The Effects of VR on Consumer Attitudes: Lessons from Implementing VR Technologies in Real Estate Marketing

(1)

The Effects of VR on Consumer Attitudes -

Lessons from Implementing VR

Technologies in Real Estate Marketing

Master’s Thesis 30 credits

Department of Business Studies Uppsala University

Spring Semester of 2020

Date of Submission: 2020-06-03

Filip Gunnarsson

François Mirza Ali Khan

Supervisor: Jukka Hohenthal

(2)

Abstract

Virtual Reality (VR) has gained tremendous interest within the last decades and is rapidly becoming a disruptive marketing tool across all industries. Predictions show that the VR market is expected to reach a staggering US$120 billion by 2026 (Fortune Business Insights, 2019).

While previous research has investigated the impact of VR on consumer attitudes, the connection between its immersive characteristic, i.e. telepresence (Steuer, 1992), and the complexity of consumer attitudes (Eagly & Chaiken, 1993), is still relatively unknown. Hence, this paper addresses the discrepancy in extant literature by applying a qualitative research approach. We conduct in-depth interviews using a quasi-experiment design, displaying an apartment through 2D (pictures) and followingly 3D stimuli (VR). The increased telepresence from VR and its impact on participants’ responses towards the apartment is thereby captured.

Results show that both technological variables of telepresence (vividness & interactivity) are positively influential in the three-dimensional spectrum of consumer responses (affective, cognitive & conative). Moreover, this study reveals that telepresence contributes to enhanced product knowledge, produces a hedonic experience and aids in decision-making. Conclusively, managerial recommendations are provided, detailing aspects to consider before implementing such immersive technologies.

Keywords: Virtual Reality, Telepresence, Interactivity, Vividness, Consumer Attitudes, Tripartite Model, Affective, Cognitive, Conative, Real Estate

(3)

Acknowledgements

The completion of this undertaking would have not been possible without the support of the following individuals, for whom we are extremely grateful.

First, we would like to thank our supervisor Jukka Hohenthal for guiding and helping us through the entire process. He provided invaluable insights which made our thesis much more focused, well-formulated and structured. We also wish to thank the professors at Uppsala University FEK that taught us the foundational knowledge and skills needed to write this thesis.

Our gratitude is extended to our peers in the seminar group for interesting discussions and giving us thoughtful feedback on every aspect of our paper. Furthermore, to all respondents who made this thesis possible by sharing their thoughts and feelings with us. Their engaging participation set the foundation for new knowledge creation and extending the frontier of marketing academia.

We also want to voice our sincere appreciation to Sara Brandt and Caroline Hedlund of HSB.

Your insights into the real estate market and how VR is implemented in practice, paved the way for our whole paper. In addition, we received high-quality VR content that was paramount in the successful data collection. Immense praise also goes out to Emil Rydberg and Thea Hamrén at First Lady Agency, who kindly lent us their Oculus Quest for our interviews.

Finally, we would like to express our deepest gratitude to our family and friends who have greatly supported us throughout this thesis and our entire education.

Uppsala University, June 3, 2020

Filip Gunnarsson & François Mirza Ali Khan

(4)

1. Introduction

1.1 Background

The developments within Virtual Reality (VR) are rapidly becoming a truly disruptive force for marketers. What until recently was an expensive and unwieldy fringe gadget is now considered by many to be the next technological market breakthrough (McKone, Haslehurst &

Steingoltz, 2016; Capgemini, 2018). Possibilities arising from these technologies are already abundant, promising to create a whole new generation of powerful marketing tools. Huge leaps have been made during the last decade in terms of accessibility and processing power, exemplified by VR headsets such as Oculus (Facebook), HTC Vive and Samsung Odyssey.

According to Fortune Business Insights (2019), the VR market currently represents US$7.3 billion, but is expected to skyrocket to US$120 billion in 2026. While the gaming industry is the major utilizer of VR technologies (40% in 2018, ibid.), other industry verticals such as healthcare, education, retail and even manufacturing are taking advantage of its potential.

These industries are progressively boosting their VR market share and substantial investments are made in R&D for VR adoption (Fortune Business, 2019). In Capgemini (2018), we see that key industrial actors such as Ford, BMW and Airbus are already utilizing the capabilities of VR to drastically improve efficiency and work conditions.

One particular area where VR is predicted to gain tremendous momentum is within the real estate sector. From enhancing architectural processes to marketing apartments and buildings to customers, VR is set to revolutionize how we design spaces and buy property (RICS, 2019). In Sweden, several real estate actors have already enabled VR apartment viewings for new productions, allowing consumers to view and interact with virtual 360° representations of projects. While the creation of these virtual spaces is not cheap, market mechanisms are sure to drive the prices down significantly over the coming years. In short, the signs are pointing towards a symbiotic convergence of VR market penetration, coupled with progress in rendering technology and lower costs of production (Bellini et al., 2016; RICS, 2019).

But despite its promising outlook, not much literature has been devoted to investigating the empowering impact of VR for marketing practice (Boyd & Koles, 2019). Presently, main themes within VR literature mostly include its implementation within the gaming and entertainment industry (Ermi & Mäyrä, 2005). According to Xue, Parker and McCormik (2018), a few studies have also examined VR in contexts of online retailing and shopping, as

(7)

2

well as tourism (e.g. Yung & Khoo-Lattimore, 2017). Notable research by Steuer (1992; 1995), Biocca (1992), Pine & Gilmore (1998) and Klein (2003) were also conducted during the 90’s and early 2000’s to conceptualize VR into a theoretical framework – telepresence – and to investigate its impact on constructs such as brand attitudes, offer perception and product knowledge. The concept of telepresence has arguably paved the way for the entire academic subject. However, much of these early studies utilized simple 3D models, and not the complex technology of Head Mounted Display (HMD) existing today.

1.2 Problematization

Since then, VR research has been stagnating, continuously revolving around similar subjects and utilizing similar methods. A major discrepancy concerns the influence of telepresence on consumer attitudes. Although some research has been devoted to this topic, much remains unclear. Several studies have shown positive correlations between enhanced telepresence and increased consumer responses (e.g. Grudzewski et al., 2018; Klein, 2003; Van Kerrebroeck, Brengman & Willems, 2017), but the literature is still lacking qualitative data. As a result, the underlying factors bonding consumer preferences and telepresence are still largely unknown – a gap that needs to be addressed for both academia and practice to progress.

