1
The effects of consistency on luxury
perceptions and attitudes toward the
brand.
Master thesis author: Tommaso Alchidi (11372982)
Supervisor: Dr. Frank Slisser
MSc. In Business Administration – Marketing Track University of Amsterdam
2
Statement of Originality
This document is written by the UvA Student Tommaso Alchidi 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.
In this thesis, the use of the pronoun “we” is being adopted for writing style purposes only. The research paper was written by one author, Tommaso Alchidi.
3 TABLE OF CONTENTS AKNOWLEDGEMENTS……….…… 4 ABSTRACT………... 5 1. INTRODUCTION……….……….. 6 2. LITERATURE REVIEW………...…… 8
2.1 The Luxury concepts and its dimensions………..………... 8
2.2 The Luxury brand paradox and the Internet as communication tool……….……… 12
2.3 Luxury brands and Social media……….………... 14
2.4 What is Consistency……….……… 15
2.5 Visual stimuli and Content……….………. 18
2.6 Non-personal luxury perceptions and brand attitudes formation……….……….. 21
Gap and Research Question………... 23
CONCEPTUAL FRAMEWORK………...26
Explanation of the relationship in the model……….…………..26
Hypotheses………..27 3. METHODOLOGY………..………….. 30 3.1 Experimental Design………..………. 30 3.2 Subjects………..……….. 30 3.3 Procedure………..……….……….. 30 3.4 Pre-Test results………..……….……….32 3.5 The Experiment………... 33
3.6 Operationalization of the Variables………..….………... 34
3.6.1 Independent Variable………... 34
3.6.2 Mediators………... 35
3.6.3 Dependent Variable………... 37
3.6.4 Control Variables……….. 38
4. DATA ANALYSIS AND RESULTS……… 40
4.1 Analytical Strategy………..……… 41
4.2 Overview of the main hypothesis……….…………..………... 53
5. CONCLUSION AND GENERAL DISCUSSION ………..…………. 54
5.1 Limitations and Future Research directions ………..………. 56
REFERENCES……….…... 57
4
AKNOWLEDGEMENTS
I would like to thank my supervisor, Dr. Frank Slisser, for his assistance, encouragement and genuine belief in the final result.
Mr. Frank Slisser’s advices helped me make it to the final objective of handing in the thesis within the conclusion of the first year’s Master, which because of my full-time job seemed to be very hard to accomplish at a certain moment back in time.
Moreover, I would like to thank my parents, who through their constant support and belief helped me undergo the difficulties met during the Master’s Thesis “journey”.
5
ABSTRACT
This study will follow the open discussion regarding the role of consistency and the luxury brands “paradox” on the Internet. This paper, though, will analyze it from a different point of view, innovative and contemporary, precisely from a consumer’s standpoint. In fact, the majority of the studies about communication consistency and its effects are solely studied from the side of the companies and never from the side of consumers.
Then, the purpose of this work will be to answer these questions; What is the impact of luxury brands online consistency if it studied from consumers' point of view? Does it actually have an impact over perceptions' and attitudes' formation of consumers? We will try to demonstrate, through empirical results, whether the physical consistency, we define as graphical appearance and visual characteristics displayed by luxury brands’ screen interfaces through their Websites and Instagram’s home-pages, represents a concrete positive boost for their images, in terms of customers' perceptions and attitudes elicited, or only a strategy adopted by companies because representing of “the way it is”. Also, the aim of this paper will be to add practical insights about the image management of luxury brands in the online context and contributing to the developing of the luxury marketing literature.
6
1. Introduction
The concept of consistency is a topic that has been extensively addressed by many scholars in the literature. The main scope of this research will be of studying the concept of consistency in relation to luxury brands in their online environment. Specifically, we will dig in the content that luxury brands displayed through Websites and Social Media (i.e Instagram).
We started out by reviewing the literature regarding the concept of luxury with the aim of defining its meaning and facets. In particular, we will follow the research of Vigneron and Johnson (2004) and the scale they developed for the measurement of the perceived luxury contained in one brand, also known as the BLI scale. Subsequently, we will move on defining the role that luxury brands can play on the Internet, addressing what is being defined by the literature as the “luxury paradox” (Geerts&Sala, 2011).
Contrary to many studies, which addressed the concept of consistency only from the companies’ side, we propose a new approach through which the importance of consistency will be studied in connection to the perceptions and attitudes it elicits to consumers. In fact, browsing the literature we realized that the majority of the studies about consistency were, mainly, referring to the role it plays for companies, without ever proving its beneficial effects on consumers.
Since the many and divergent meanings consistency can assume according to the context and the object to which it is referred, we scrutinized the literature to look for a type of consistency (i.e physical consistency) that could have properly fit the scope of our paper, namely, a type of consistency that would function properly when considering the online presence of luxury brands, mainly displayed through Websites and Social Media (i.e Instagram).
7
We will investigate on the visual in/consistency, look or graphical appearance, of a fictitious luxury brand’s (i.e Casa Futon) Website and its two Instagram home-pages, designing the content in a way to propose a visual consistent and inconsistent scenario. We expect in/consistency to impact on the brand’s perceived luxury perceptions and eventually on people attitudes towards it, leading to more positive attitudes when the content between the two channels is visually consistent than when it is not. Such relationships will be, further explained in the literature review.
Instagram was chosen as the most pertinent social media to consider in the context of this research paper because, besides being the fastest growing Social Media of the last years with a number of 400 million active users per month (Kemp, 2016), it has quickly become the social network of choice for luxury brands and their followers given its photo and video-friendly layout that allows an easy and user-video-friendly visual storytelling (Scott, 2015; Sherman, 2015), high engagement rates and appeals to an attractive demographic (Jones, 2015).
8
2. Literature review
In this section we will be giving a general overview about the topics that this work aims to address. In particular we will browse over the relationship between luxury brands and the digital world. We will review the literature about three particular topics; the luxury concept, the definition of content and content consistency. In particular, we will refer to the luxury codes and the need of luxury brands to adapt them in the social media context.
2.1 The luxury concept and its dimensions
In this paragraph we will try to provide a distinctive definition of what is luxury. Further, we will focus on the so-called; luxury signals, also defined as “luxury codes”, drawing from the most relevant works related to this argument. Then, we will dig into the “Internet paradox” for luxury brands and how they need to manage their content over many different channels, we will focus in particular on the channels deployed thorough the adoption of the Internet; Social Media and Websites.
One of the most acknowledged luxury's definitions was provided in the 1997 from Kapferer who presented the semiotics of the word ‘luxury’, its sociological references and the pragmatics of luxury-brand management: ‘Luxury defines beauty; it is art applied to functional items. Like light, luxury is enlightening. [...] They offer more than mere objects: they provide reference of good taste. That is why luxury management should not only depend on customer expectations: luxury brands are animated by their internal program, their global vision, the specific taste which they promote as well as the pursuit of their own standards ... Luxury items provide extra pleasure and flatter all senses at once ... Luxury is the appendage of the ruling classes.’
