The role of consumption motivation
in the evaluation of
category-spanning music festivals
M.C. (Marieke) te Poele
11417323
Master’s thesis
Entrepreneurship and Management in the Creative Industries
Thesis supervisor: J.F.E. de Groot MSc
Second reader: dr. M. Piazzai
June 22, 2018
Final version
Statement of originality
This document is written by Student Marieke te Poele 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.
Abstract
Consumers use categories in order to make sense of the market and its products. Producing organizations use categories to receive legitimacy for their products. Nevertheless, transcending the boundaries of categories, i.e. category spanning, is performed as well in order to develop new product opportunities. Category spanning is also known for inducing consumer confusion and aversion. However, the negative effect category spanning has on the evaluation is not applicable for every type of consumer. Innovation literature shows there are consumers who are interested in category-spanning products. This study investigates the distinction that can be made between consumers based on their consumption motivations in the empirical setting of the Dutch music festival. As it appears, there are consumers who predominantly have enjoyment motivations to attend the music festival and there are consumers who predominantly have music motivations. The expectation of this study is that enjoyment motivated consumers will respond less negatively to category spanning as their consumption motivation has less to do with the product category, i.e. genre. In contrast, consumption motivations of music motivated consumers are dominated by the category, so the expectation is that they will also attach more value to a ‘clear’ category. Nevertheless, the results of this study provide no proof to accept or disprove these assumptions. Key words: category spanning, evaluation valence, consumption motivation
Acknowledgements
I want to express my great gratitude to everyone who has advised me, stood by and supported me during the writing process of my master's thesis and during my bachelor’s and master’s in general. It gave me new insights, courage and motivation.
Special thanks to Jan de Groot for his professional and pleasant thesis supervision. From the start, he provided me with extensive feedback via the mail or in the supervision appointments. He provided me direction and made sure that I remained motivated. I could not have a better supervisor.
Moreover, I want to thank my mother for providing me the structure to keep studying and stay motivated, for her scientific knowledge and support. My housemates also deserve a thank-you here, since all those weeks they had to listen to my struggles during this process and were always available to talk about thesis issues. I want to thank Lucia and Rick for all those ‘UBakkies’. Those moments were the reason I started my master’s at all. Lastly, I want to thank the crew of BazarMedia for providing me the necessary distraction during the last weeks of the thesis process. Those Saturdays made me do my best every week.
Table of contents
1. Introduction 06 2. Literature review 10 2.1. Categories and category spanning 10 2.2. The role of consumer motivations 16 3. Data and method 23 3.1. Research design 23 3.2. Operationalization of the variables 24 3.2.1. Dependent variable 24 3.2.2. Independent variables 26 3.2.3. Control variables 29 4. Results 33 4.1. Descriptive statistics 33 4.2. Testing the hypotheses 36 4.2.1. Hypothesis 1 36 4.2.2. Hypothesis 2 40 4.2.3. Hypothesis 3 42 4.2.4. Hypothesis 4 43 4.2.5. Hypothesis 5 44 5. Discussion 45 5.1. Empirical findings 45 5.2. Implications, limitations and future directions 47 6. Conclusion 49 7. References 50 8. Appendix 57 8.1. Appendix A – Survey 57 8.2. Appendix B – Scatter plots 631. Introduction
People use categories to make sense of the world (DiMaggio, 1987; Douglas, 1986; Kuijken, Leenders, Wijnberg & Gemser, 2016; Wijnberg, 2011). Categories serve as reference frameworks in order to determine where the subject belongs to (DiMaggio, 1997). Categories help consumers to create familiarity with a certain product. They offer guidelines, reference points and standards of evaluation (Ouellet, Savard & Colbert, 2008). People use these categories in their purchasing behavior as well. For example, artists who launch a new album portray their album in a particular genre category. At that moment, consumers know what kind of music they could expect. The creative industries are a pre-eminently example where categorization fulfills a key role. Since products of the creative industries are predominantly hedonic and do not possess clear and distinctive functionalities to differentiate and evaluate on, categories offer guidelines, reference points and standards of evolution (Ouellet, Savard & Colbert, 2008). Additionally, much research has been conducted on the effects of categorization during the selection, purchase and evaluation process by consumers (Bowen & Daniels, 2005; Gelder & Robinson, 2009; Gemser, Van Oostrum & Leenders, 2007; J.S. Lee, C.K. Lee & Choi, 2011; Lena & Peterson, 2008; Redondo & Holbrook, 2010; Paleo & Wijnberg, 2006; Peltoniemi, 2015).
Categorization represents the process of dividing products into frames which guide stakeholders to assign legitimacy, meaning and value to the subject (Hsu, Hannan & Koçak, 2009; Khaire & Wadhwani, 2010; Navis & Glynn, 2010). The shared understandings belonging to categorization give guidance to what lies within the category and what does not have a fit with the other subjects. These constructions contribute to the stabilization and transparency of the market (Hsu, Hannan & Koçak, 2009).