In addition, a widespread ignorance exists towards the complex attitude concept in VR literature. Many studies on the subject simply neglect to define attitude in any real sense, often referring to the construct as if the term “attitude” is self-explanatory. For example, Klein (2003) examines the relationship between vividness and attitude towards a product but does not specify how attitude is to be understood in her paper. Similarly, Grudzewski et al. (2018) investigates telepresence and its impact on the perceived attractiveness of an offer, giving no theoretical disclosure to what “perceived attractiveness” actually entails. This neglection is harmful for both marketing academia and practice, since it is often concluded that telepresence influences attitudes positively, without fully understanding why. We therefore argue for the necessity of sealing the hole in present academic VR marketing literature, and attempt to answer the following research question:

“How does the experience of VR technology in marketing influence consumer attitudes?”

(8)

3

1.3 Purpose of the Study

The purpose is thus to study how enhanced telepresence affects consumer attitudes through an in-depth perspective of both frameworks. Our research intends to provide significant contributions to the VR literature and the use of VR in marketing activities. This will be achieved by presenting a modernized extension of the theoretical framework of telepresence (Steuer, 1992; Klein, 2003) with the study of consumer attitudes (Eagly & Chaiken, 1993;

2007). The main appeal of telepresence is its experience-oriented nature, which makes it applicable to practically every communication medium – from a book or a picture, to complex virtual worlds in VR or augmented reality (AR) (Steuer, 1992). As a result, the paper also contributes to research on immersive marketing tools in general and their applicability to consumer marketing. AR, mixed-reality and interactive videos are all examples of technologies to which our research will extend. Finally, in hopes of providing relevant managerial implications, this paper attempts to help executives build profitable VR marketing strategies directed towards the influence of consumer attitudes.

Real estate industry is selected as an illustrative example of how VR is implemented to enhance marketing efforts. The industry is arguably a ‘ground zero’ for the technology, where its true potential as a marketing tool is established (Bellini et al., 2016; RICS, 2019). From a consumer perspective, it also offers a completely new customer experience with a global reach, making it possible to ‘visit’ properties from anywhere in the world, at any time. Moreover, apartments are products of spatial nature, and their implementation as attitude objects in previous VR research has shown promising results (see Grudzewski et al., 2018). Therefore, to present valuable insights from the implementation of VR in real estate, we draw support from HSB, a Swedish real estate company which is already utilizing VR technology as an integral part of their marketing strategy. HSB is a cooperative organization that plans, finances, builds and manages real estate throughout Sweden – currently having 25,000 rental apartments and 340,000 tenant-owner apartments (HSB, 2020). Recently, HSB has augmented their existing multi-channel marketing approach of new estate projects with 360° virtual experiences. As one of Sweden’s largest actors in real estate, HSB serves as a perfect example to respond to our problematization.

(9)

4

1.4 Thesis Disposition

To methodically answer our posed research question, the two chosen theoretical constructs of consumer attitudes and telepresence are firstly defined and elucidated. We summarize crucial insights and identify shortcomings in the extant literature to provide a sound foundation of theoretical understanding. After, we describe our chosen methodological approach. A series of in-depth consumer interviews are conducted, aided by HSB’s marketing material. Thereby, we gather rich data that captures the small but very important details and nuances in consumer attitudes. The data is then examined through the lens of our conceptual framework and our findings are presented accordingly. Following, the most important insights are brought together, compared to extant theory and discussed in detail. Our conceptual model is revised, and new insights are integrated into the framework. Lastly, we conclude our study and delineate managerial implications, as well as contributions and suggestions for future research.

2. Literature Review

2.1 Consumer Attitudes

Attitudes and their implications on behaviour have fascinated scholars and practitioners for a long time, with extensive research already being conducted in early-20^th century (e.g.

Thurstone, 1928; Thurstone & Chave, 1929; La Piere, 1934). Drawing its roots from social psychology (Olson & Zanna, 1993), the concept has been studied in both psychological and marketing contexts because of its predictionary role to human (i.e. consumer) behaviour (Argyriou & Melewar, 2011). Substantial research has been made on the nature of attitudes and their implications, relating to areas such as knowledge structures and memory, level of elaboration, behaviour prediction and attitude change (Fabrigar, Macdonald & Wegner. 2005;

Olson & Zanna, 1993).¹ In the following chapter however, we focus our review on the attitude definition and how it can be conceptualized into evaluative responses.

1 For a comprehensive review on a ttitude resea rch, we refer to the work of Fa briga r, Ma cdona ld & Wegner (2005), Olson & Za nna (1993) a nd Argyriou & Melewa r (2011).

(10)

5

2.1.1 Attitude Definition

As with many abstract constructs within social sciences, the definition of attitude is elusive.

Despite the considerable amount of research conducted within the field, no universally accepted definition of attitude is present in academia (Greenwald, 1968; Olsen & Zanna, 1993).

To clarify our conceptual model, we provide a brief overview to existing definitions prevalent in the attitude literature – laying the groundwork for how attitude is seen in this paper.

Beginning with the definition from Greenwald (1989), who regards attitude as “the affect associated with a mental object”, which can be either negatively or positively constructed.

Emphasis is on a predisposition of affect, which shifts focus away from active evaluation. Fazio (1989; 1995), on the other hand, advocates for viewing attitudes as a semantic two-node network model of object-evaluation associations. Here, one node represents the object, the other node the evaluation schematic, and the link between these two nodes determines the association strength. Coherent to the more traditional view of attitudes, from scholars such as Fishbein & Ajzen (1975), Zanna & Rempel (1988) propose that attitudes should be seen as the evaluative response itself. This definition is subscribed to by Kruglanski & Stroebe (2005), who see attitudes as evaluative judgements towards an object in terms of its positive or negative degree.

Solomon (2013, p.273) confers that “an attitude is a lasting, general evaluation of people (including oneself), objects, advertisements, or issues.” Special attention should be given to the aspect of ‘lasting, general evaluation’. This specification is crucial, because it pinpoints attitudes as overarching mental structures that often persists over longer periods of time. Two implications spring from this assessment, (1) attitudes are difficult to influence; and (2) when established, they are often stable. While this may be true for some attitudes, it is not true for all. Research show that attitudes follow a hierarchical structure of commitment, where some are more important and firmly held than others (Eagly & Chaiken, 2007; Solomon, 2013).

While the strength/weakness aspect of attitude is not relevant for our research, it is necessary to reject the idea of attitude as an always stable construct.