9
Broadly, it can be defined as a way of being, a way of life but also a way to buy, consume and use, rather than know-how or a way to do something (Roux and Floch, 1996; Nyeck, 2004). It is definitely a hard task to define the concept of luxury in a unique fashion. In fact, it is being noticed by many recent studies how a significant chunk of it stems from subjective interpretations of the single individual or of the environmental characteristics in where a certain brand is presented. In other words it can relate to where and when one lives and the development state of the society in which this individual lives in (Heine, 2012).
Further often what is considered as expensive is often associated with luxury, thus the relativity of the luxury concept can be also strictly dependent on the economic resources of one individual. What looks as superficial and expensive for one person can be just mere comfort or lifestyle conforming for another individual. Plus, the definition of one product as luxurious is not only related to the object-product, per se, but also to the imaginary subject associated with the product usage stored in consumers' minds, which is dependent on the aesthetics of one individual which is as well strongly subjective (Parent, & Berthon, 2009).
This leads to the recognition that the luxury perception cannot be considered as absolute and that the amount of perceived luxuriousness of one brand, if recognized as it, can differ from one subject to another. Not all luxury brands are deemed equally luxurious (Cornell, 2002).
First of all, it is important for brands ensuring to get recognized as luxury brands. It is, indeed, considered from practitioners as important distinguishing trait to differentiate it in a given product category (Allérès, 1991; Kapferer, 1997), as important driver of consumer preference and usage (Dubois and Duquesne, 1993) as well as universal factor which prompts consumption across cultures .
Having this said, it is important to understand what are the features and traits that make one brand being perceived as luxurious. Many studies have attempted to identify
10
characteristic traits of luxury, for instance Phau and Prendergast (2000) emphasized “luxury brands compete on the ability to evoke exclusivity, a well-known brand identity, brand awareness and perceived quality”.
This gap in the literature is being filled in by Vigneron and Johnson (2004). They said, at that time, despite many researches were being made about the symbolic use of brands, there was still little research about the development of scales that would have enabled to measure the amount of luxury of one brand on the basis of the elicited luxury perceptions of such brands. Their study was meant not only to address this gap but also to help managers creating new luxury brands, monitoring the perceived level of luxury among people of their luxury brands and distinguishing what is luxury and what is not.
In order to support our research, the luxury framework developed by Vigneron and Johnson (2004) will be of crucial importance. In particular, the authors have developed an index, also known as the Brand Luxuriousness Index (BLI), which defines the dimensions of luxury. According to the BLI, luxury brands offer superior quality and performance (i.e., perceived quality), are scarce (i.e., perceived uniqueness), signal status and wealth (i.e., perceived conspicuousness), integrate meaning into consumers’ identity (i.e., perceived extended-self) and provide emotional benefits and intrinsically satisfying properties (i.e., perceived hedonism).
These five dimensions that the authors have drawn from other previously studies about luxury and brands, make up for two bigger categories, personal and non-personal perceptions. The former category accounts for the hedonic and the extended-self dimensions and relate to an individual inner desires and need while the latter, the personal perceptions, accounts for the conspicuousness, uniqueness and quality dimensions which refer to the individual concerns about how being perceived by others when using and consuming luxury
11
brands. Most people perceived a certain amount of luxury in one brand thorough a combination of these perceptions (Wiedman, Hennings, 2013).
Figure 1
Source: Vigneron, Johnson (2004) pag 488
The strong subjective interpretation of luxury associates this concept to a real philosophy, “with its own culture and identity” (Okonkwo, 2009).
As we have noticed from the luxury framework provided by Vigneron and Johnson there are certain aspects related to the luxury concept that, even if have not helped defining it uniquely, can set it apart from what luxury is not. This finds confirmation also in recent and less recent literature in which scholars seem to propose more consistent views about what are the components of luxury than what is luxury itself. So, drawing from the literature, the components of luxury which we will be calling luxury “codes” (Larraufie, 2015) represent the distinctive characteristics of luxury that Okonkwo (2009) described as a “culture and identity”. They, usually, resemble symbols of social status, lifestyle, quality, exclusivity but also brand heritage, authenticity and history.
12
Understanding the impact of such luxury codes over people perceptions represents another important step of the research here implemented. Moreover, their impact will be studied in an environment that many scholars told not to be adapted to what luxury was meant for; the Internet.
2.2 The Luxury brand paradox and the Internet as communication tool
Recent studies have risen significant concerns regarding the role of the Internet for luxury brands and if it can be considered as actually beneficial for them (Neuborne, 2000, Muller, Kocher, & Ivens, 2007). In fact, luxury brands on the Internet have a hard task to accomplish; they have to maintain a fragile equilibrium between high exposure and awareness but a controlled level of sales. This problem is best summed up in the ‘luxury brand paradox‘ (Radòn, 2012).
Many recent studies have focused on examining the role of the Internet for luxury brands either as communication tool, purchasing tool or customer acquisition channel. Many authors have conveyed to the vision that for luxury brands using Internet as communication tool can be definitely helpful; “for luxury brands, the Internet is very likely to fulfill a communication and information role, but less likely to be used as a customer acquisition channel” (Riley, Lacroix, 2003).
The aim of this paper, though, will not be about discussing the positive or negative sides of the Internet as customer acquisition channel or purchasing tool for consumers (e.g e-commerce), but to examine its effects as marketing communication tool over people luxury perceptions and brand attitudes. At first sight, the luxury “codes” that we defined as distinctive components and drivers of consumers' preferences seem to be incompatible with the mass-communication tool such as the Internet. Indeed, as we have already said, luxury codes resemble concepts such as exclusivity (Fionda, Moore, 2009), quality (Heine, 2012),
13
heritage (Beverland, 2006) etc.. which are definitely at odds with a “democratic” tool such as the Internet which is by definition ubiquitous and easily accessible (Anderson, Rainie, 2009). So, many marketing luxury managers are still trying to figure out how to integrate the luxury components in the Internet context, without “loss of image control, which would lead to a prestige loss as well” (de Chernatony, 2001): the more differentiated information luxury brands share online, the more potential distorted reuse of their stories they face. This signals how important it is for luxury brands managing their information sharing and communication with caution when adopting Internet channels.
So, how could luxury brands overcome the Internet paradox they face in the Web 2.0 era? How can they integrate their codes in the Internet context? Answers to this questions are being provided only by recent studies which have tried to find solutions for luxury brands to minimize, at least, their incompatibility with the Internet. An interesting work by Geerts and Sala (2011) has been one of the first trying to fill in the gap of overcoming the image dilution that luxury brands may face when expanding from traditional communication channels to the adoption of Websites, according to the authors the key to avoid dilution for luxury brands is consistency. They said that the Web 1.0, at that time, represented a unique opportunity for luxury brands to communicate their identity in a “creative way” as long as they had remained consistent with their brand's narratives. So luxury brands should develop and transmit their values from traditional channels to their websites in a way to strengthen and to be consistent with their brand's values (brand “narratives”). They said “in order to maintain the consistency of the brand on the Internet, it is important to resume on all pages on the website the same color codes, the same universe, the same word as used in other communication media”. Geerts and Sala have not proceeded in studying the effects that such visual consistency (websites' colors and codes..) has on people. The aim of the authors was, mainly, to contribute to the luxury brands' management on the Internet without studying the plausible
14
impact such content consistency has on consumers, and if it is actually represents a boost for luxury brands not only in terms of brand image but also in terms of consumers' perceptions and evaluations.
