are not free of charge on demand, by recording them through their digital television set-top box. Then there also is the option to become a paid VOD network subscriber to eg. Netflix or HBO. The monthly fee for these networks is very low compared to the amount of content that is available, accumulating all pay-per-view episodes, and buying DVD-box sets of series. As discussed above in the content gratification section, the content is of superior quality nowadays, specially that on VOD subscription networks. With the increase in media came no increase in hours a day. So the time that people now spend on watching VOD must be at the expense of the use of other media. The Internet has already proven itself more efficient in terms of use, cost, and quality (Henten & Tadayoni, 2008). Because VOD is more convenient, less costly, and has superior content, and therefore meets all three requirements by Lin (2001), it is expected that VOD will lead to a reduction in use of linear television. H6: The frequency of VOD usage will have a negative relation to linear television viewing.

2.7 Mediation When looking at these hypotheses altogether, a mediation effect is revealed for the effect of VOD user motives on linear television watching via frequency of VOD usage. The content, process, and social uses and gratification for linear television have yet been studied and confirmed extensively (E.g. by Bantz, 1982; Conway & Rubin, 1991; Hawkins et al., 1991; Lull, 1980; Palmgreen & Rayburn, 1979; Rubin 1981, 1983), and therefore, this research does not try to do so as well. The main focus of this research is on discovering uses and gratification for VOD usage, and additionally to identify the effect of VOD usage on linear television viewing. This research proposes that content, process, and social gratifications, in line with television and Internet studies, also account for VOD usage, and that this form of media consumption leads to a decrease in use of linear television (Cha, 2013; Kaye, 1998; Vitalari et al., 1985). Therefore, a full mediation effect will be tested as an addition to the testing of the six proposed hypotheses. The first part of the mediation uses the motivations for using VOD as an independent variable, and frequency of VOD usage as a dependent variable, and this effect is predicted to be positive. Subsequently is the negative effect of frequency of VOD usage, now as an independent variable, on linear television viewing a dependent variable. 2.8 Theoretical model A total of six hypotheses were proposed in the sections above. Hypothesis 1 to 5 represent the direct and positive effect of motives for using VOD on the frequency of that usage. Following, the sixth hypothesis represents the effect of the frequency of VOD usage on the

model that illustrates the relations between the variables and the sign of their effects is displayed in figure 1.

Figure 1. Theoretical model.

3. Method. This chapter outlines the research methodology. The first section will outline the empirical strategy for this research. The second section will elaborate more on the characteristics of the survey respondents. Following, the conceptualization and operationalization of the used variables will be discussed. Finally, the procedures and statistical analysis that are conducted in order to test the hypotheses and answer the research questions will be described. 3.1 Empirical strategy Based on the uses and gratifications theory, and the media substitution theory, the purpose of this research was to test the effect of VOD user motives on the frequency of VOD usage, and additionally, the effect that VOD usage has on linear television use. As the previous chapter formed hypotheses to specify the directions between independent and depended variables, this research is explanatory in nature. Quantitative data are collected using a cross-sectional survey design, with a sample that contained VOD users. Surveys were used, as this is the most appropriate and common way to collect data for research on human behavior and uses and gratifications (Haridakis & Hanson, 2009; Saunders, Lewis & Thornhill, 2000). In this research, the survey was used to discover people’s motivations for using VOD, and to discover if their use of linear television had changed, and if so, if it increased or decreased. The surveys were conducted online, as this is a fast way to reach a large number of participants, highly efficient, and low in costs (Bernard, 2000; Katz et al.,