Subscribing to the view of attitudes as evaluative structures is the definition first presented by Eagly & Chaiken in 1993, and later revisited in 2007. For them, attitude is “a psychological tendency that is expressed by evaluating a particular entity with some degree of favor or disfavor” (2007, p. 582). The formulation is made to encompass the overarching concept of attitude, which for Eagly & Chaiken (2007) consists of the key elements: tendency, entity and

(11)

6

evaluation. Within this framework, evaluation includes all classes of evaluative responding, which are beliefs and thoughts, feelings and emotions, and intentions and overt behaviour (Eagly & Chaiken, 2007, p.583). These classes are drawn from the widely used tripartite model of attitudes, which categorizes attitude responses in the three dimensions of: affect, cognition and conation (behaviour) (Fishbein, 1989; Solomon, 2013). The evaluative responding is directed towards ‘attitude objects’ – entities that can be liked or disliked, ranging from direct objects such as a person (Vladimir Putin) or a brand (Amazon), to more abstract concepts such as belief sets (liberalism, conservatism, Christianity, etc.). A key element in the definition by Eagly & Chaiken (2007) is ‘tendency’: “an individual’s past experiences establish a tendency to respond with some degree of positivity or negativity to an attitude object” (p. 585). By acknowledging that previous experiences and existing attitudes can affect an individual’s response towards an object, the authors confirm that attitude doesn’t necessarily translate into overt behaviour (Fishbein & Ajzen, 1977). Rather, it asserts that attitude is a predictor for behaviour.

Despite the discrepancies within the attitude research field , the fundamental idea of attitude as a subjective structure of object evaluation permeates and persists (Fabrigar, Macdonald &

Wegner, 2005). In fact, Bagozzi (1999) observes that some social psychologists even regard attitudes as isomorphic with evaluation and use the terms interchangeably (e.g. Fishbein &

Aizen, 1975). For our purpose, we lean on the definition presented by Eagly & Chaiken (1993;

2007), stemming from two key factors. Firstly, we confer that the definition is broad enough to encompass most relevant aspects of attitude, whilst steering clear of similar concepts, such as emotions and moods (Bagozzi, 1999). The idea of attitude as evaluative responses per se (Kruglanski & Stroebe, 2005) is rejected. Instead, these responses are manifestations of an individual’s attitude towards an object. The element of tendency also illustrates that attitudes can, to an extent, predict behaviour, and that its possible for attitudes to change as a result of new experiences. Second, the definition is written against the structure of the ‘tripartite model of attitudes’ (Fishbein, 1989) which will be elucidated in the following segment. This ensures that the definition can be clearly modelled and then integrated with the telepresence framework.

2.1.2 The Tripartite Model of Attitude

With the basic idea of attitude in place, one fundamental question to address is how attitudes should be modelled. With the basic idea of attitude in place, one fundamental question is how

(12)

7

attitudes should be modelled. Already in the pioneering work by Thurstone (1928), it was concluded that attitudes seem to represent complex constructs with multifaceted properties.

Therefore, Thurstone argued that it is inappropriate to measure attitude with a single index (Fabrigar, Macdonald & Wegner, 2005). Notwithstanding of Thurstone’s early postulation that attitude is multifaceted, several studies in the mid-20^th century suggested that attitude is actually a single-dimensional construct. Fishbein (1967) for example, proposed that attitude is best characterized as a single dimension of affect for or against an object.

Ajzen (1989) highlights that attitude is a hypothetical construct, thus being inaccessible to direct observation. Instead, attitude must be inferred from measurable responses, which are either negatively or positively directed towards the attitude object. According to Ajzen (1989), there are no limitations to the kind of responses that might arise. Therefore, categorization of attitude responses into subgroups are often made to simplify the analysis process. The most prevalent of these categorizations, and the one utilized in this paper, is a three-dimensional response framework: affective, cognitive, and conative (behavioural). The so-called ‘tripartite theory’ has enjoyed an extensive history with much attention diverted to it (Ajzen, 1989;

Rosenberg & Hovland, 1960; Smith, 1947). The earlier conceptualizations of the tripartite model tended to consider response as direct displays of attitude (Smith, 1947). Later studies would instead view attitude as “a general evaluative summary of the information derived from these [three] bases” (Fabrigar, Macdonald & Wegner. 2005, p.80). Thus, a general attitude does not consist of affect, cognitive and conative responses, but is instead a separate construct which may be inferred from all types of evaluative responses (Crites, Fabrigar & Petty, 1994).

In addition, it is common to further separate between verbal and non-verbal responses within the three subgroups (Ajzen, 1989). For our research however, only verbal responses will be considered, as our ability to measure and analyse nonverbal responses is limited in both resources and time. Cognitive-verbal responses reflect beliefs about the attitude object, centring on characteristics and attributes. Affective-verbal responses are expressions of feelings towards attitude objects, such as excitement, sadness, frustration or joy. Conative-verbal responses are predispositions, intentions and commitments of actions with respect of the attitude object.

These indicate how a person does, would , or plans to act towards an entity (Ajzen, 1989; Park, Stoel & Lennon, 2008). In the tripartite model, the three classes of evaluative responses are considered equivalent in their ability to reflect an individual’s attitude (Ajzen, 1989; Fabrigar, Macdonald & Wegner, 2005).

(13)

8

Eagly & Chaiken (2007) agree with the distinction between attitudes and evaluative responses.

As previously illuminated, they define attitude as an inner tendency that influences the response towards a particular entity with some degree of favour or disfavour. Along a continuum, we can see attitude as an already existing construct from previous experiences. When a person encounters an object, the attitude influences the favourability of responses that the person will show (see Figure-1). For this thesis, we outline our attitude model based on the definition from Eagly & Chaiken (1993; 2007) and a contemporary three-dimensional model of evaluative responses. Because we study the effect of telepresence on consumer attitudes, a framework which focuses on the evaluative responses themselves is preferable, instead of a model which has behaviour intention or overt behaviour as the focal component (e.g. Theory of Reasoned Action, Fishbein & Ajzen, 1975; Theory of Planned Behaviour, Ajzen, 1985).

Figure 1. Conceptual model of attitude and evaluative responses.

The model presented above displays how tendency influences the evaluative responses towards an attitude object. It highlights attitude as a separate construct – distinct from the evaluative responses – which might reflect aspects of the general attitude; but does not determine it. The framework is founded on the tripartite model (as seen in Rosenberg & Hovland, 1960) and advanced with the more contemporary attitude definition presented by Eagly & Chaiken (1993;

2007).