Organizations can choose to target multiple product categories. This is called category spanning. The advantage of spanning multiple categories is that the potential audience for the product will be multiplied as well. This will result in an increase in the product’s performance in terms of, for example, visitor numbers or box office revenues. The more categories a product addresses, the more it makes a claim on the audiences that are consuming in those categories. Nevertheless, as it appeared from former research (Hsu, 2006), spanning categories has also its deficits. The principle of allocation dictates that spanning categories and therefore addressing a larger audience means that, at the same time, the organization will face a lower appeal to its audience (Hannan & Freeman, 1977, 1989). In order to be highly appealing to a particular group, an organization must trade off in terms of performance (e.g., visitor numbers). This results in the situation where organizations have to make a choice between appeal and performance. Choosing for a single category, and therefore obtaining an audience that feels strongly appealed to the product, will result in a more limited audience and therefore lower performance. Concluding: targeting multiple categories will result in a larger audience (i.e. higher performance), although the disadvantage is that this audience feels less connected to the product (Hsu, 2006; Hsu, Hannan & Koçak, 2009; Levins, 1968). Despite the cons, category spanning is not a rare phenomenon within the creative industries. As said before, the industry shows a high frequency of category spanning (see, for example, Kuijken et al., 2016; Leenders, Van Telgen, Gemser & Van der Wurff, 2005). Since categorization and category spanning are dependent on the shared understandings of people, the audience plays a key role in these dynamics since organizations must conform to expectations of the audience in order to obtain legitimacy (DiMaggio & Powell, 1983; Meyer & Rowan, 1977; Suchman, 1995). Nevertheless, whether it is due to the presumed unobservability of diversity within audiences or due to the only low-level
conceptualizations (Hsu, 2006), former literature is dominated by the assumption that audiences have a shared perspective on categorization and the consequences they attach to it (Goldberg, Hannan & Kovács, 2016). Although categorization is dependent on a certain level of share of mind (DiMaggio, 1997; Goldberg, Hannan & Kovács, 2016), this does not mean that they also respond to the category-spanning activities in the same way. Innovation is the first signal that consumers could react differently to category spanning. According to the literature, innovation requires at least a minimal amount of category spanning activities since new and valuable ideas require someone to think out of the box and to make new connections between existing products (Glynn & Navis, 2013; King, Clemens & Fry, 2011). As argued by Phillips, Turco and Zuckerman (2013), because innovation is a category-spanning activity, the first reaction of consumers will be negative since the item shows inconsistency and therefore unreliability. Nevertheless, after a period of resistance the innovation, if valuable, will be accepted by a group of consumers who are open minded to novel and innovative ideas (Uzzi, Mukherjee, Stringer & Jones, 2013). This proves that rejecting category-spanning is not always the standard. The fact that there are at least some innovations that have been adopted, implies that there are consumers who are interested in these category-spanning activities.
This study tries to contribute to the exploration of a more elaborated view of a diverse audience concerning their behavior on category-spanning activities of organizations, especially in the creative industries. In this process, the focal point will be a distinction based on the consumption motivations of the audience. In order to understand consumption in general, one needs to look at the underlying consumer motives that drive consumption (Barbopoulos & Johansson, 2017). This study focusses on the field of music festivals as the empirical setting. It does so for several reasons. To begin with, categories have a very dominant and pervasive role in the music industry because
they offer quality guidelines, grounds for comparison and selection criteria (Paleo & Wijnberg, 2006; Schmutz, 2009; Van Venrooij, 2009). Second, the industry contains a very clear and observable categorization process, namely genres (Lena & Peterson, 2008). Third, music festivals are important in the daily form of cultural consumption these days (Paleo & Wijnberg, 2006). The growth of festivals is enormous. In 2017 alone, around 500 music festivals took place in The Netherlands. Dutch music festivals attract a total of 14.4 million visitors and their expenses contain in total € 513.1 million – 9% more than the year before (Consultancy.NL, 2017). Assuming that there is a difference in the offered products, this thesis expects to detect an observable difference in consumption motivations of visitors as well. The pressing question at this point is to what extent the category spanning effects on festival evaluation are influenced by the initial consumer motivations. If a person visits a music festival with the primary reason to attend a social event – as was found by Bowen and Daniels (2005) - does this have a moderating effect on how he will perceive category spanning? Does a person who attends a music festival because one particular genre will be played, react more negatively if a more diverse palette of genres is played? Does the reason to attend the festival has an impact on how visitors perceive category spanning? All these questions come together in the following research question: to what extent are the (adverse) effects of category spanning on evaluation by visitors dependent on their consumption motivations?
Restressing the impact of the consequences of categorization as mentioned above, more data and information about the reasons of audience members to visit a particular festival and the categorization that is involved during this process would be a considerable value to the development of the categorization theory in the creative industries. This study tries to develop knowledge that is actually useful for music festival
managers by creating better insights in the reasons why the consumer is attending the festival and to what extent this reason holds any consequences on how he/she will react on category spanning. Since music festivals are growing fast, a contribution to the music festival literature will help academia to generate a more elaborated view of the concept. In the subsequent section, a review of the literature about categorization, category spanning and customer motivations will be presented. In addition, a conceptual model will be developed in order to display the variables and their potential connection. Thereafter, the measures and data analysis will be described. Then, the findings and the discussion of the study will follow. This thesis will end with a limitation section and the conclusion.