2.3 Luxury brands and Social media
Up to date, few studies have analyzed how luxury brands should manage their image consistency among more traditional channels, such as newspapers or billboards, and new ones, such Social Media. Through the latter, in fact, luxury brands have, finally, the chance to interact with their customers as they have never done before. Social Media give people the chance to start their own conversation with brands; Rubinstein & Griffiths (2001) pointed out ; “successful customer communication is customer initiated rather than company initiated”, plus using social media is a great way for luxury brands to gain more information about their needs from the consumers themselves, since their interactive nature, indeed (Okonkwo, 2009).
On the other hand, though, the rise of the Social media has, with no doubts, changed the power relationship between brands and their audiences, particularly for the case of luxury brands. Blogs, communities, forums have led to people disclosing their ideas and opinions and having their voices spread and heard by their peers, so that luxury brands must be careful when communicating with them. “The clients control their own online universe and can literally navigate the luxury cyberspace while shutting luxury brands’ direct influence out” (Okonkwo, 2010).
Differently from common brands, though, luxury brands can not settle on any short-term online demands of their customers, but they need to develop strategies over a long-short-term period where design and vision are at the center of it, to maintain a certain reputation. So, luxury brands must be able to understand the effects that communicating online can have on
15
their targeted audience. The study of Lee and Moisseaux (2013) has delved for the first time into how perceptions of customers can be affected by the presence of luxury brands on Social Media. In particular they took the case of an already existing top luxury brand such as Luis Vuitton and see how people coming across its Facebook's page would have perceived it, without considering the type of content that they displayed through this platforms, their conclusion was that “young consumers are slightly skeptical about the presence of luxury fashion brands on non-exclusive Social Media like Facebook”. The Lee and Moisseaux's study seems to provide subsequent proves of the fact that luxury brands' presence online can be risky for their images if it is not managed properly. This adds particular interest on the aim of our study. We think, in fact, that online content consistency as the luxury brands' key factor not only to reduce as much as possible the dilution risk they encounter when communicating over multiple online channels but also to capitalize on the new opportunities provided by the Internet.
2.3 What is Consistency
In order to be coherent with the aim of our study, we will focus especially on the few studies developed about the role that graphical and visual consistency plays when it comes to user-interfaces in a digital and online context (Moran, 1981, Nielsen, 1998, Satzinger&Olfman, 1998, Mendel, 2010).
To Moran (1981), user interfaces can be defined as “everything the user comes into contact” while using system based processes, from PC software packages to the World Wide Web (WWW), defined also as “applications” (Satzinger&Olfman, 1998). With regard of the definition of user interfaces’ consistency, despite the multitude of scholars’ views about this argument, still looks hard finding a “one-fits-all” definition of it. As mentioned by Rhee, Moon and Choe (2005), drawing from the study of Grudin (1989), even a group of 15 experts
16
were unable to produce a definition of consistency during a two day workshop. Thereby, this study will follow the definition of consistency user interface provided by Mendel (2010), who defines it in a less scientific but more intuitive and simpler terms, as “the extent to which two interfaces or systems share a common look, layout and functionality”. Moreover, consistency user interfaces are meant to facilitate the transfer of learning from one application to another, because one user can draw on existing knowledge when using a new application (Satzinger&Olfman, 1998).
As mentioned in Satzinger&Olfman (1998), though, despite the interface consistency guidelines and standards, based on transfer of learning assumptions, “few studies have investigated the effects of interface consistency, especially when a user interacts with multiple applications”. At the time the authors developed their research, the Internet “era” was at its dawn and they refer to the role of user-interface consistency only with regard of computer-based “end-user application programs”. The authors’ aim was computer-based on studying the effect of user-interfaces applications’ consistency across multiple applications on users’ performances, regarding the accomplishment of a certain task. Satzinger&Olfman (1998) distinguished between internal and external consistency, defining the former as the “design choices made at any level of the user interface”, and the latter as “the extent to which similar design choices of the interface are made for two or more applications”. Like Satzinger&Olfman did, we decided to direct our attention on, what the authors have defined as, the external consistency based on a comparison, content wise, between more than one application. Contrary to our research, though, they have studied the effects of visual appearance and action language syntax consistency, which refers to the consistency of the commands labels (e.g. F1 = help), for example, on users’ performances regarding the accomplishment of a specific task. This differs from our research’s objective which studies the effects of the external visual consistency, images-based, on luxury perceptions and attitudes toward brands.
17
Other studies (Kellogg, 1987, Adamson&Wallace, 1997) have, differently from what Satzinger&Olfman (1998) did, defined consistency distinguishing it into three facets. According to these studies, in fact, there are three main dimensions of graphical interfaces’ consistency, respectively: conceptual, communicational and physical. As explained in Olzok&Salvendy (2000), the consistency’s conceptual dimension consist of “metaphor applied to an object or an action that is embodied within an object”, where the word object as meant by the authors, can stand for an end-user program or a website home page, for instance. The communicational dimension comprises “the consistency of the input and the output”, while the physical consistency component involves the “graphical appearance or the visual characteristics of an object” (Ozok&Salvendy, 2000). Some studies, like the one of Adamson and Wallace (1997) studied the effects of these three internal component of user interface consistency on performance and user satisfaction. Contrary to the most of studies, as also the aforementioned work of Satzinger&Olfman (1998), which focused on the effects of user-interfaces’ consistency on pc-based programs, the study of Ozok&Salvendy (2000) has shifted the attention toward the graphical user interface consistency of home-page websites which they define as “a single page that is the default page of a website on the WWW”. They significantly contributed to the literature regarding the effects of Website’s interfaces consistency through the developing of the very first tool which enables the measurement of web pages’ interfaces in/consistency, or web pages’ screen consistency, on the World-Wide Web (WWW), known as the Purdue Consistency Testing Questionnaire (PCTQ).
To fit the scope of our paper we will focus only on the screen Website and Instagram interfaces’ look, following the definition provided by Mendel (2010). In order to do so we will adapt the concept of physical consistency which by definition, as aferomentioned, is the “graphical appearance or the visual characteristics of an object” (Ozok&Salvendy, 2000). The physical consistency among the three consistency’s dimensions is the only consistency’s
18
facet which can fit the scope of this paper, namely, to measure the visual consistency between a Website and an Instagram linked page to see its effects on luxury perceptions and brand attitudes. Both the conceptual and communication consistency are, in fact, meant and, subsequently, adopted by scholars in their studies to measure the functionality of computer based programs to study their effects on tasks’ accomplishment performances of users. Thus they are not applicable in our research, which focuses exclusively on the, “look”, the “graphical appearance and visual characteristics” of our fictitious luxury brand as displayed through its Website’s and Instagram’s home-page to study their impacts on people luxury perceptions and attitudes toward the brand.