(14)

9

2.2 Virtual Reality and the Concept of Telepresence 2.2.1 Defining Virtual Reality

Virtual Reality and immersive tools alike are still relatively novel elements in marketing theory, although the concept arguably appeared in the 1960s (Mazuryk & Gervautz, 1996). In literature, there is a relatively vast array of heterogeneous and diversified definitions of VR.

The founding research conducted on VR adopts a tech-centric perception of the phenomenon, and most of the definitions describe VR as specific technological systems requiring hardware equipment such as headsets, computers and gloves (Gold, 1993). For example, Greenbaum (1992) viewed VR as “an alternative world filled with computer-generated images that respond to human movements. These simulated environments are usually visited with the aid of an expensive data suit which features stereophonic video googles and fiber-optic data gloves” (p. 58). Similarly, McCloy & Stone (2001) refer to VR as “a collection of technologies that allow people to interact efficiently with 3D computerised databases in real time using their natural senses and skills” (p. 912). Even in the recent literature, researchers continue to describe VR as a computer-based technology delivering a sense of immersion within a particular environment. For instance, Lopreiato et al. (2016) define VR as “a computer- generated three-dimensional environment that gives an immersion effect” (p. 40).

Overall, most scholars agree that this collection of technologies transfer the user into an immersive, virtual or simulated environment, in which the consumer can interact with his surroundings to a pre-determined degree. In this sense, Riva (2002) provides a slightly more complete definition of VR, specifying that this technology is “an advanced form of human–

computer interface that allows the user to interact with and become immersed in a computer- generated environment in a naturalistic fashion” (p. 230).

However, in his seminal paper, Steuer (1992) asserts that a device-driven typology of VR is unacceptable for researchers, software developers, policymakers and media consumers. The fallacy in such a definition lies in its inability to withstand new innovations and its unsuitability to act as a conceptual framework (Steuer, 1992). Within a device-driven typology, a system either is VR, or it is not – leaving little room for comparisons between different systems.

Because of these fallacies, is it postulated that VR must instead be defined within the context of human experience (Pine & Gilmore, 1998). The technological gadgets are not at the centre of attention, but the sensations that these technologies mediate. For Steuer (1992), the key to

(15)

10

accurately defining VR is by utilizing the concept of presence, which can be described as ‘the feeling of being somewhere else’.

This notion is echoed by Biocca (1992), who describes VR as an “environment created by a computer or other media, an environment in which the user feels present” (pp. 5-6).

Correspondently, Li, Daugherty & Biocca (2001) suggest that VR is best theoretically conceptualized within the context of consumption experience,, defining VR as “a vivid, involving, active, and affective psychological state consumers encounter when interacting with 3-D products in a computer-mediated environment” (p. 4). In this form, VR is only achieved when telepresence is mediated by 3D products, and not 2D stimuli. The concept of telepresence is then portrayed as a fundamental characteristic of VR and help provide another description without referring to any hardware systems. Indeed, as Steuer (1992) details it, “a virtual reality is defined as a real or simulated environment in which a perceiver experiences telepresence ” (p. 7). In that sense, the current phenomenon of VR through HMD is simply one way that telepresence can be experienced. It is merely a part on the entire spectrum of telepresence- inducing mediums. Thus, the particularity of the definition of VR as telepresence, as opposed to computer-based definitions, is that it refers to an experience rather than a hardware system.

It can be applied to past, present and future media technologies, rather than a specific technological moment (Steuer, 1992).

2.2.2 The Concept of Telepresence

Since Steuer’s paper in 1992, the concept of telepresence has been extensively scrutinized in VR literature and is often presented as a key element to describe the sense of presence in a virtual or mediated environment (Steuer, 1992; Sheridan, 1992; Klein, 2003; Kim and Biocca, 1997; Nowak and Biocca 2003; Jung, 2011). Steuer (1992) construed VR using the approach of telepresence and defined the latter as “the experience of presence in an environment by means of a communication medium” (p. 76). As opposed to the term ‘presence’ (the sense of being in a physical or natural environment), the concept of ‘telepresence’ entails the notion of mediated perception of being in that environment. In other words, the perception of presence in a natural environment is obviously taken for granted as no medium forces this perception.

For telepresence, the feeling of presence is mediated by technology and one is therefore compelled to feel presence in a simulated environment. Nowak and Biocca (2003) delineated the term presence into three categories: social presence, copresence and telepresence. While the first two refer to the sense of social interaction and to a psychological link to another

(16)

11

individual, the concept of telepresence is characterized by the experience of a virtual environment (Nowak and Biocca, 2003; Jung, 2011). The social presence and copresence will not be examined here, as the focus of our study concerns consumer responses towards a product within a mediated environment, rather than the social links and interactions between actors. In addition, the concept of telepresence is more appropriate and often utilized to help understand the attitudes and behavior of consumers within a virtual environment (Cheng et al., 2014).

Steuer (1992) identifies two technological variables contributing to the experience of telepresence: vividness and interactivity (see Figure-2). The author describes the two components as following: vividness represents the “richness of a mediated environment as defined by its formal features, that is, the way in which an environment presents information to the senses”(p. 11), whilst the term interactivity is defined as “the extent to which users can participate in modifying the form and content of a mediated environment in real time” (Steuer, 1992, p. 14). When considering these dimensions, Steuer adverts that the contribution of these variables is relative in the creation of telepresence. This means that the feeling of telepresence will vary according to one’s individual perception of presence. In fact, the two dimensions vividness and interactivity influence the power of the technology rather than the perceiver. The relative contribution of these variables explicates the reasons to use attitude models (Van Kerrebroeck, Brengman & Willems, 2017; Riva et al., 2007) in order to comprehend the various stimuli influencing an individual’s behaviour. In this section, we analyse the two constructs in-depth as they are foundational for our analytical model.

Figure 2. Technological variables influencing telepresence. Source: Steuer (1992)

(17)

12 Vividness

The notion of vividness refers to “the ability of a technology to produce a sensorially rich mediated environment” (Steuer 1992, p.10). In that sense, vividness is characterized as a variable of telepresence as its richness and depth induce a sense of presence. The concept of vividness has seen a fair amount of research within the marketing communication literature.

For instance, Klein (2003) labelled the term as ‘media richness’, as it would better describe the phenomenon and with the purpose to distinguish it from the unrelated term of ‘vividness’ in marketing according to Kisielius and Sternthal (1986) – however, for the purpose of this study, the term ‘vividness’ will be utilized, as it arguably better encompasses the full scope of the telepresence framework. In addition, Bone and Ellen (1992) mentioned the spectrum of the concept, that is, the richness of the media ranges from poor presentation to an enriching experience of the environment. In their paper, Cheng et al. (2014) also emphasize the impacting role of vividness on telepresence, as well as on interactivity, the second technological variable explained by Steuer. The authors stipulate that vividness can trigger sensory perceptions of consumers to significantly improve interactivity, as it provides a richer environment in which the user can interact (Cheng et al., 2014).