2. Literature review
2.1. Categories and category spanning In all kinds of industries, categories are used to make sense of things, persons, concepts, events and other entities (DiMaggio, 1987; Douglas, 1986; Kuijken et al., 2016; Glynn & Navis, 2013; Negro, Koçak & Hsu, 2010; Ross, 1996; Wijnberg, 2011). Categories offer a framework in which one can place the subject in a bigger picture, differentiate it from other subjects and evaluate it (DiMaggio, 1997). Categories in markets span a multiplicity of literature within the sociology of organizations and institutional theory (DiMaggio, 1987; Hitters & Van de Kamp, 2010; Negro, Koçak & Hsu, 2010). Categories are important because they form the structures on which both organizations, i.e. the producers, and consumers obtain clarity and order of the market form (Durand & Khaire, 2017; Goldberg,Hannan & Kovács, 2016; Hsu, Koçak & Hannan, 2009; Rao, Monin & Durand, 2005). Categories help consumers to create familiarity with the product. In less abstract words, categories help in answering the questions ‘what is it’ or ‘what kind of thing is it’. Due to categories, one is able to categorize a bank as an organization, a festival as an event and Stayin’ Alive from the Bee Gees as a music song (Glynn & Navis, 2013).
In industries, like the creative industries, where products are predominantly hedonic and not possess clear and distinctive functionalities to differentiate and evaluate on, categories offer guidelines, reference points and standards of evolution (Ouellet, Savard & Colbert, 2008). In these industries, categories are essential in the process of gaining legitimacy from suppliers, consumers and other stakeholders (Hitters & Van de Kamp, 2010; Glynn & Navis, 2013; Singh, Tucker & House, 1986). Legitimacy contains the perception that something is "desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions" (Suchman, 1995: 574). Whether a category is broadly accepted and seen as legitimate depends on the adherence of stakeholders to the same conceptual scheme (Suarez, Grodal & Gotsopoulos, 2015).
Categories are socially constructed orderings that do only exists because a particular group of people share understandings about the subject (Rosa, Porac, Runser-Spanjol & Saxon, 1999). They organize the social space and the associated meanings with them (Durkheim & Mauss, 1963; Goldberg, Hannan & Kovács, 2016). Categories are present at multiple level of analysis; from the market level, to the organization level such as types of businesses and the product level containing specific types of products (Navis & Glynn, 2010). Since categories are socially constructed, there is a certain degree of fluidity. Therefore, it is possible that different stakeholder groups have different interpretations of the category (Kuijken et al, 2016). Even despite this fluidity categories
are related to a certain level of ‘share of mind’ (Kennedy, 2008) – even if not all information of the target object is known (DiMaggio, 1997).
Categorization is the process of dividing the market into different categories (Jacob, 2004; Suarez, Grodal & Gotsopoulos, 2015). Products that are easy to categorize, are simple for consumers to understand and will as a result attract more attention. A bad establishment in a category will lead to confusion by consumers and therefore to less attention (Zuckerman, 2004). Firms or products that do not fit into established and broadly accepted categories face the risk of being overlooked or ignored (Kennedy, 2008). Less attention leads to less market share. Many organizations are therefore triggered to conform their product to a particular category or reference frame in order for consumers will understand their product (Zuckerman, 2000). Easily categorizable products and organizations will benefit concerning the social, cultural and material resources that are made available to them (Negro, Koçak & Hsu, 2010).
In order to minimize problems and disputes about categorization, categories are constructed in such a way it is clear for the broad public what belongs to the category and what does not (Zuckermann, 1999). The categorical boundaries determine the ‘territory’ of a particular category. They indicate when one category ends and where the other starts. The notion must be made that these boundaries are dependent on ‘the share of mind’ of stakeholders as well, so they are not fixed lines of demarcation. There are cases of fuzzy boundaries which lead to a lack of clarity among stakeholders (Lamont & Molnar, 2002; Rao, Monin & Durand, 2003; Wijnberg, 2011). When a product crosses these boundaries and as a result spans multiple categories, one speaks of a category-spanning product (Wijnberg, 2011). These category-spanning activities could lead again to a lack of clarity among stakeholders. This is because it becomes more difficult to make sense of. Hsu, Negro and Perretti (2012) argue that the
producing organization of the product can wait for significant appeal penalties when it does not stick to the established categorical boundaries. Consumers find themselves less appealed to the category-spanning product since this product does not match the expectations they had before they made the purchase decision. It has become more difficult for consumers to legitimize and understand the product within a certain category because the product spans categories. The category spanning has led to fuzziness and ambiguity regarding the reference frame of the product (Hsu, 2006; Pontikes, 2012; Rao, Monin & Durand, 2005; Zuckerman, 1999).