2.4 Visual stimuli and Content
As aforementioned, since we are investigating the “graphical appearance or visual characteristics” among different media, Website and Instagram, we, further, needed to the define what type of graphical appearance or visual characteristics we will take into account.
Browsing the literature, interesting studies have pointed on the role played by visual stimuli when designing branding strategies. Visual stimuli such as Websites, products' display and logos can assist in the building of strong brand images by providing differentiation, creating loyalty and standing out of the competition (Hutton, 1997, Schmitt&Simonson, 1997). This being confirmed by Wallace (2001), who found a strong correlation between the quality of the visual stimuli and the financial performance of one company, serving as cues to trigger perceived quality. Also, recent studies have demonstrated how the adoption of visual elements, in particular in Asiatic countries, resulted in more positive effects in terms of recalling and purchasing intentions than the adoption of verbal elements (Tavassoli&Han, 2002).
19
But what is the exact type of content we should be focusing on when talking about luxury brands? A great contribution to the current, few, literature available about this argument is being provided by the recent work of Anne-Flore Larraufie (2015) in her work titled “the E-semiotics of Luxury brands”. The aim of Larraufie (2015) was to offer a new contribution to the debate regarding the luxury paradox on the Internet, following what Geerts and Sala had started previously. How can luxury brands adapt their codes in a seemingly counter-luxury medium as the Internet? How can people recognize them in a digital context? In order to answer these questions the author scrutinized carefully the digital content of some of the most notorious luxury brands.
This led to six luxury content’s dimensions, respectively; know-how, history/heritage, lifestyle, story-telling, product display and visual emotion. Larraufie also distinguished between the most and the least utilized dimensions by the luxury brands she considered in her research, respectively; heritage/history (5.5), know-how (5) and story-telling (7.5) resulted to be the most frequently used in their websites.
source: Larraufie (2015). Table 4, pag.8.
She also explained how each dimension was characterized by specific online luxury brands' “behaviors”. The know-how dimension, for instance, is typically depicted online through the adoption of video/photos creation of the making of a luxury product or the usage of color of oldness, while the product display's dimension usually comes with images pointing on the
20
qualitative features of one product (see Figure 1). Depictions we faithfully adopted for the creation of our fictitious luxury brand’s Website and Instagram home-pages.
Figure 2
Also, we needed to adapt this content to the Instagram context took into account in our research, thus of content displayed through Website and Social Media.
Instagram pages, indeed, offer a way to give the fan base a closer look at the brand in the form of a more personal and authentic content than what the old-fashion advertising campaigns conveyed. Examples of this content could be videos and pictures of behind-the-scenes, fashion shows or the making of certain products, in line with the content’s dimensions of: know-how, product-display, heritage/history and storytelling as it has defined by the aforementioned Larraufie’s framework (2014).
21
2.5 Non-personal luxury perceptions and brand attitudes formation
We chose to study the luxury perceptions elicited by the physical in/consistency of our fictitious brand on the basis of the luxury dimensions that Vigneron and Johnson have distinguished as non-personal and personal dimensions of luxury. Following other studies (Delécolle T., Parguel B., Valette-Florence, 2014), we will focus only on the luxury's non-personal dimensions, namely; quality, uniqueness and conspicuousness. These dimensions' type of luxury are, in fact, more strictly related to one brand's characteristics, can be perceived by people also through “virtual” channels such as Instagram and are less susceptible to one's personality type, perfectly suiting the scope of the paper. On the other hand, the so-called personal dimensions, namely; hedonism and self-extension, are more strictly related to one's personality and can strongly differ among subjects, so possibly biasing the final results.
Researches have focused over the importance of the perceived quality dimension as guarantee of long-term success for brands. In particular, high perceived quality level is considered to be pivotal for brands aiming to allocate their resources in more than one industry, as in the case of brands' extensions (Aaker, 1990). But what is the role played by perceived quality in the luxury industry? Here, in fact, where “quality over quantity” is the motto, brands signal their luxury status through the expensiveness of their products. In fact, the products' price seems to be one of the most effective quality distinctive signal that luxury brands adopt, because “people are not able to readily identify a product as qualitative or not” (Keller, 2013). Studies on the effect of high priced goods on consumers' luxury perceptions and eventually its impact on brands' attitudes, have shown how actually luxury consumers tend to identify high prices with quality, positively enhancing their attitudes toward brands (Delécolle T., Parguel B., Valette-Florence, 2014). This is also confirmed by the highly
22
notorious Veblen effect, according to which the higher the price, the higher the desirability of one product (Bagwell, Bernheim, 1996) .
On the same line, uniqueness dimension of luxury is often associated in the literature with concepts like rarity, exclusiveness or scarcity (Vigneron&Johnson, 2014), which, aims to make something people are looking for, as difficult to obtain (Snyder& Fromkin, 1977). Also in this case, this concept reflects partially the, already mentioned, Veblen effect. In fact, price can not only signal quality but also exclusivity, scarcity or rarity which can enhance consumers' desire of products. It is common to find marketing practitioners adopting statements that recall such features in order to entice consumers products' purchases, such as; “limited release”, “limited edition”, “limited time only” etc.. Some studies have demonstrated how scarcity perceptions can increase the perceived value of one product (Brock, 1968, Lynn, 1991). Here, the concept of value is being defined as a product's “potency to affect attitudes and behaviors” (Brock, 1968, p. 246), showcasing possibilities that perceptions of uniqueness can lead to modifying consumers' attitudes toward brands. These studies are based on what is known as the commodity theory, according to which “any commodity will be valued to the extent that is unavailable”, where the commodity's concept stands for anything, messages, experiences or objects which respects three main criteria. They must be useful, transferable and can be possessed (Brock, 1968, Lynn, 1991).
The conspicuousness dimension of luxury, on its side, involves the purchasing of a product not only for self-satisfaction but also to impress significant others. When it comes to luxury, this dimension seems to be the most significant of the three, given that “ luxury goods are perceived as one of the most visible forms of conspicuous consumption” (Chen et al., 2008; Vigneron and Johnson, 2004). Moreover, luxury goods consumers are “ primarily concerned about others' perspectives of the self ” (Bushman, 1993) and “highly sensitive to others’ evaluations of them” (Yim et al., 2014). So that, the role played by these, so-called,
23
interpersonal motivations which get consumers to possess luxury goods, seem to strictly be dependent on the motive to demonstrate social superiority to others (i.e. status seeking) (O’Cass and Frost, 2002; Tsai, 2005). Although many scholars sustain that conspicuousness consumption, for the motives listed above, create value for consumers, some more recent empirical observations (Hung et al, 2011) seem to go against this belief. They showed, indeed, how this dimension negatively impacted on the purchasing intentions of Chinese Luxury consumers, raising doubts on its absolute positive effects on consumers' attitudes toward luxury brands. As it is also highlighted in Delécolle (2014), the conspicuousness effects on brand attitudes then, seem to be more ambivalent than in the case of the other two previously explained dimensions. This is in line with other empirical researches (Yim et al, 2014), about the relationship between inter-personal motivations to buy luxury goods and attitudes toward luxury brands. The main research's findings show how external factors such as the inner cultural characteristics of one society, habits and customs of its inhabitants as well as internal personal features or cultural background of one individual can lead to different attitudes toward luxury brands, that center their position on being perceived as conspicuous.