Vividness can be analysed through two factors: sensory breadth and depth. The breadth of a media corresponds to the number of sensory channels that can be measured simultaneously (Steuer, 1992; Klein, 2003; Cheng et al., 2014). For instance, thanks to contemporary technologies, various senses can be stimulated in virtual environments, such as sounds provided by the headset, visual effects, the feeling of touch via a haptic glove or controller, etc.

The juxtaposition of these senses enhances the richness of the media and contribute to an augmentation of the vividness (Biocca, 1992; Steuer, 1992). Vividness also depends upon the sensory depth of a media, which refers to the quality of each sensory channel (Steuer, 1992).

Depth often corresponds to the resolution of the images, the quality of sound and other effects created in the virtual environment. Hence, a higher quality of sensory channels generates a higher perception of mediated presence (Steuer, 1992; Cheng et al., 2014). Both dimensions, breadth and depth, are simultaneously serving each other to enhance vividness and thus to create a sense of telepresence.

Interactivity

The second technological variable influencing telepresence is called ‘interactivity’. Steuer (1992) defines interactivity as “the extent to which users can participate in modifying the form and content of a mediated environment in real time” (p. 14). As seen earlier, Steuer’s (1992)

(18)

13

telepresence framework is present throughout large parts of the VR literature, sometimes with minor revisions. For example, the concept of interactivity was deemed too indistinct for Ariely (2000), who instead use the narrower term “user control”, a formulation also adopted by Klein (2003) to describe the potential for a user to influence and modify a (virtual) environment. Yet again, the term ‘interactivity’ is chosen here as this study is based on Steuer’s framework (1992) and it includes the full sphere of technological factors essential for our purpose. More recently, Van Kerrebroeck, Brengman & Willems (2017) also studied the application of VR in the context of transformational brand experience and judged that VR in marketing communications provided too little interactivity. In theory however, interactivity plays an equally important factor as vividness in the influence of telepresence on user attitude and is therefore studied as an integrated concept (Steuer, 1992).

Like vividness, interactivity is determined by key factors. Although many contributing factors exist in extant literature (Klein, 1998; Ariely, 2000; McMillan and Jang-Sun, 2002), we decide to focus on the three factors delineated by Steuer (1992): speed, range and mapping. First, the speed of interactions corresponds to the pace at which an action can be assimilated into the environment (Steuer, 1992). For instance, when a user interacts with his environment, such as opening a door, the highest possible rate of motion for that action would correspond to a real- time scenario.

The second characteristic contributing to interactivity is range. According to Steuer (1992), range refers to the number of actions made possible in the mediated environment. In a simulated apartment, a high level of range could entail the ability to open doors, pull/push furniture, handle objects, open/close the tap, etc. Hence, the greater the number of attributes that can be manipulated within the environment, the greater the range of interactivity (Steuer, 1992).

Finally, the third and last factor is mapping. Mapping is determined by the connection between a user’s natural action and the action reflected within the given medium (Norman, 1988). The objective of environment designers is to create controllers that are handled by users in the most natural and intuitive way possible. For example, when looking at the action of opening a door, the natural movement of a hand reaching for the handle, pulling it down and pushing the door to open it, should be reflected within the mediated environment to simulate a real-world setting.

Consequently, mapping is increased when controllers react in the mediated environment as the human body would do in reality (Steuer, 1992).

(19)

14

In conclusion, ‘vividness’ is how close to reality the mediated environment is materialized for the user, which fluctuates according to the number of sensory information channels measured simultaneously in the environment (breadth), and the quality of this information (depth).

‘Interactivity’ is how close to reality the actions of a user are performed in the mediated environment, which differ according to the number of attributes a user can manipulate (range), the rate of his actions (speed), and the intuitiveness/naturalness of his actions (mapping). All of which have a fundamental influence on the user’s feeling of telepresence within the given medium.

2.3 Telepresence and Consumer Attitudes – A Novel Conceptual Model

In prior literature, researchers have demonstrated that telepresence significantly influences consumer’s attitudes and behaviors in VR experiences (Klein, 1998; Hoffman and Nowak, 1996; Jung, 2011). For example, Li, Daugherty & Biocca (2001) studied the impact of 2D versus 3D product representations and determined that the role of telepresence favourably affects the customer attitudes towards online advertisements. Hoffman and Nowak (1996) also explored the influence of telepresence on web users’ attitude and concluded that the greater level of telepresence induced a greater effect on attitude. Klein (1998) posited that consumers develop a better attitude toward the product and its attributes when the product is experienced through a mediated environment. She explains that a high degree of telepresence enables a more persuasive communication, meaning that the consumer response and attitudes are positively affected by the intensity of telepresence (Klein, 1998; Barnes, 2016). Park, Stoel &

Lennon (2008) showed how increased telepresence through online product rotation positively impacts consumer responses through increased perceived information, a heightened mood and more favourable attitudes. The study by Manis & Choi (2018) showed that perceived enjoyment from VR is the most powerful predictor for using VR hardware, which points to telepresence as a facilitator for hedonic affective responses.

However, a fault in extant telepresence-attitude research is the absence of diverse VR experiences. The positive connection between the two concepts can arguably be drawn to the pleasant nature of the content utilized in previous studies. This is made evident by Riva et al.

(2007), who compared affective attitude responses between two different types of VR experiences: One designed to be inducing calm, the other to be inducing anxiety. The results

(20)

15

clearly showed that telepresence acts as a response facilitator, capable of strengthening both positive and negative emotions.

The proposed conceptual model (Figure 3) is a development of the telepresence model presented by Steuer (1992), integrated with research conducted on the telepresence-attitude relationship (Grudzewski et al., 2018; Van Kerrebroeck, Brengman & Willems, 2017; Li Daugherty & Biocca, 2002; Park, Stoel & Lennon, 2008; Riva et al. 2007). We postulate that a thorough understanding of the relationship between the two concepts cannot be reached without modelling both in detail. Therefore, we highlight the influence of technological variables on telepresence. Secondly, we present the attitude construct as seen in the tripartite model – focusing on observable responses of an individual towards a defined attitude object (e.g. Ajzen, 1989; Rosenberg & Hovland, 1960).