Nevertheless, previous research has also shown that the number of categories a product spans could have positive outcomes for its performance and market success (Hsu, 2006; Kuijken et al., 2016). The more categories a product spans the more performance (in terms of, for example, box office returns) it generates (Hsu, 2006). Targeting multiple categories, will lead to a greater audience rate. Category spanning in unstable environments is even seen as risk reducing, since producers can spread the risk over multiple categories and compensate low turnovers in one category with the higher turnovers in the other (Hannan & Freeman, 1977). However, this statement is contradicted by, among others, Hitters and Van de Kamp (2010) and Kuijken et al. (2016) who state that category spanning leads to unoptimized adaptation. This is problematic since the finite resources of producing organizations (e.g., time workers, financials, etc.) must be divided over all these categories. The product’s quality will turn out to be lower because instead of devoting all the resources to one category, more categories have to deal with lesser resources – and therefore a decreasing performance (Dobrev, Kim & Hannan, 2001; Hannan & Freeman, 1989; Hsu, 2006; Levins, 1968; Negro & Leung, 2013). As DiMaggio (1992) claimed as well, an organization’s ability to access these material and symbolic resources are mediated by the meaning that audiences hold about them. In the
end, as mentioned above, the audience’s ability to make sense of the product determines its legitimacy and therefore its ability to attract resources (DiMaggio, 1987, 1992; Glynn & Navis, 2013; Pontikes, 2012). This debate on the effect of category spanning on the valence of how consumers evaluate the appeal and performance of the product is formulated in the following hypotheses:
H1 Spanning multiple categories will lead to lower appeal of the product by
consumers.
H2 Spanning multiple categories will lead to lesser performance of the product.
In multiple studies, a grade of the evaluation in total is asked because in some cases the audience holds negative feelings towards the appeal and performance of the product, although does shows these negative feelings in the total final judgement (Freling, Crosno & Henard, 2011). However, since former studies show such a strong negative effect of category spanning on evaluation, the final judgment could also act as a recap of the individual assessments of appeal and performance and therefore shows the same negativity. This expectation is formulated in the following hypothesis:
H3 Spanning multiple categories will lead to a lower evaluation grade of the
product.
Nevertheless, the question that prevails here is, if these unfavorable consequences are well known, why are there producers who try to span categories with their products? According to Goldberg, Hannan and Kovács (2016), category spanning is also seen as a form of atypicality seeking. In order to differentiate from competitors, producers are
looking for the vague areas of a categories or undiscovered category combinations. Category spanning makes it possible that anchored structures are broken and new possibilities are discovered. Multiple studies in the innovation literature suggest that category spanning is the first step towards innovation (Hsu, Negro & Perretti, 2012; Reid & de Brentani, 2004; Wijnberg, 2011). Innovation requires a certain level of ‘thinking out of the box’. Spanning categories is a powerful tool towards new opportunities for innovation (Glynn & Navis, 2013; King, Clemens & Fry, 2011). A study by Phillips, Turco and Zuckerman (2013) shows consumers adhere to a relatively negative assessment of category-spanning activities because to them it shows inconsistency and therefore unreliability (Ferguson & Hasan, 2013; Hsu, Hannan & Koçak, 2009; Kovács & Hannan, 2010). Nevertheless, Hannan (2010) argues that innovation requires organizations to provoke a certain level of spanning-aversion in order to be innovative at all. After a short period of resistance, the innovation will be accepted by a small group of consumers and later the mass will follow by adopting the innovation (Uzzi et al., 2013).
Although relatively negative assessments by consumers of these differentiation or innovation actions keep coming back in the literature, Goldberg, Hannan and Kovács (2016) claim that there is a group of consumers who practice and prefer category spanning. The fact that there are at least some innovations that have been adopted, implies that there are consumers who are interested in these category-spanning activities. Ollivier (2008) describes these consumers as being open-minded and having a preference for atypicality or a wide variety of categories. Durand and Khaire (2017) make a clear differentiation between the resistant consumers who prefer categorical purity and the consumers who are more omnivorous. Nevertheless, literature that dived into the effects of category spanning on appeal and performance and the so called principle of allocation generally neglects the fact that
there are also different consumer categories: the category omnivores and the category purists. As stated by Kovács and Liu (2016), there was a tradition to advocate a dominant role for the consumer in categorization processes, although there was no mention of any divergence between consumer attitudes. Only a few recent studies included this differentiation, also called consumer heterogeneity - the opposite of consumer homogeneity - in their research. See, for example, Kovács and Hannan (2010), Kovács and Sharkey (2014) and Pontikes (2012). They all showed consumers do not react as a homogeneous group to the possible ambiguity of category spanning, to reviews, to word-of-mouth or to other quality signals. 2.2. The role of consumer motivations This study focusses on the field of music festivals as the empirical setting. As mentioned in the introduction, music shows a clear categorization process in the form of genres. Genres are very useful in the process of assigning legitimacy, they offer grounds for comparison and quality guidelines (Paleo & Wijnberg, 2006; Schmutz, 2009; Van Venrooij, 2009). Paleo and Wijnberg (2006, 51) describe the music festival as the “key role [player by linking] artists to the consumers.”