1974). Furthermore, online surveys lead to quantitative data that can be directly downloaded into SPSS in order to statistically test the hypotheses. Principal component analysis was first conducted, in order to group all motivational statements into the hypothesized motivational factors (Field, 2009). Following, correlation analyses were done to examine the strength of the association between the independent and dependent variables (Field, 2009). More important, regression analyses were done in order to actually determine the effect of the motivations on frequency of VOD usage, and the effect of VOD usage on use of linear television, respectively. 3.2 Respondents In order to gain the most appropriate insights in VOD user motivations, it was desired to gain as many respondents as possible that use VOD or otherwise are acquainted with VOD. An online survey was set up and distributed in several ways. First, the survey link was spread using snowball sampling through email and Facebook. Additionally, the researcher placed a post with the survey link on her LinkedIn page and in a special Dutch Facebook group for gaining respondents, called respondenten gezocht. The survey was started by 276 respondents, of which 253 fully completed it, resulting in a response rate of 92%. The final 253 participants had a close to equal gender distribution, as 52,2% was female and 47,8% was male. Only 1 respondent indicated to be under the age of 18, most respondents belonged to the group ’18-29’, also known was generation Y (70,6%), and the age groups ‘30-49’ and ‘50 or older’ both comprised 14.5% of the respondents. The majority of the respondents completed their education at university (31,9% bachelor,

41,7% master), 10,2% completed university of applied sciences, and the remaining 16,2% had completed vocational / technical school or a lower education. The most used forms of VOD, in descending order, were uitzending gemist (90,2%), Netflix (77%), RTL XL (53,6%), recording through digital television (27,7%), program’s own website (25,2%), HBO (21,7%), Videoland (11,1%), KIJK.NL (8,9%), and NL ziet (4,3%). 73,6% of the respondent indicated to own a Netflix account, but 77% of the respondents indicated to have used Netflix. This indicates that people without an own account can still use Netflix, either through a guest account, by using the account of a friend, or by co-viewing with someone that owns an account. The vast majority of the respondents (92,8%) indicated to have ever watched 2 or more episodes in a row, which is equivalent to binge watching (Pittman & Sheehan, 2015). When taking a closer look at these binge watchers, 73,9% of them belonged to generation Y, which is in accordance with Matrix (2014). A table with complete demographics of the respondents is presented in appendix 1. 3.3 Measurements The items that were used to construct the questionnaire were derived from leading authors in the field of uses and gratifications research. Solely items from Internet and television studies that were fully appropriate for this research were used, that had the same variable conceptualization and no highly divergent survey populations. The complete questionnaire of this study is included as appendix 2.

3.3.1 Video on Demand VOD in this research referred to all ways of television consumption other than linear television viewing. VOD thus includes watching play-back content recorded through digital television, catch-up sources, VOD subscription networks, the program’s own website, and other (inter)national streaming websites. 3.3.2 Entertainment The entertainment motives identified the degree to which participants were entertained by the content they watch, by experiencing joy, relaxation, and amusement. Motivations by Rubin (1981, 1983) for watching television were used to measure entertainment. As this research focusses on VOD, the word ‘television’ was replaced by ‘VOD’ for the statements of the current survey. The entertainment variable consisted of seven statements to which respondent indicated their level of agreement on an ordinal 5-point Likert scale with the following responses: (1) Disagree, (2) Somewhat disagree, (3) Neither agree nor disagree, (4) Somewhat agree, (5) Agree. The statements were “I watch VOD because it amuses me”, “I watch VOD because it is exciting”, “I watch VOD because it is enjoyable”, “I watch VOD because it is entertaining”, “I watch VOD because it relaxes me”, “I watch VOD because it allows me to unwind” and “I watch VOD because it is a pleasant rest”. 3.3.3 Content The specific characteristics of the content of programs comprised the content variable. Motivations by Rubin (1981, 1983) and Matrix (2014) were used, because a more recent article was desired as content has strongly developed the last couple of years. Again, the