Figure 3. Conceptual model of Telepresence and Attitude Responses, derived from Steuer (1992), Ajzen (1989) and Rosenberg & Hovland (1960).

We derive that an increased level of telepresence will facilitate attitudinal responses. As seen in Riva et al. (2007), this mediation can be favourable or unfavourable, depending on the character of the presented content. In the proposed framework, the relationship between the two concepts is illustrated by “+”, which symbolizes the mediation of telepresence on the attitude responses. The novel model will be utilized to structure and analyse our collected data, facilitating a better understanding of how telepresence – and thus immersive technologies in general – affect consumer attitudes.

(21)

16

3. Method

The following section encompasses and justifies the method adopted to address our research problem and details how the data collection was executed. Furthermore, this chapter presents our approach in terms of collection and interpretation of data. Conclusively, we discuss the limitations related to the chosen method.

3.1 Research Approach

The aim of our study is to further develop the understanding of VR technologies and their influence on consumer attitudes. A thorough comprehension of the phenomenon entails, by using the adapted method, that we elucidate the variation of consumer attitudes when exposed to 2D versus 3D stimuli. Comparable investigations can be found in existing literature on VR and telepresence (Biocca, 1992; Klein, 1998; Hoffman and Nowak, 1996; Jung, 2011).

However, most of these researchers have analysed the concept of telepresence using quantitative methods, with a few exceptions (e.g. Li, Daugherty & Biocca, 2001). Similarly, there is a deficiency of conceptual rigidity in the researchers' approach to the attitude construct.

Because of the quantitative nature of the studies – and the operationalized survey methodology – attitude is often consigned to a favour/disfavour continuum on a Likert scale (Grudzewski, et al., 2018; Van Kerrebroeck, Brengman & Willems, 2017). In addition, the conceptual nature of attitudes is rarely clarified in detail, which neglects the rich attitude literature and causes confusion in the academic discourse.

Consequently, there is a growing necessity to delve into the consumer’s experience and conduct a qualitative study to saturate existing theory with qualitative data. Indeed, according to Bryman and Bell (2011; 2015), using a qualitative approach allows for a more in-depth and richer investigation of a phenomenon. Quantitative studies do not fully represent the interviewee’s perception and viewpoints, as opposed to what a qualitative approach would enable (Bell et al., 2018). From these arguments, we confer that a qualitative research approach is the most suitable for our purpose and thus applied in this study. Although some critics would argue that this form of study might be considered too subjective (Yin, 2014), it is exactly this subjectivity that we were trying to examine.

(22)

17

3.2 Research Design

In concurrence with our research question and our qualitative approach, this thesis applied an exploratory research design. According to Saunders, Lewis & Thornhill (2016), exploratory research designs are suitable for studies that seek to develop a deeper understanding about a problem or phenomenon, whose precise nature is not yet fully elucidated . Extant quantitative data suggests a telepresence framework with a set of five variables (depth, breadth, speed, range

& mapping), yet the subjective impact of telepresence is still relatively unknown. The advantages of exploratory studies lie in their flexibility and focus on building in-depth data.

Therefore, these designs are valuable when open questions are needed to gain further insights about a certain topic, which correlated well with our qualitative approach (Saunders, Lewis &

Thornhill, 2016).

To gain in-depth insights, we structured our data collection around a series of semi-structured interviews. These are characterized by a relatively loose interview structure which includes a predetermined set of questions, formulated to facilitate open and in-depth responses (Bryman

& Bell, 2011). The conversation is consequently guided but remains flexible, which allows researchers to respond to new information from the interviewee, such as posing follow-up questions. The interviewer can also change the order of questions to better suit the flow of conversation. However, the general structure and wording will remain intact between the interviews (Brinkmann & Kvale, 2009). These aspects of semi-structured interviews were deemed fitting to our research purpose, question and approach. By allowing our respondents to talk openly about their VR experience – thoughts, feelings, intentions – we hoped to capture the currently lacking understanding of telepresence and consumer attitudes.

Based on our research purpose, we decided to conduct the interviews in the form of a quasi- experiment. According to Bryman & Bell (2011, p.50), these studies display “characteristics of experimental designs but (…) do not fulfil all the internal validity requirements”. For example, the aspect of random assignment or using a control group – neither of which was implemented in this study. As we conduct qualitative research, the small respondent group does arguably not allow for rigorous comparisons between interview groups. Instead, we imposed an internal control factor by comparing 2D and 3D stimuli for each respondent. Thus, we focused our efforts into capturing the effect of increased telepresence (independent variable) on attitude responses (dependent variable).

(23)

18

3.3 Respondent Selection

Due to the extraordinary situation facing our research from the pandemic of COVID-19, respondent selection was heavily impacted. The Swedish Public Health Authority had recommended every citizen to stay at home, avoid social contact and conduct meetings remotely. This presented a major dilemma since our interview methodology rested on visual stimuli (an apartment) being presenting in 2D and directly after in VR. Thus, we were forced to limit our respondent pool to family and friends that would allow us to meet them. This also limited our options on imposing restrictions on respondent’s previous VR experiences, as our potential respondent group was so small. We still enacted socio-demographic requirements that was relevant to our purpose. The person had to be within the typical age-span for moving and purchasing an apartment, which we deemed to be above the age of 20. According to Boverket (2019), the age for moving to the first own apartment is usually around the age of 20.

Restrictions on the other end of the age-spectrum had to be set at 70 years maximum, because of the health risks associated with the COVID-19 pandemic (Folkhälsomyndigheten, 2020).

Table 1 below presents an overview of the respondents, their socio-demographic profile, previous VR experiences and interview time.

Table 1. Overview of the respondents and how they are referred to in the analysis.

(24)

19

3.4 Data Collection 3.4.1 Interviews

When conducting in-depth interviews, certain factors must be observed for meaningful, rich and engaging conversations to take place. Firstly, researchers must ensure that environmental factors do not inhibit the interviewee’s expression of thoughts and feelings (Bryman & Bell, 2011). The location should be where both the respondent and the interviewer feel safe and relaxed. To strengthen this aspect, it might be prudent for the respondent to determine where the interview should be conducted. We followed the recommendations from Bryman & Bell (2011) and aimed at conducting the interviews in spaces where these aspects were met.

Furthermore, we informed the respondents beforehand about the outlining structure of the interviews. We detailed the practical aspects of the interviews, such as approximative time needed and that a VR headset would play an integral part. However, to minimize the risk for strong Hawthorne² effects, the exact purpose of our study was left obscured. Moreover, we informed all respondents that the data collection would be anonymous and that the interview could be paused or stopped if they wanted to. All interviews were held in English; an aspect also confirmed with the respondents beforehand. The possibility to conduct the interview in Swedish was provided. When the transcription for each interview was finished, we offered to send it to the respective respondents to ensure that the information was understood correctly.