The visitors of the festival, i.e. the audience, are the consumers of the product which is the festival itself. The business or group of people that organizes the festival is the producing organization. These organizations span categories with their product, i.e. the festival, when they program multiple genres on the same festival. However, at a certain point, organizations need to decide whether they follow a more generalists strategy, which means that they span multiple categories with their product and therefore attract more audiences, or to follow a specialists strategy, which will enable the organization to
master itself in that specific category (Hannan & Freeman, 1977; Hsu, 2006). Multiple studies promote the importance of the role of the product’s audience in this process (Kovács & Liu, 2016), because the organization’s ability to receive legitimacy and meaning for its product is mediated by the opinion of the audience (DiMaggio, 1992). Nevertheless, as stated before, the audience does not contain a homogenic set of preferences, conceptualizations, judgments and reflections. In reality, it is quite complicated and sometimes even neglected as well because of a presumed unobservability (Hsu, 2006).
Multiple studies have made a contribution to the mapping process of the role of audiences: Kuijken, Gemser and Wijnberg (2017) enriched the literature by analyzing the audiences’ frames of references which shows difference to the ones of the producing organizations; Leenders (2010) examined the degree of impact of brand equity aspects on the audience appeal; Oakes (2003) found support for the need for enhanced managerial awareness of the demographic profile of the audience; Leenders et al. (2005) focused on the effect of the role of format and content of the product on the product’s performance assigned by the audience, and so on.
Analyzing the differences between audiences regarding their preferences, conceptualizations, judgments and evaluations, is even more crucial in the creative industries since the typical product generally does not possess clear, distinct and objective functionalities to evaluate the product on. Therefore, these subjective value judgments become the basics on which the product and organization obtain legitimacy and value (Boorsma, 2006; Pearce, 2005; Priem, 2007; Wijnberg & Gemser, 2000). In order to understand these judgments and consumption in general, one must look to the underlying consumer motives that drive consumption (Barbopoulos & Johansson, 2017). Motives explain the internal processes of consumer behavior (Iso-Ahola, 1980). According to Sirgy, Rahtz and Portolese (2014) consumer motivations are “the drive to
satisfy needs and wants, both physiological and psychological, through the purchase and use of products and services.” Consumer motivations are created by the desire to achieve certain goals (Bagozzi, 1993). Like other social and cognitive constructs, these goals are subject to the environmental situation (Moskowitz & Grant, 2009). This means that consumer goals and therefore consumer motivations are not fixed but could change over time (Batra & Ahtola, 1990). Knowing what goals the consumer is looking for to achieve and what his/her motivations are, provides producers and the market in general with valuable information for their strategies (Barbopoulos & Johansson, 2017). There is a variety of literature about different consumer motivations for different type of consumptions (Pegg & Patterson, 2010). The creative industries are dominated by forms of hedonic consumption (Holbrook & Hirschman, 1982). Hedonic consumption means that consumers are more focused on the emotional aspects of products and are driven to consume goods that will have an influence on their mood and experience – as opposed to utilitarian consumption that consists of the fulfillment of rather basic requirements such as hunger, clothes and medical care (Dichter, 1960; Hirschman & Holbrook, 1982; Khan, Dhar & Wertenbroch, 2004; Lindenberg & Steg, 2007). The distinction between hedonic and utilitarian consumption is important because the nature of hedonic consumption is relatively unstable and short-sighted. This has a negative effect on marketing strategies such as pricing strategies and information campaigns (Barbopoulos & Johansson, 2017).
Literature about hedonic consumer motivations is termed as Motivation Research. During the 1950s, Ernest Dichter started this movement by emphasizing the emotional aspects of consumption. Until the 1980s, Dichter and other researchers had a hard time in defending this approach of consumer motivation since it’s rigor and validity would be questionable (Kassarijan, 1971; Wells & Beard, 1973; Hirschman & Holbrook, 1982).
Since the end of the 20th century, the number of studies on hedonic consumption has multiplied and has become more specific. These studies focus mainly on consumer motivations for cultural events. For example, Uysal, Gahan and Martin (1993) specified the consumer motivations to five dimensions: escape, excitement/thrills, event novelty, socialization and family togetherness. Crompton and McKay (1997) investigated the possibility of a better measurement tool to measure different consumption motives of events. They came up with six dimensions: cultural exploration, novelty/regression, recover equilibrium, known group socialization, external interaction/socialization, and gregariousness. Tomljenovic, Larsson and Faulkner (2001) created four motivations dimensions: to socialize, to party, to experience novelty and to attend multiple program parts. Bowen and Daniels (2005) conducted a small metanalysis and made an ordering of the most investigated consumer motivations of former studies. Based upon these former studies, three consumer motives were reoccurring. The pattern of dimensions that becomes visible throughout these studies is the focus on music, the focus on novelty and the focus on a (social) enjoyment. Bowen and Daniels (2005) conducted an explatorial factor analysis in order to find conformation for these dimensions. As it appeared, the most diverse consumer motivations can be clustered in the following three groups: “Enjoyment”, “Music” and “Discovery”. Noteworthy is that, as was mentioned by Bowen and Daniels (2005), although music is a form of culture and one can say culturally bound as well, none of the studies mentioned “culture” as a possible consumer motivation. According to their findings, the social group, which they called “Just Being Social”, turned out to show a low interest in the musical aspects of the festival. Their motivations were to spend time with friends and family, to have nice day out, to do something else than the daily routine, to party or to visit the non-musical attractions such as the food
trucks, amusement rides or markets. Regarding festivals with mainly these group of visitors, it is important to include other attractions or activities that are not immediately related to the main purpose of the festival. In addition, a part of the festival visitors appears to visit the festival because their friends or family go or have invited them. Emphasizing the social aspects of the festival would help in getting and retaining the attention of this group of visitors (Bowen & Daniels, 2005; Daniels & Norman, 2003).