Gap and Research Question
Although the numerous studies debating the relationship between luxury brands and the Internet, only few relevant studies are being brought up regarding the role of content consistency in the online context. Many and divergent studies are being provided about the concept of consistency while only a few are being provided on the role it plays for luxury brands. Moreover, there is no current study in the literature that has proved the isolated effects of physical consistency as “look” and “graphical appearance and visual characteristics” (Mendel, 2010, Ozok&Salvendy, 2000) on users perceptions and attitudes,
24
always being studied from companies’ standpoint. Also, this study will dig in the, still existing, necessity of analyzing the effects of user-interfaces’ “when a user interacts with multiple applications”(Satzinger&Olfman, 1998).
As previously mentioned, the impact of content consistency over people perceptions and attitudes, in the Internet context, gains even more importance when considering luxury brands. In fact, by relying on the work of Geerts&Sala (2001) we have seen how the authors have stressed out the importance of keeping luxury brands' narratives consistent when communicating through different channels in order to not dilute the luxury brands' images. So, to fulfill the scope of our paper we narrowed down the concept of “brand narratives”, a là Geersts&Sala, coming down to the definition of brands' content, following the Larraufie's content framework (2015) and the studies on the role of brands' visual stimuli and their impact on consumers' evaluations (Hutton, 1997, Schmitt&Simonson, 1997, Wallace, 2001).
Browsing through the literature we saw how the concept of brands' consistency often arise among the marketing and branding literature and its relevance is always stressed for many different reasons, going from brands' extension evaluations to brands' images management. None of the studies we have come across, though, have never actually proved the effect of consistency on people perceptions and evaluations, studying it only from the companies' standpoint. Since “consistency is at the heart of luxury brands' management” (Roux, 1991) and the Internet for luxury brands is still considered, at least, potentially dangerous as potentially beneficial, we decided to, firstly, check if the perceptions of luxury are affected by in/consistent content, providing new contributions to the luxury brands' Internet paradox open discussion. Subsequently we will focus on the effect of content consistency on people attitudes toward brands, mediated by the luxury perceptions.
For the first time, we will provide empirical results on the effects of visual (images-based) consistency, aiming to prove its beneficial effects on people evaluations. The novelty
25
of the study here proposed even increases considering that the effects of consistency will be studied in the online context, addressing the role played by both Website and Social Media.
Based on the previously mentioned gap regarding the linkage between consistency and consumers' perceptions and attitudes formation, the goal of this research will be to study the effect of phisical online content consistency on luxury perceptions and attitudes toward luxury brands, from a consumers' standpoint. Specifically, the research question we will address is:
“Does physical content online consistency impact on consumers' luxury perceptions? And do these perceptions mediate the relationship between luxury brands' online content consistency and brands’ attitudes formations?”
To say that, as already explained, we will take into account only the non personal luxury perceptions to avoid biased results. Meanwhile, to comply with the scope of the paper, we will control for other external factors such as wording, age and nationality of the final experiment’s participants to verify the “solo” effect of the online physical consistency on perceptions and attitudes.
26
CONCEPTUAL FRAMEWORK
Figure 3
Source: Reproduction of the Author
Explanation of the relationship in the model;
So drawing from the literature, this study will investigate into the relationship between online physical consistency and luxury brands’ attitudes. According to this research the relationship between our two main variables is mediated by, what Vigneron and Johnson (1999, 2004) have defined as non-personal luxury perceptions. We hypothesize these perceptions generated and influenced by the perceived online content in/consistency, to have a direct impact on the attitudes toward luxury brands. Specifically, we expect that the more a luxury brand’s content will be perceived as graphical consistent between its Website and its Instagram home-page, the more the brand will be perceived as effectively luxurious, and so, more favorable should be the attitudes elicited towards it.
27 Hypotheses
Luxury brands must be able to transmit their luxury codes in a consistent way through their channels in a way to avoid diluting their brand image. Following, in particular, the Geerts and Sala (1999, 2004) construct regarding the importance of consistency for luxury brands, we hypothesize that consistent content, given that it reduces the dilution effect derived from the adoption of different channels, should positively affect the attitudes of, already existing or potential, consumers toward luxury brands. From this assumption we come to our first hypothesis, H1: Luxury brands physical online content consistency positively enhances brands' attitudes.
By relying on the Vigneron and Johnson’s brand luxury index framework (BLI – Appendix 2), we have identified the luxury’s dimensions, so that we are now able to quantify the perceived luxuriousness amount of one brand. Our other assumption is that, the linear relationship between online physical, colors and symbols, content in/consistency and brands' attitudes is mediated by the non-personal luxury perceptions of the Vigneron&Johnson framework so that consistent content should lead to an increase of the perceived quality, uniqueness and conspicuousness of the brand in a way to, subsequently, positively impact the attitudes toward our fictitious brand.
This reasoning leads to H2: For luxury brands, online physical content consistency enhances their perceived quality. To H3: For luxury brands, online physical content consistency enhances their perceived uniqueness.. And to H4: For luxury brands, online physical content consistency enhances their perceived conspicuousness.
As aforementioned, we hypothesized that those three dimensions work as mediators in the positive relationship between online content in/consistency and attitudes toward the brand.
28
Drawing from the literature we have seen how luxury goods perceived as expensive are often associated with high level of perceived quality leading to an overall increase of positive attitudes toward the brand (Delécolle T., Parguel B., Valette-Florence, 2014). On the same line perceived uniqueness usually associated with perceptions of rarity or scarcity can enhance the desirability of one product, and this is likely to affect positively, the attitudes towards it. This is also confirmed by what is known as commodity theory, according to “one commodity will be valued to the extent that is unavailable”. Thus, we expect perceived quality and perceived uniqueness to lead to a, generally, positive enhancement of attitudes toward luxury brands, leading to H5: Perceived quality, due to online physical content consistency, positively enhances people attitudes toward luxury brands. And to H6: Perceived uniqueness, due to online content consistency, positively enhances people attitudes toward luxury brands.
On the other hand the conspicuousness dimension of luxury, as we have seen, it is connected not only to the self satisfaction derived from the consumption of luxury goods but also to the need of impressing significant others. This can lead to more ambivalent reactions according to the subjective background, culture or habits of one individual (Delécolle, 2014, Yim et al., 2014). So, we expect perceived conspicuousness to lead to ambiguous effects on people attitudes, either negative or positive. This leads us to H7: Perceived Conspicuousness leads to ambiguous effects on people attitudes toward the brand.