3.4.2 Interview Guide and Pilot Studies

A pilot study was conducted, consisting of three interviews, which permitted us to test the structure and flow of our interview guide. The purpose behind the pilot-study was to observe if any questions were misunderstood or did not stimulate the respondent to properly express the evaluative responses. Pilot studies is a common tool to detect faults in the chosen research approach and identify questions which generate uninteresting answers (Saunders, Lewis &

Thornhill 2016).

From the three initial interviews, we asserted that several questions needed a revision, since they did not allow for proper respondent elaboration or were too difficult formulated. We also noticed that some questions elicited very similar responses. During the revision process, the

2 The Ha wthorne effect is a phenomenon where the respondent is a ssuming the purpose of the study a nd tries to provide a nswers tha t – a s they believe – meet the resea rchers ‘expecta tions’ (Bryma n & Bell, 2011)

(25)

20

questions deemed redundant were removed completely. The amount of 2D pictures was increased to a total of 25 to better depict every detail of the apartment. We also adjusted our interview structure, since our feedback indicated that some questions were better asked while wearing the VR headset, particularly questions on specific aspects of the apartment (size, layout, light etc.), but also about perceived emotions during the VR exhibit. Thus, we asked more questions while the respondents were still in the VR experience. The full and final interview guide can be viewed in Appendix 1. Interview Guide.

3.4.3 Selecting Technological Tools and Content

The research operationalization of a quasi-experiment was structured around the presentation of an attitude object in 2D and then in VR. As highlighted in the introduction, real estate is seen as a market with considerable VR potential, and therefore an apartment was selected as a fitting attitude object. The utilized VR HMD was an Oculus Quest – at the time of study a relatively new VR model with integrated software and hardware and two remote controllers, thus being completely wireless. The wireless character was a major reason for choosing the Quest, as well as its advanced head tracking, stereoscopic view and high optical resolution (Oculus, 2020). Combined, these features point to a solid technical foundation for creating a strong sense of telepresence.

The utilized virtual apartment was selected based on several telepresence factors (see Figure 2). First, it was important for us that the media richness, i.e. the optical quality of the content, was of high standard. This included a significant degree of depth, which was concretized in high resolution and good lighting conditions. The content should also display the different rooms of the apartment in an adequate fashion. Additionally, a strong media richness entails the dimension of breadth: “the number of different sensory channels that a medium utilizes”

(Klein, 2003, p.45). This aspect is particularly important in VR simulations, since 2D pictures only stimulate the visual sensory channel, while the Oculus Quest also stimulates kinaesthetic senses through the aforementioned head tracking, allowing the user to actively move their head to look around. As such, we only looked at content that was compatible with this technological feature. The second requirement concerned the degree of user control: “the extent to which users can participate in modifying the form and content of a mediated environment in real time.” (Steuer, 1992, p.84). Although the degree of enabled control in virtual apartment exhibits is generally limited, we wanted to present content which, at the minimum, allowed users to

‘move around’ in the projected space.

(26)

21

With these requirements in place, we decided to opt for a virtual apartment provided to us by HSB. The visualization had been used as a marketing tool to display a real apartment in the Stockholm area. The apartment had one large living room with an open kitchen, a guest room, balcony, two bathrooms, a kid’s room and master bedroom. The visual quality was deemed to be of high standard, including well-lit rooms and good quality 360° photographs. Furthermore, it was compatible with VR head tracking and allowed users to move between locations in the apartment via a point-and-click feature³.

Since we wanted the content to be the same between the 2D and 3D stimuli, we extracted the 2D pictures from the web-based 360° visualization of the apartment. Efforts were made to present the different angles from each 360° photograph accurately, making the photos as informative and intriguing as possible. Changes in the respondents’ attitude should thus not be a result of dissimilar content, but instead from the increased telepresence of VR (see pictures in Appendix 2. The Apartment). The pictures were printed with a high-quality printer in an A4- format.

3.5 Operationalization

The operationalization of our research method intended to delineate our conceptual model into respective interview questions. Indeed, this method properly illustrates how the various concepts of the extent literature are incorporated throughout the interviews. Hence, we decided to split the interview structure into three sections corresponding to concepts of the literature and the used mediums of pictures and VR.

3.5.1 Interview Introduction

The first section was an introduction to the theme of VR, with questions relating to socio- demographic factors, as well as previous VR experiences and general attitude towards new technology. This was done to account for any strong novelty effects (Berlyne et al., 1963) and influences from the respondents’ tendency. While neither construct is a focal component in this study, we still concluded that it is necessary to take their potential influence into consideration.

Edwards & Gangadharbatla’s (2001) study is of particular relevance here, because of its focus

3 The point-a nd-click fea ture a llows users to tra nsport themselves in the projected spa ce by pointing the controller towards a ‘position icon’ and clicking once, thus being directly transferred to the chosen position

(27)

22

on 3D product presentations and novelty effects. They concluded that novelty seems to play a role in facilitating both product attitude formation and purchase intention.

3.5.2 Scenario & 2D Pictures of the apartment

In the second section, the respondents were first presented with an adapted scenario to minimize the risk for biased answers, resulting from personal factors such as living situation, salary income and motivation to move (see Appendix 1. Interview Guide). The scenario was built to feel relevant for all respondents and served as a vignette, a methodological tool described as a “very short story that, when carefully constructed and pretested, simulates real life experiences.” (Schoenberg & Ravdal, 2000, pp. 63–64). The purpose behind a vignette is to place questions into a tangible perspective, thereby eliciting rich but focused responses that are relevant to the study. A vignette was considered an appropriate measure to avoid responses that were excessively based on the interviewees current living situation. It was assumed that a proportion of the respondents had no intention to actually move, which could have led to a perceived lack of product relevance.

Afterwards, we showed 25 pictures displaying the selected apartment and asked questions relating to their affective, cognitive and conative evaluations. We wanted to observe in-depth patterns of how telepresence facilitated attitude changes. Therefore, we specifically asked questions regarding the respondents’ perceived level of telepresence, making sure to address each conceptual component of media vividness and interactivity.