The music group, named “The Music Matters”, scored, as one may expect, relatively high on the music-related dimensions. Motivations that popped out were ‘to see a particular band or artist’, ‘they liked the associated genre’ or ‘they wanted to listen to music in real life’. The interesting thing that emerged here was that this group of visitors scored very low on the items related to discovery. Bowen and Daniels (2005 described these visitors as the hard-core music fans.
As described by Bowen and Daniels (2005), items that loaded high on the discovery dimension ‘were to experience new and different things’ and ‘to increase knowledge about undiscovered music and culture’. The innovation literature argues that ‘thinking out of the box’ and ‘the wish to discover’ are primary drivers of innovation. At the same time, discovery implies a certain level of category-spanning activities (Hsu, Negro & Perretti, 2012; Reid & de Brentani, 2004; Wijnberg, 2011). Nevertheless, the enjoyment and discovery audience groups showed multiple similarities in their scores on the motivation items. Especially on the item of their interest in the music, on which both groups scored very low (Bowen & Daniels, 2005). This creates the expectation that enjoyment motivated consumers hold multiple similarities with the discovery motivated consumers.
According to the literature, as mentioned above, discovery motivated consumers, who are open to new experiences, are probably more open to innovation and therefore
more tolerant towards category spanning. It has not been investigated to date whether the enjoyment motivated consumers, who show multiple similarities with discovery motivated consumers, will have the same reaction on category spanning. To formulate it differently: are there consumers who are not predominantly driven by discovery motivations who show the same tolerance towards category spanning? At the same time, music motivated consumers did not show any similarities with the discovery group. They were only interested in the music related motivations specifically. This does not raise any expectations that this group will react in a more positive way on category spanning. On the contrary, consumers who are driven predominantly by music motivations can react even more negatively on category spanning since their prior motivations are related to the music categories they expect. This divergence in consumer motivations and how this relates to the effect of category spanning on the evaluation valence are formulated in the following hypotheses:
H4a The negative effect of category spanning on the evaluation appeal will be
stronger for music-motivated consumers.
H4b The negative effect of category spanning on the evaluation performance will
be stronger for music-motivated consumers.
H4c The negative effect of category spanning on the evaluation grade will be
stronger for music-motivated consumers.
H1 H2 H3 H4a H4b H4c H5a H5b H5c
H5a The negative effect of category spanning on the evaluation appeal will be
weaker for enjoyment-motivated consumers.
H5b The negative effect of category spanning on the evaluation performance will
be weaker for enjoyment-motivated consumers.
H5c The negative effect of category spanning on the evaluation grade will be
weaker for enjoyment-motivated consumers.
FIGURE 01 Conceptual model of the hypotheses (simplified)
FIGURE 02 Conceptual model of the hypotheses (extensive)
Category spanning Evaluation of appeal Evaluation of performance Evaluation grade Music motivated Enjoyment motivated Category spanning Evaluation valence Consumer motivation
3. Data and method
3.1. Research design The study was conducted in the form of a survey. The population of interest was visitors of Dutch music festivals. A probability sampling technique is used in order to retrieve the sample. Because consumers vary in terms of knowledge about music genres, this study tried to control for a minimal level of knowledge by selecting the respondents on the condition that they have visited at least one Dutch music festival in the past twelve months (Kuijken et al., 2016). Subsequently, it was made clear to respondents that the survey questions that followed would focus on their last visited music festival. For this festival their memories were the freshest and therefore the expectation was that these answers would be the most reliable. Respondents were approached by using the researcher’s personal and professional network through Facebook, LinkedIn and telephone numbers. Data was collected during twenty days in May, 2018. Because the empirical setting for this study is the Dutch music festival, the survey took place in Dutch. Nevertheless, no question of the respondent’s nationality was included, so as long as respondents mastered the Dutch language they could participate in the survey. Therefore, it cannot be excluded that there were non-Dutch participants. Since this study is looking for differentiation between customer motivations in general and not specifically for these differences between nationalities, this was not regarded as a constraint.The study contains a cross-sectional designed survey. The survey consists of open -, list – and rating questions. The data was collected through the digital survey tool ‘Qualtrics’ and thereafter exported and analyzed in ‘IBM SPSS Statistics’. First, the dataset