Thus for luxury brands, the relationship between online content in/consistency and brand attitudes is mediated by perceived quality and perceived uniqueness, which we expect to elicit more positive attitudes toward the brand , and by perceived conspicuousness which we expect to elicit ambiguous attitudes toward the brand. In particular, following the Geerts&Sala study, according to which luxury brands must keep their brand’s narratives consistent among their channels to avoid dilution, we hypothesize if the same precept can
29
also apply when studying the graphical appearance consistency of luxury brands, also defined as physical consistency, following the studies developed about it by Ozok&Salvendy (2000) and Mendel (2010). Thus, we expect one brand which keeps its “look” consistent between its channels to generate more positive attitudes than one displaying its graphical content in an inconsistent fashion through the same channels.
This leads to H8: Physical Online content consistency elicits more positive attitudes toward luxury brands than physical online content inconsistency, when mediated by the effects of luxury perceptions.
30
CHAPTER 3 - METHODOLOGY
3.1 Experimental Design
This research will be of predictive nature, thus it aims to identify a linear combination between an independent (physical online content in/consistency) and a dependent variable (attitudes toward the brand) with the purpose of predicting in the best way the value of the latter. Our research’s approach will be of experimental nature by means of independent variable manipulation. The final experiment can be defined as cross-sectional study since it studies a particular contextual phenomenon at a specific moment in time.
3.2 Subjects
As a source, male and female university students aged between 20-26 years were mainly used for the realization of the experiment. Requiring diverse gender participation both in the pretest and in the final experiment is of crucial importance. In fact, a number of recent studies have demonstrated as the respondents’ gender plays an important role in attitudes’ formation (e.g Theodoridis et al., 2013).
We used a non-probability convenience technique for the selection of the final experiment’s participants. 110 respondents were gathered, reached by means of UvA email student system and Social Media (i.e Facebook Messenger).
3.3 Procedure The pretests
In order to achieve the goal of our study, namely proving that the consistency of visual content positively enhances consumers’ attitudes toward luxury brands and, thus, must be preferred over online inconsistent visual content, we created a fictitious brand named as CasaFuton. In particular, we created the CasaFuton’s Website (see Appendix 3) and two
31
CasaFuton’s Instagram profiles (see Appendix 4-5). The two Instagram profiles which are different, content-wise, one being physically, or also visually, consistent with the Website and one being visually inconsistent with it, are of crucial importance to satisfy the aim of this paper.
Previously, though, in order to prove that the effects on people perceptions and attitudes was only due to the in/consistency between the two communication channels (Website vs. Instagram) has been necessary to prove, that both the CasaFuton’s Instagram profiles were perceived, on their own, as equally luxurious and also that they were actually perceived, respectively, as visually consistent, content-wise with our fictitious luxury brand’s Website and visually inconsistent, content-wise, with it.
We started out by, firstly, designing and creating our fictitious luxury brand’s Website home-page and subsequently the two fictitious luxury brand’s Instagram home-pages. Further, we manipulated the content of the two Instagram profiles, following the Larraufie’s content framework (see Table 6), in order to create a visual consistent and inconsistent scenario (see Pretest Appendix 8).
Then, we needed to run two pretests to make sure that the two fictitious luxury brand’s Instagram home-pages were actually perceived as physically consistent and inconsistent compared to our fictitious luxury brand’s Website, respectively, which was kept the same for both the scenarios. Through these two pretests we also have checked for the perceived luxuriousness of both our luxury brand Instagram home pages (see Appendix 3-4), to make sure that the variation in the luxury perceptions we aimed to see in the final experiment was actually due to the physical in/consistency between the Instagram home-pages and the Website and not to the Instagram content’s home home-pages on their own.
In the consistent scenario (see Appendix 3-4), our brand’s Instagram content was designed to show content’s dimensions representing the brand’s heritage/history and the
32
brand’s know-how (see Table 6, Larraufie, 2015). This being visually consistent with the brand’s Website, displaying images recalling brand’s heritage and know-how, as well. While in the inconsistent scenario (see Appendix 3-5) the Instagram content was manipulated to show different content’s dimensions, namely, the product display’s dimension and the storytelling dimension (see Table 6, Larraufie, 2015), inconsistently with the brand’s Website content which was kept the same as the former scenario. So, the pretests’ aim was to ensure that both the brand’s Instagram profiles were actually perceived as equally luxurious, on their own, and to check for the visual perceived in/consistency between CasaFuton’s Instagram pages and CasaFuton’s Website (Appendix 8 – pre-test form).
3.4 Pre-test results
We run two pretests, in order to prove that both the fictitious brand’s Instagram home-pages were perceived about equally luxurious and to verify if the Instagram home-page’s graphical manipulated appearances were perceived as actually consistent and inconsistent, respectively, with our fictitious brand’s Website content (see Appendix 8).
Twenty people participated to each of the test. The total number of participants (n=40) was reached by means of Google modules, in a way to ensure a quicker and secure procedure of data gathering for pretests, which were handed out mainly through the UvA e-mail system and Social Media (i.e Facebook Messenger). Pretests’ participants were asked to firstly have a look at the Casa Futon’s Instagram home-page in a way to, as already pointed out, evaluate their luxury perceptions over, only, the Instagram home-pages. In particular, they were asked to rate on a 5 point scale each luxury’s dimensions of the BLI scale. During the pretest also the personal luxury dimensions of the Brand Luxury Index (Vingneron&Johnson, 2004 – Appendix 2) were considered since in this phase we were only focused on making sure that the people overall perceptions of both the expected inconsistent
33
and consistent Instagram home pages were actually perceived as about equally luxurious. Before answering to the last three questions, attaining to the type of physical consistency we addressed in this paper, regarding the perceived colors, symbols and general consistency perceptions (Ozok&Salvendy, 2000), the samples were asked to carefully look, also, at the CasaFuton’s Website to enable them to compare the “graphical appearance and the visual
characteristics” between the aforementioned luxury brand’s Website home-page and its
Instagram home-page of reference (Appendix 8).
Pretests’ results confirmed our expectations. The overall luxury perception, obtained as average of the values of each item, was 3.6/5 for one Instagram home-page and 3.1/5 for the other, which we considered as about the same. The pretests’ have also shown one luxury brand’s Instagram home-page being perceived as graphically consistent with the luxury brand’s Website while the other luxury brand’s Instagram home-page being perceived as graphically inconsistent with the luxury brand’s Website, as expected.
3.5 The experiment
In order to answer our research question, the final experiment was run by means of Qualtrics, which is more adapt for the developing of experiment of larger dimensions. In particular, a between-subjects 2(online physical content: consistent/inconsistent) X 2 (+/- attitudes toward the brands) experiment took place. It consisted of two main scenarios to which respondents have been randomly assigned. One scenario consisted of a Website interface and an Instagram interface with consistent content, in terms of “graphical appearance and visual
characteristics”, (both the interfaces will display know-how and heritage of the brand) and
the other scenario consisted of a Website interface and Instagram interface inconsistent content, as pretested (Website: Heritage and Know-how vs. Instagram: Product display and
34
Story-telling). In both the cases, the content was manipulated following the Larraufie’s framework ( 2014 - Appendix 6).