word ‘television’ was changed into ‘VOD’ when necessary. Four statements, “I watch VOD when there is a specific program I want to see”, “I watch VOD because I like to see certain content”, “I watch VOD because of the high quality of the content”, and “I watch VOD because of the good storylines”, were the content motivations, to which respondents indicated their level of agreement on the same ordinal 5-point Likert scale. 3.3.4 Diversion The diversion variable contained statements referring to mental and physical escapism and ways to pass time. Motivations by Rubin (1981, 1983) were used, altered to VOD. The diversion motivation contained six statement on the same ordinal 5-point Likert scale, and they were “I watch VOD when I have nothing better to do”, “I watch VOD because it is just possible”, “I watch VOD because it gives me something to keep me occupied”, “I watch VOD so I can get away from what I am doing”, “I watch VOD so I can get away from the rest of the family, friends, or others”, and “I watch VOD so I can forget about school, work, or other things”. 3.3.5 Convenience Items from more recent studies were used for convenience, as this concept was found to be gratified only by new media. Six statements together formed the convenience variable, and items were taken from Haridakis and Hanson (2009), Matrix (2014), and Stafford and Stafford (1996). Where necessary, the words ‘television’ and ‘YouTube’ were changed to VOD. Convenience refers to features like zipping, availability, mobility and navigability that all contribute to the ease of use of VOD. Respondents indicated their level of agreement on

the same ordinal 5-point Likert scale to the following statements: “I watch VOD because it is easy to get to desired content”, “I watch VOD because I can watch it any time I want”, “I watch VOD because I can watch it anywhere I want on mobile devices”, “I watch VOD because there are no commercial breaks”, “I watch VOD because I can pause, resume, and replay at any moment I want”, and “I watch VOD because it is convenient”. 3.3.6 Social Utility The social utility variable identified the motives for VOD use in terms of social interaction and the watching of VOD as an activity on itself. Television viewing motivations by Rubin (1981, 1983) were used to measure the social utility variable, and the word ‘television’ was replaced by VOD for the current items. The five items were “I watch VOD because it is something to do when friends come over”, “I watch VOD so I can be with friends, family, and others”, “I watch VOD so I can talk to other people about the programs I have seen”, “I watch VOD when there is no one else to talk to or be with”, and “I watch VOD because it makes me feel less lonely”. Participants indicated their level of agreement on the same ordinal 5-point Likert scale. 3.3.7 Frequency of VOD usage The frequency of VOD usage variable in this research was operationalized on an ordinal scale ranging from ‘hardly ever’ (0) to ‘very frequently’ (100), presented to the respondents on a slider bar. In accordance with Pittman and Sheehan (2015), there is deliberately chosen not to use a minutes per day or minutes per week indication as this is more reliable for respondents in a longitudinal study. VOD usage can differ strongly per day or week and

so use in minutes can be quite abstract for respondents, were indicating frequency on an ordinal scale is more concrete (Rindfleisch, Malter, Ganesan & Moorman, 2008). 3.3.8 Linear television Respondents were asked to indicate the change in amount of use of linear television since they became acquainted with VOD, following the example of Kaye (1998). Respondents had the following response options on an ordinal 5-point Likert scale: “decreased” (1), “slightly decreased” (2), “about the same” (3), “slightly increased” (4), “increased”. It is deliberately chosen to use a closed and qualitative approach for this item as well, for the same reasons as frequency of VOD usage. 3.3.9 Control variables Age, gender, and educational level were added as control variables. Since Matrix (2014) posits that generation Y (age 18-29) is the most active VOD consumer, the answering options for “What is your age?” were divided on an ordinal scale into 1= “under 18”, 2= “18-29”, 3= “30-49”, and 4= “50 or older”. Educational level was asked through the question “What is the highest degree or level of school you have completed? (If currently enrolled, mark highest degree received)”. It was operationalized on an ordinal scale from 1 to 7 (1= No education, 2= Elementary school, 3= High school, 4= Vocational / technical school, 5= University of applied sciences, 6= University bachelor, 7 = University master). Gender was operationalized as a dichotomous variable (0= female).