was checked for cases with missing data. It appears, 226 times the survey was opened, 170 respondents completed it entirely. In total, 56 cases were deleted. Eight of them had closed the survey after a number of questions. 46 times the survey was opened but the responded did not click on the next step. Two times the survey was closed after a number of questions. Two respondents completed the survey from a concert perspective; they were also deleted. After this, the cleaned dataset was checked for any errors, but no errors were found. 3.2. Operationalization of the variables 3.2.1. Dependent variable Evaluation valence: As elaborated in the literature review, the evaluation valence consists of three parts: the appeal, the performance and the general grade. The grade acts as a figurative umbrella that includes evaluation appeal and evaluation performance and is used to measure a more general evaluation of the festival in total. Appeal: The validated seven-point bi-polar adjectives scale of Freling, Crosno and Henard (2011) is used to measure evaluation appeal. In their two studies the Cronbach’s alphas were respectively 0.95 and 0.93. An example item is: “How did you experience the festival in general? Unpleasant / pleasant.” In order to detect any acquiescence biases, two items of the were stated reversed. Recoding was applied to these counter-indicative items. This means that rEva4 and rEva5 now represent Eva4 and Eva5. A Principal Axis Factoring (PAF) was also conducted on the six evaluation items that represented appeal in order to determine if it resulted as one dimension. The Keiser-Meyer-Olkin measure verified the sampling adequacy for the analysis, KMO = 0.811. Bartlett’s test of sphericity χ2 (15) = 439.239 p < 0.001, indicated that correlations
between items were sufficiently large for PAF. An initial analysis was run to obtain Eigenvalues for the component in the data. One component had an Eigenvalue over Kaiser’s criterion of 1 and in explained 48.42% of the variance. In agreement with Kaiser’s criterion, examination of the scree plot revealed a levelling off after the first factor.
In order to test the internal reliability and consistency of the scale, the Cronbach’s alpha was measured. The scale has a high reliability (α = 0.839). This indicates that all the items have a good correlation with the total score of the scale (Nunnally, 1978).
Performance: For the evaluation valence that represented performance, the net-promotor score was used. The net-promotor score is used to get more insights into the division of the customer who would recommend the product and customers who would discommend the product to family and friends (Kerin, Jain & Howard, 1992; Reichfeld, 2003). In order to grow in performance, an organization needs more recommenders and less discommenders. The lower the score, the more someone is a discommender. The higher the score, the more someone is a recommender. The most effective manner to get the net-promotor score is to ask the question: How likely is it that you would recommend organization/product X to family or friend? There are differentiations to this question possible, although those were not that efficient in predicting the actual customer behavior (Reichfeld, 2013). Nevertheless, the net-promotor score is not perfect as well. It is, for example, not able to measure the drivers behind the customer’s reaction. Thus, it will not give the organization any insights into what specific items caused the low or high score (Keiningham, Cooil, Andreassen & Aksoy, 2007). Grade: A general assessment in the form of a grade was asked in order to measure the total evaluation of the festival (Bathra & Ahtola, 1991). The descriptive information of these variables can be found in table 04.
3.2.2. Independent variables
Category spanning: This variable was measured by asking the respondents to classify the genre of the festival they had visited – as was done in Kuijken et al. (2016). A list of seventeenth genres and an option to fill in not-mentioned genres were provided in order to give the results more structure and avoid a gigantic varied outcome. These seventeen genres are based upon the outcome of a multi-step process, as was done in Rentfrow and Gosling (2003). First, a list of genres and well-known subgenres was generated based upon the genre list of the popular streaming service Spotify (http://spotify.com). In order to ensure the list was varied adequately, multiple websites of radio stations (e.g., http://npo3fm.nl, http://nporadio2.nl, http://538.nl, http://qmusic.nl, http://slam.nl, etc.) consulted to enlarge the initial list. A list of 20 genres and subgenres were gathered through this procedure. After this, the selection has been assessed by a group of 40 individuals in order to check if the list would provide participants of the survey with a proper representation of the genres played at Dutch festivals. This resulted in a list with the following seventeen genres, subgenres or groups of genres: classical, blues, jazz/soul, world music/country/folk, schlager/‘levenslied’, mainstream/pop, indie/alternative, rap/hip-hop, house, techno, disco, drum ‘n bass, reggae/ska, rock, hard rock/metal, hardstyle and psychedelic trance/tekno. An option ‘other’ was added to enable participants to add an extra genre if they thought this would describe the festival they had visited in a better way. The number of genres that respondents linked to the festivals are counted, see table 01 (Kuijken et al, 2016).
Respondents were asked as well to fill in the name of the festival. It makes sense that a larger festival will span more categories because it has more stages, more visitors and therefore creates a bigger line-up for a larger number of visitors – all compared to a smaller festival. This does not necessarily mean that this immediately leads to uncertainty
among the visitor regarding the categories. The most obvious thing would be to divide the number of acts of the festival by the number of genres assigned. Unfortunately, this information was only known for a limited number of festivals. Because the study focused on festivals in 2017, many festivals websites had already been adapted to the 2018 edition. The second option is to share the number of visitors by the number of genres. To control for the size of the festival, the number of visitors is a clear indication of the size. In order to retrieve the visitor numbers, websites of all the mentioned festivals have been used. The formula used for the standardized category spanning variable is: X = sum number of genres / number of visitors * 100,000. In order to create a workable number, the outcome has been multiplied by 100,000. Additionally, the final outcome of the formula is rounded to 1 decimal he outcome of this has been rounded to 1 decimal.