We will further display the final experiment’s results which accordingly to our hypothesis should lead to more or less positive attitudes toward our fictitious brand due to the positive or negative mediation effects of the elicited luxury perceptions.
3.6 Operationalization of Variables
3.6.1 Independent Variable
The independent variable in the experiment is: the Online Physical In/Consistency (CFIG). We pretested if the two fictitious luxury brand’s Instagram home-pages were actually perceived as physically consistent and inconsistent compared to our Website, thus we needed a scale able to measure the graphical appearance in/consistency between our fictitious Website’s home-page and our two fictitious Instagram’s home-pages, respectively. Since almost any scale found was mainly based on the measurement of consistency within the same “application” focusing on elements others than visual elements, or images, we adapted the Purdue Consistency Testing Questionnaire (PCTQ) introduced and developed by Ozok&Salvendy (2010). We have not taken into account the entire test, which consists of 67 questions addressing each consistency’s facet, but only two dimensions of one consistency dimension, the aforementioned physical consistency, respectively what Ozok and Salvendy defined as colors and symbols. For the latter dimensions we adapted the definition of symbol as “representation of an object” (Dictionary.com) in a way to adapt this scale to our experiment and to the definition of content provided by the Larraufie’s framework (2014)(see Appendix 8). These two dimension, are, in fact, the only ones in line with the aim of our research and applicable to our experiment (with the necessary aforementioned adaptation of the meaning of symbol) and the channels through which the brand’s content is displayed.
35
This will allow us to uniquely analyze the graphical appearance’s consistency we address in the present paper, cohesively with the definition of physical consistency provided by Ozok&Salvendy (2000) and Mendel (2010) . Also, in line with the most of the studies debating the role of user interfaces’ graphical consistency (Ozok&Salvendy, 2000, 2001, 2010, Satzinger&Olfman, 1998, Mendel, 2010) , we used a 5 point Scale to measure these two dimensions plus the overall perceived consistency as in Ozok&Salvendy (2010).
The authors showed how the items included in the PCTQ (see Appendix 7) for the measurement of the physical consistency (“visual characteristics and graphical appearance”), including our two items; colors and symbols, were highly internally reliable with a Cronbanch’s Alpha significantly over the acceptable level of 0,7 (α=0.91).
3.6.2 Mediators
The mediators of the experiment are the non-personal luxury dimensions, namely: perceived quality, perceived uniqueness and perceived conspicuousness (QUA, UNI, CONS).
These three dimensions were being identified and defined by the studies of Vigneron and Johnson (1998, 2004). Who, as already mentioned in the literature, have distinguished between what they named as personal and non personal luxury dimensions through the development of the Brad Luxury Index (BLI) scale. The BLI scale besides being the most recent updated scale was chosen also because it was meant by the authors to, precisely, measure consumers’ luxury perceptions of the luxury of certain brands or products, contrary, for instance, to the scale of Dubois and Laurent (1994) which measures consumer perceptions on a more general level. The scale was meant, in particular, to distinguish between high and low end luxury goods. The five luxury’s dimensions included in the BLI scale, are: quality, uniqueness and conspicuousness, which account for the non-personal luxury dimensions while the hedonic and the extended self account for the personal luxury dimensions, for a
36
total of twenty items. Vigneron and Johnson obtained the final five luxury’s dimensions and the respective twenty items, through a deep selection of items which lead to a scale both internally valid and reliable. The reliability test included a sample of 418 business students and some brands selected on the basis of being highly recognizable and having a potential luxury image. The results showed both the scale (α = 0.86) and each items (0.7 < α > 0.9), to be internally reliable with all values over the acceptable level of 0.7. Both exploratory and confirmatory factor analysis was run by the authors to develop the scale, with the disregard of all the factor loadings less than 0.6. Scale validity was, further, run with results which confirmed a significant association between the open-ended answers provided by the respondents and their scores, providing further evidence to support the validity of the scale.
37 Table 1 – BLI scale
Source: Vigneron and Johnson (2004)
Nb: for the final experiment the items of the scale have been reverted to ensure that the least luxurious value to be the first (on the left side) and the highest luxurious value to be the last (on the right side). So the item with the asterisk (*) have been the only ones which we have not reverted in the final experiment.
3.6.3 Dependent Variable
The dependent variable of the experiment is; attitudes toward the brand (ATT).
The attitudes toward the brand were measured, following the studies of Spears and Singh (2004) who were the first scholars to develop a standard scale, namely “scales that provide
38
measurement of brands’ attitudes. This scale proved to be the most efficacious among the other competing ones in terms of measurement and synthesis.
Their studies, in particular, aimed to address the shortcoming observed in the literature in the measurement of attitudes toward brands by the development of valid measures that can be consistently used across studies. In order to do that, a set of items concerning attitudes toward brands was generated from the literature. Exploratory factor analysis was run in order to reduce the number of items, followed by a confirmatory factor analysis to make sure the set structure previously generated was actually correct. Also following the studies of Bagozzi and Yi (1998), all the items that had an extremely high item-to-total correlation (.90 greater) were removed. This led to five lasting items: unappealing/appealing, bad/good, unpleasant/pleasant, unfavorable/favorable, unlikable/likable. Internal reliability was also highly verified with this scale showing a Cronbach’s Alpha of 0.97 (Appendix 2).
3.6.4 External factors
In order to ensure that the attitudes’ change toward our fictitious brand was due to, uniquely, the physical online in/consistency comparing the graphical appearance of its Website and Instagram home-pages, it was essential to cosnider external factors that could have influenced the variables’ relationship here studied.
Content’s Wording
First of all, as we already pinpointed, since the pivotal aim of our study is to take into account, only, the graphical appearance, in terms of colors and symbols (“representation of
39
brands’ attitudes, we decided, in order to avoid interferences of diverse nature, derived for instance from the usage of a different set of words displayed in the expected consistent Instagram home-page compared to the inconsistent one, to keep the words in both the scenario exactly the same. This allowed us to avoid any possible wording’s biases derived from using certain words that could have, on their own, led to different attitudes toward our fictitious luxury brand regardless of the content displayed.
Participants as Instagram users:
Since the main aim of the paper is scrutinizing what are the effects of in/consistent visual content displayed through Website and Social Media (i.e Instagram), we also asked respondents to indicate if they are Instagram users as it has done in Hietanen (2016). Participants who indicated to be not Instagram users have been excluded from the main experiment.
We considered other variables such as: Age, and Nationality, to obtain an overview regarding the characteristics of the samples that participated in the experiment.
Regarding the Age of the experiment’s participants, we decided to externally control it by putting a threshold of 34 years old. This in line with recent statististics, showing that the two largest Instagram users’ groups are represented by people aged between 18 and 34 (Statista, 2016). We did this, also, to ensure that the participants’ knowledge of the topics addressed in our final experiment was adequate.