3.3 Procedures The online survey was operated in Qualtrics, for which a personal account of the University of Amsterdam was used. The quantitative survey data retrieved from Qualtrics were imported into IBM’s statistical program SPSS, and several analyses were conducted to test the hypothesis and accordingly, answer the research questions. The procedures for both the survey and the statistical analysis will now be discussed. 3.3.1 Survey After a pre-test was conducted amongst 14 respondents who checked for any ambiguities, survey flow, and spelling errors, the final version resolved one spelling error as a result the pre-test. The final version of the online survey opened on February 16th _{2016 and was} closed 9 days later on February 25th_. The respondents first read a cover letter in which they were thanked for their participation, and the confidentiality of their answers was assured, followed by the conceptualization for VOD in this research. Respondents could continue to the actual questions when they had indicated to have read the explanation and understood how VOD was conceptualized. In order to gain more insight in the respondents, they first indicated if they were acquainted with VOD and if they had ever used VOD. Following, respondents were asked if they had personal access to Netflix, as this is the most popular paid VOD subscription network in the Netherlands (van Hoek, 2013), and what forms of VOD they had ever used. Then, respondents were asked to indicate their frequency of VOD usage. Subsequently, the respondents indicated their level of agreement to a total of 28 statements about reasons for using VOD. After these statements, the respondents were

asked if they had ever watched two or more episodes of a program in a row, on what devices they prefer to watch VOD, and how their amount of linear television viewing had changed since their use of VOD. At the end, some demographic questions were asked for the use of control variables. After finishing these last questions, a default message was shown in which respondents were thanked for their participation. A progress bar was presented at the bottom of every page to give respondents insight in their progress within the survey. All questions needed to be answered before being able to move on to the next page, and respondents could not move back in the survey in order to prevent them from altering earlier given answers. 3.3.2 Statistical analyses After the survey was closed, the data were downloaded from Qualtrics and entered into SPSS. First, the 23 respondents who did not completely finished the survey were deleted, in order to make statements about the same group of respondents, and to continue the statistical analysis with a clear dataset (Field, 2009). First, frequencies and cross tables were computed to gain detailed insights about the sample. In order to cluster the user motives for watching VOD into separate factors, Principal Component Analysis (PCA) with varimax rotation was conducted on all 28 items. This rotation method is chosen because it “maximizes the dispersion of loadings within factors” (Field, 2009, p. 644), and leads to very well interpretable factors. All items with loadings lower than .5 were excluded (Stevens, 2012). Only factors with an eigenvalue bigger than 1 that consisted of a minimum of 3 items were retained (Field, 2009; Pittman & Sheehan, 2015). Furthermore, reliabilities of the factors were computed, and summated

factor scales were made in order to test the hypotheses. Summated scales are more easy to interpret than factor scores, and they were made by summing all items within the factors and dividing them by the number of items within that factor. This could be done because all items were measured on the same 5-point Likert scale (Field, 2009). Subsequently, Pearson correlations were computed to examine the strength of the relations between each reliable factor and the frequency of VOD usage for the first five hypotheses, and additionally between the frequency of VOD usage and linear television use. After the correlation analysis, ordinary least square linear regression analyses were conducted. In the same order, all variables, and now also including the control variables, were entered separately into the regression analyses in order to examine the prediction values of the dependent variable from the independent variables (Field, 2009). Only hypotheses with a confidence level of at least 95% were accepted. The regression model is a straight line representing the ratio between the dependent and independent variables. It explains how the value of the dependent variable (on the Y-axis) changes, when the value of the independent variable moves 1 step on the X-axis, whilst control variables are held fixed (Field, 2009). For the first research question, this means that the hypotheses propose an increase in frequency of VOD usage, when respondents’ identification with the motivational variables increases with 1, on the 5-point Likert scale. For the second research question, this means a decrease in use of linear television when the frequency of VOD usage increases with 1, on an ordinal scale from 0-100. An additional analysis was done for the mediation effect. Only the factors of the accepted hypotheses were entered into linear regression analysis (Field, 2009). These

factor scales were entered as independent variables and linear television viewing as dependent variable. Additionally, frequency of VOD usage was entered as a control variable to see if the effect of user motives on linear television use did indeed decrease, according to the established method of Baron and Kenny (1986).