3.5.3 3D Visualization with the VR headset (Oculus Quest)

In the third interview section, the respondents were shown the apartment anew via a VR headset. We placed no limit on the time spent viewing the object, neither the 2D pictures nor the 3D environment. Since we asked the same questions twice, we began by stating that the questions were going to be repeated , to minimize confusion. However, we added that the respondents should mention any perceived changes in their evaluation (if any such changes existed). The questions directed towards specific characteristics of the apartment and the respondents’ perceived feelings were asked during the VR experience.⁴ Afterwards, the headset was removed, and the interview continued. During the entire third section, the respondents had

4 This a pproa ch wa s a result from the previously mentioned pilot studies, which indica t ed tha t the respondents wa nted to a ctively view the a pa rtment while a nswering these questions.

(28)

23

no access to the pictures. A full overview of all questions and the respective theoretical foundation are displayed below.

Table 2. Operationalization of the interview questions and their conceptual foundation.

(29)

24

3.6 Data Analysis

The analysis of qualitative data, called coding, crucial in qualitative research to address the general weakness of validity (Bryman & Bell, 2011). According to Baralt (2012, p.230), codes

“can be single words, phrases, utterances, or even entire sections of highlighted text”. Coding entails that researchers read the collected data intensely, identifies important codes and then sort these into themes (Bryman & Bell, 2011). For our analysis, we opted for a widespread coding method called “colour coding”. The collected data is categorized into themes by using different highlighting colours, which allows for a more structured and manageable data pool.

This will help the consecutive analysis to be carried out with better precision and focus (Saunders, Lewis & Thornhill, 2016).

The interviews were recorded, saved, and later transcribed to their full extent. Using the adopted colour coding method, we categorized themes in accordance with the components of our conceptual model: Attitude responses (affective, cognitive & conative) and telepresence (vividness & interactivity). To exemplify: respondents’ expressions of feelings and emotions were coded as affective responses, which were sub-divided into spontaneous or elaborated affective responses. Then, we categorized these as favourable or unfavourable. Lastly, when these initial coding phases were completed, we analysed the composed data categories for themes that represented more general response patterns. We also addressed the aspect of tendency and novelty effects, which were made in a separate chapter.

3.7 Methodology Limitations

Our adopted methodology is not without limitations. Critics argue that qualitative studies are based on subjective opinions and are difficult to generalize (Bryman & Bell, 2011, Yin, 1994).

Arguably, the small number of interviews can be viewed as a limitation in terms of replicable findings to other contexts, as well as a risk of receiving biased responses from interviewees (Saunders, Lewis & Thornhill, 2016). However, we base our research on the support of prior quantitative analysis in literature, thus provide an in-depth extension of the existing framework.

As evident in Table 2, there is a disproportionate number of female respondents in our study – over 70%. This might affect our analysis in a gender-biased fashion. Our respondent pool also displays a lacking diversity in VR experiences. No participant had had extensive VR experiences, and most had no previous experience whatsoever. This makes it difficult to

(30)

25

account for novelty effects. Furthermore, the applied interview structure (allowing respondents to view the apartment through pictures, and then followingly in VR) presents its own limitations. Primarily, we ponder that it could lead to a ‘mere exposure effect’⁵. Any favourable attitude changes from VR could therefore be partially drawn to the recurring presentation of the apartment. Research has shown that the mere exposure effect can be more potent with subliminal stimuli (Zajonc, 2001). However, it is undoubtedly possible to detect its existence in more lucid circumstances, such as our interviews.

4. Empirical Findings and Analysis

In this chapter, we analyse the empirical findings from the interviews conducted. The findings are examined and modelled according to our theoretical framework. Therefore, the following sections are delineated into two main concepts: attitude responses and telepresence. Each segment highlights the clearest trends in our collected data, as well as outliers deemed important to capture the answers’ heterogeneity.

4.1. Attitude Responses 4.1.1 Affective

In the attitude framework, affective responses are defined as expressions of feelings and emotions towards attitude objects (Ajzen, 1989; Park, Stoel & Lennon, 2008). These responses are verbal descriptions of perceived subjective emotions, such as: “I feel comfortable”, or “I’m angry”. They can also be emotional expressions of approval/disapproval of an object: “I love this”, or “That’s disgusting”.

From our data, it is evident that overt affective responses towards the 2D-representation of the apartment are almost non-existent amongst our respondents. In fact, most participants bluntly answered that they “don’t feel anything” when looking at the pictures. On the question of why that is, answers centred around the perceived limitations of the photo-medium:

5 “Repea ted exposure of the individua l to a stimulus is a sufficient condition for the enha ncement of his a ttitude towards it.” (Zajonc, 1968, p.1).

(31)

26

R1: “No, it's just pictures I'm looking at. I can see it looks nice, but it doesn't evoke any feelings for me.”

Meanwhile, affective responses in the VR experience are, to some extent, also conspicuous with their absence. When asked to elaborate on perceived emotions during the virtual exhibit, several respondents showed signs of confusion and had difficulty expressing themselves:

R8: “Emotions? I dunno. It feels just... Like a nice apartment. I think that… it feels nice. I don't have anything else to add.”

Like Respondent 8, some interviewees resorted to talking about cognitive aspects, stating that the apartment “feels nice”, or “seems spacious”; thereby focusing on how they perceived the apartment or the VR experience, instead of how they actually felt. This tendency to describe the apartment instead of emotions was especially prevalent when looking at the pictures, but it is also present during the VR exhibit. Notwithstanding, a clear accentuation of affective responses with VR can be identified. An observable pattern were expressions of excitement and having fun. This hedonic aspect is sometimes derived from the VR experience itself. As few respondents had tried VR before, we suspected some novelty effect in our data, which is illustrated by R2 and R5:

R2: “Yeah, I feel a bit excited. Perhaps mostly because I’m trying VR.”

R5: “It's exciting to use the VR and obviously feels a bit more realistic. As if you are in the flat now, in the apartment.”

The above quotes are in line with extant research on novel stimuli and an increased level of arousal, as briefly elucidated in chapter 3.5 (Edwards & Gangadharbatla, 2001). From our fifteen respondents, thirteen had either no previous experiences with VR, or they had tried it only once – often with relatively limited interactions and simpler VR models. Thus, it comes as little surprise that this pattern of ‘excitement from trying VR’ should be prevalent in our study. However, the articulated hedonic sensation is not exclusively attributed to ‘first VR experience’, but also since VR replicated the act of going to an exhibit and exploring the apartment.

R2: It’s exciting to go to a viewing. I got sort of the same feelings from VR as when I’m in a viewing.”

The Effects of VR on Consumer Attitudes: Lessons from Implementing VR Technologies in Real Estate Marketing