A dummy variable was created in order to distinguish among Italian respondents (= 1) and non-Italian respondents (= 0) and it will be considered as control variable in our final experiment (NATCOD).
40 4. DATA ANALYSIS AND RESULTS
A total of 131 responses were collected. Out of the all responses collected, interviewed with less than 18 years old and over 34 years old were not alowed to participate to the final experiment. So we remained with a total of 86 relevant responses. 42 subjects were randomly assigned to the physical non-consistent Instagram scenario while 44 were randomly assigned to the physically consistent Instagram scenario. As aforementioned, respondents were collected by means of Social Media and UvA email-system. 51 respondents indicated their Nationality as Italian, 8 as Dutch, 5 as Greek, 4 as German, 3 as Brazilians, 3 as Turkish, 2 as Romanian, 2 as British, 1 as Chinese, 1 as American, 1 as Moldavian, 1 as Swedish, 1 as Polish, 1 as Canadian, 1 as French and 1 as Peruvian.
The collected data will be analyzed in several steps. First we will check all data for missing or errorous items. Subsequently, the reliability of scales will be tested, using the Cronbach’s alpha. The goodness of the scales adopted will be also tested through a factor analysis, and thus the correlations among the items studied were also analyzed.
The hypothesis were tested by means of a Mediation Analysis, run by Process Macro (Hayes, 2012).
4.1 Analytical Strategy
The first step of the analysis was to check the data for errors and counter indicative items. No missing values or data errors were found. There was no need of recoding variables, since no counter-indicative items were found.
So, we proceeded with verifying the internal reliability of the set of items adopted by following the Vigneron&Johnson (2004) Brand Luxury Index (see Table 1) , to measure the luxury perceptions elicited by our fictitious Luxury brand.
41
The conspicuousness items set represented by its four items that we named as CONS1,
CONS2,CONS3 and CONS4, showed an high reliability with a Cronbach’s α significantly
higher than the general accepted cut-off point of 0.7 (Field, 2009), equals to 0.866. We verified also the correlation between the Conspicuousness’s scale items and the total score of the scale. With the former, showing a Corrected Item-Total correlation far over the cut-off point of 0.3 (Table 2). Reliability Statistics Cronbach's Alpha N of Items ,866 4 Item-Total Statistics Scale Mean if Item Deleted Scale Variance if Item Deleted Corrected Item-Total Correlation Cronbach's Alpha if Item Deleted CONS1 13,57 18,333 ,612 ,868 CONS2 12,68 15,432 ,688 ,843 CONS3 12,39 17,227 ,831 ,798 CONS4 12,94 13,743 ,792 ,799
Table 2 – Items Reliability – Cronbach’s Alpha
We, then, proceeded with analyzing the reliability of the Uniqueness items set, represented by four items, namely, UNI1, UNI2, UNI3 and UNI4. As for the case of the conspicuousness also the Uniqueness set of items showed high reliability with a Cronbach’s α = 0.896. The uniqueness’s scale showed a good correlation with the total score of the scale, with items’ Corrected Item-Total correlation well over the cut-off point of 0.3.
42 Reliability Statistics Cronbach's Alpha N of Items ,896 4 Item-Total Statistics Scale Meanif Item Deleted Scale Varianceif Item Deleted Corrected Item-Total Correlation Cronbach's Alpha if Item Deleted UNI1 12,61 18,410 ,795 ,858 UNI2 12,86 20,826 ,662 ,905 UNI3 12,58 20,246 ,805 ,856 UNI4 12,61 18,776 ,832 ,843
Table 3 – Items Reliability – Cronbach’s Alpha
Finally, we checked for the reliability of the last set of items used to measure luxury perceptions, namely the perceived Quality, characterized by five items: QUA1, QUA2,
QUA3, QUA4 and QUA5. The Quality items’ set showed high reliability with a Cronbach’s α
= 0.938. Also, the Quality’s scale items showed a good correlation with the total score of the scale, with a Corrected Item-Total correlation well over the cut-off point of 0.3 (Table 4).
Cronbach's Alpha N of Items ,938 5 Item-Total Statistics Scale Meanif Item Deleted Scale Varianceif Item Deleted Corrected Item-Total Correlation Cronbach's Alpha if Item Deleted QUA1 18,25 43,176 ,810 ,929 QUA2 18,53 42,703 ,879 ,915 QUA3 18,35 45,807 ,805 ,929 QUA4 18,38 45,646 ,798 ,930 QUA5 18,50 45,099 ,888 ,915
43
Afterward, we studied the reliability of our 7 point-scale measuring attitudes toward the brand, following Spears&Singh (2004, Appendix 2). The attitudes’ scale showed to be high reliable with a Cronbach’s α = 0.953, and all items displayed a Corrected Item-Total correlation well over the cut-off point of 0.3, meaning that this scale is well correlated with the total score of the scale (Table 5).
Reliability Statistics Cronbach's Alpha N of Items ,953 5 Item-Total Statistics Scale Meanif Item Deleted Scale Varianceif Item Deleted Corrected Item-Total Correlation Cronbach's Alpha if Item Deleted ATT1 20,69 21,877 ,785 ,956 ATT2 20,11 21,678 ,851 ,945 ATT3 20,07 20,742 ,923 ,933 ATT4 20,29 20,379 ,908 ,935 ATT5 20,11 21,283 ,883 ,940
Table 5 – Items Reliability – Cronbach’s Alpha
Factor Analysis
Furthermore, a factor analysis was run in order to investigate the similarities between the set of items utilized in this study to measure our luxury perceptions (CONS1-2-3-4, UNI1-2-3-4,
QUA1-2-3-4-5). A factor is defined as a dimension which represents the concise statement of
the correlations between a set of variables (Kline, 2014). The factor analysis will help us to evaluate how good these set of items we used are.
As already mentioned, the non-personal luxury perceptions, namely our mediators, conspicuousness, uniqueness and quality (Vingero&Johnson, 2004), were measured following the Vigneron&Johnson’s BLI scale, on a 7-point scale (Table 1 – pag.37).
44
A principal axis factoring analysis (PAF) was conducted on the scales of our mediators. The Kaiser-Meyer-Olkin measure verified the sampling adequacy for the analysis, KMOO =.906. Bartlett’s test of sphericity χ² (78) = 969.944, p < .001, indicated that the correlations between items were sufficiently large for PAF (Table 6). An initial analysis was run to obtain eigenvalues for each component in the data. Only one component (CONS1) out of the 3 items (CONS, UNI, QUA) considered, had eigenvalues over Kaiser’s criterion of 1 and it explains 68.5 % of the variance (Table 7). In agreement with Kaiser’s criterion, examination of the scree plot revealed a leveling off after one factor (Appendix 10).
KMO and Bartlett's Test
Kaiser-Meyer-Olkin Measure of SamplingAdequacy. ,906
Bartlett's Test of Sphericity
Approx. Chi-Square 969,944
df 78
Sig. ,000