4. Results. The results of the statistical analyses that were conducted in SPSS will be presented in this chapter. First, the results of PCA will be outlined. Subsequently, to answer research question 1, the results of the correlation and regression analyses for the first five hypotheses will be presented. Following, the results of the correlation and regression analyses for the last hypothesis will be presented, for the purpose of answering the second research question. Furthermore, the mediational effect of frequency of VOD usage will be presented as an additional analysis, since this mediation was not the main focus of this research, but emerged from the six proposed hypothesis. 4.1 Principal Component Analysis The first research question asked about motives for using VOD, and its effects on the frequency of VOD usage. PCA with varimax rotation was conducted in order to factor all 28 motivational statements for using VOD. After this first analysis, the item “I watch VOD because it is just possible” was the only item with a factor loading below .5, and so it was excluded. A second round of PCA with varimax rotation was conducted for the remaining 27 motivational statements, and resulted in factor loadings that were all above .5 divided over six factors with an eigenvalue greater than 1. The sixth factor, however, comprised of only two statements that were originally attributed to content (“I watch VOD because of the high quality of the content”, and “I watch VOD because of the good storylines”), and so this factor was dropped because it did not meet the requirement for a factor to consist of at least three statements (Field, 2009). The final result of the PCA with varimax rotation was a

deviation of 25 of the original 28 motivational statements over five factors, all with a loading higher than .5 (not rounded off), and together they explained 59.67% of the variance. The five factors are entertainment, content availability, diversion, convenience, and social utility. The Cronbach’s alpha of all factors was computed with a reliability test. Besides the reliabilities, the output of this test also showed the Cronbach’s alpha for the factor if any item would be deleted, and thus if the measurement of the factors could be improved (Field, 2009). Very small changes in the reliability scores occurred for three factors. The results showed that the Cronbach’s alpha for social utility was .739. When deleting the item “I watch VOD so I can talk to other people about programs I have seen”, the Cronbach’s alpha would increase to .744. The consequence of deleting this item would be that the social utility factor as a whole had to be dropped, as only two items would remain to load the factor. Since the increase in Cronbach’s alpha from .739 to .744 is so extremely small, it is judged as negligible, and the maintenance of factor 3 as a whole is judged as more important. For the same reason, it is decided to judge the minor increase of .071 in the Cronbach’s alpha of factor 4, convenience, when “I watch VOD because I watch it any where I want on mobile devices” would be deleted as negligible, since this factor also comprises of only three items. The third factor that reacted to the removal of one item was content availability (factor 5). This factor contains 5 items, which means that a sufficient number of items will remain if one of them would be deleted. The Cronbach’s alpha for social utility was .644 and would increase with .005 (α = .649) if the item “I watch VOD when there is a specific program I want to see” would be deleted. However, this increase is

so extremely small and is therefore also judged as negligible, and all 5 items were retained. The results of the PCA and reliability test are presented in table 1.

The first factor, entertainment, comprised of seven items about how people are

amused, excited, enjoyed, entertained, unwind, and relaxed by the content of VOD. Factor 2, diversion, was the factor that lost one item due its loading below .5 after the first round of PCA. Additionally, two items that were originally assigned to social utility pointed out to load on diversion (“I watch VOD when there is no one else to talk to”, and “I watch VOD because it makes me feel less lonely”). The diversion factor (factor 2) includes statements that regard using VOD for passing free time, getting away from people or tasks, fighting boredom, and to clear someone’s head from any thoughts or problems. Social utility is the third factor, and comprised of three out of the five items that were originally attributed, since two items were lost to factor 2 (diversion), as described above. The items of the social utility factor refer to social interaction in terms of companionship, social activity, and social talk. The fourth factor, convenience, comprises of three items about the process of mobility, navigability and the absence of commercial breaks. The other three items that were originally attributed to convenience were lost to the fifth factor; content availability. This last factor was originally hypothesized as content, but content availability was found a more appropriate title after the PCA because the factor also includes content statements regarding the possibility to bypass the linear broadcasting schedule, and the ease of accessing content. 4.2 Analyses for research question 1 All 5 factors were transformed into factor scales in SPSS, in order to test the Pearson correlation amongst the VOD use motives (entertainment, content availability, diversion, convenience, and social utility) and their correlation with frequency of VOD usage. Descriptive statistics of the scales are presented in table 2.

Table 2. Descriptive statistics. The means and standard deviations of the factor scales are added in the correlation matrix. When looking at the means, it is striking that the frequency of VOD usage (x̅ = 70.70) lies quite high amongst the survey respondents of this study. This implies that respondents in the sample of this study rated themselves as frequent VOD users. Furthermore, the respondents in the sample indicated to identify themselves the most with content availability (x̅ = 4.45), entertainment (x̅ = 4.11), and convenience (x̅ = 4.04) motives. The respondents could slightly identify with social utility motives (x̅ = 2.84), and was quite neutral in their identification with motives concerning diversion (x̅ = 2.54). Correlations are computed before conducting the regression. All independent variables are correlated to the dependent variable positively and reliable. This means that there is a relationship between every proposed independent variable and the dependent variable, as such that an increase in identification with all proposed motives is related to an increase in frequency of VOD usage, although the Pearson correlation coefficients are not very strong.

The entertainment motive does have a positive and significant relation with frequency of VOD usage (r = .49, p < .01). There is a weak correlation between content availability and frequency of VOD usage (r = .34, p < .01), yet, this relation is positive and significant. The correlation between diversion and frequency of VOD usage is also weak (r = .34, p < .01), yet it represents a positive relation with high significance. The same relation is found for hypothesis 4 (r = .20, p < .01). When looking at the correlation between convenience and social utility, it is shown that these factor scales have a perfect positive relationship (r = 1.00, p < .01). This means that an increase in identification with convenience motives for using VOD leads to a mutual increase of identification with social utility motives for using VOD. Therefore, the positive relation between social utility and frequency of VOD usage (r = .20, p < .01) is identical to that of convenience. A correlation matrix is displayed in table 3. As correlation analysis alone is not sufficient to establish an effect and actually test the hypotheses, regression analyses are done in order to accept or reject the proposed Table 3. Correlation matrix.

hypotheses appropriately. Again, the factor scales are used for the regression analysis, since the distribution of the factor scales is better adapted to linear regression analysis than factor scores (Field, 2009). A linear regression analysis for all hypotheses is conducted to test the actual effect of every VOD usage motive on the frequency of VOD usage. The control variables gender, age, and level of education were entered in the first step (model 1) of the regression, as this study is not particularly interested in these variables, but does like to check whether they might have any influence on frequency of VOD usage. The motivational factors were entered as independent variables in the second step (model 2) of the regression. A complete overview of all regression outputs is enclosed in appendix 3. The regression outputs pointed out that model 2 was significantly better than model 1 (F Change = 15.334 (4,227); p < .000). This means that the model with the independent variable explains the effect on the dependent variable is better than the model where only the control variables predict the dependent variable. Additionally, the ANOVA output shows if the tested model as a whole is significant. For this test, the model was statistically significant F (7, 227) = 15.477; p < .000 and explained 32,3% of the variance in frequency of VOD usage (R² = .323, Adjusted R² = .302). Accordingly, the results of the regression are judged as reliable. A summary of the regression analysis is displayed in table 4. 4.2.1 Entertainment and higher frequency of VOD usage The first hypothesis proposed a positive effect of entertainment motives (independent variable) on frequency of VOD usage (dependent variable). The coefficients output explains the actual effects, and showed that identification with entertainment motives indeed lead to a higher frequency of VOD usage (b = 11.130, p < .000). In order to appropriately