TABLE 01 Sum of the number of the assigned genres (= category spanning)
Frequency Valid percent Cumulative percent
Valid 1 genre 35 20.5 20.5 2 genres 31 18.1 38.6 3 genres 29 17.0 55.6 4 genres 27 15.9 71.3 5 genres 19 11.1 82.5 6 genres 9 5.3 87.7 7 genres 8 4.7 92.4 8 genres 4 2.3 94.7 9 genres 4 2.3 97.1 10 genres 1 0.6 97.7 11 genres 2 2.3 98.8 12 genres 2 2.3 100.0 Total 171 100.0 Consumption motivation: Based upon the study conducted by Bowen and Daniels (2005), a modified list of eight consumer motivations for festivals was presented to the respondents. Since Bowen and Daniels (2005) conducted an explatorial study on a wide
variety of consumer motivations, they included motivation statements that were not that relevant for this particular study. However, their initial Cronbach’s alphas for the different motivation dimensions ranged from 0.65 to 0.88. Since this study is looking for a distinction between enjoyment motivated consumer and music motivated consumers, it has only used the motivations that were relevant to these dimensions. The particular items can be found in the tables below.
A Principal Axis Factoring (PAF) was conducted on the eight consumption motivations items to determine if dimensions of motivation emerge as expected. The Keiser-Meyer-Olkin measure verified the sampling adequacy for the analysis, KMO = 0.587. Bartlett’s test of sphericity χ2 (28) = 138,091, p < 0.001, indicated that correlations
between items were sufficiently large for PAF. An initial analysis was run to obtain Eigenvalues for each component in the data. Two components had Eigenvalues over Kaiser’s criterion of 1 and in combination explained 28.17% of the variance. In agreement with Kaiser’s criterion, examination of the scree plot revealed a levelling off after the second factor. Thus, two factors were retained and rotated with an Oblimin with Kaiser normalization rotation. According to Stevens (1992), items with a loading less than 0.30 should be deleted for the sake of a better interpretation. Therefore, two items (item 1 and 7) were deleted for further analysis. Table 02 shows the factor loadings after rotation. The items that cluster on the same factors are interpreted as factor 1 representing music-motivated consumers and factor 2 representing enjoyment-motivated consumers.
In order to test the internal reliability and consistency of the scales, for all scales the Cronbach’s alpha was measured. According to Cortina (1993), for scales with less than six items, an alpha score of 0.60 could be considered acceptable (Bowen & Daniels, 2005). The Cronbach’s alphas were respectively for Music α = 0.627 and for Enjoyment α = 0.501.
Nevertheless, these values will be used for further analysis due to the lack of better results. This will be elaborated in the discussion.
TABLE 02 Factor analysis of Consumer Motivation
Rotated factor loadings
Item Music Enjoyment
1. Friends/acquaintances went -0.34 0.00
2. To meet new people 0.07 0.45
3. The (expected) atmosphere -0.06 0.40
4. A way to escape daily life 0.07 0.43
5. Experience nonmusical attractions -0.09 0.55
6. Because of the genre of the artists/bands 0.52 0.10
7. Because of the presence of an artist/band 8. To discover new music
0.93
0.25
-0.18 0.26
Factor loadings over 0.30 appear in bold (Stevens, 1992).
TABLE 03 Overview Factor analysis of Consumer Motivation
Item Loading Eigenvalue
Variance Explained
Cronbach’s Alpha
Music
Because of the genre of the artists/bands Because of the presence of an artist/band
Enjoyment
To meet new people The (expected) atmosphere A way to escape daily life
Experience nonmusical attractions
Total 0.52 0.93 0.45 0.40 0.43 0.55 2.008 1.478 17.608 10.566 28.174 0.627 0.501 3.2.3. Control variables
The study includes multiple control variables in order to increase the reliability and exclude the influence of any control variable on the dependent variable. Gender: the respondents were asked to indicate if he/she was a male or a female. The option of ‘different’ also exists, although no one used this option. Therefore, the option ‘different’ was excluded for further analysis. Age: the respondents were asked to fill in his/her age in numbers in years. Education: the respondents were asked to tick the box of his/her highest completed education. Five Dutch education levels formed the available options:
no education, vmbo, mbo, havo/vwo and hbo/wo. There were no respondents who ticket the first box of no education and the second box of vmbo. Therefore, these two options were excluded for further analysis. Employment status: the respondents were asked to check the box that resembled their employment status. Options for this question were: predominantly studying, predominantly working, both studying and working and not applicable. Program multiple days: in order to check if the results apply to single-day festivals as well as multi-day festivals, respondents were asked to indicate if the festival they evaluated lasts for one or multiple days or that they were not aware of it. Visited before: the last control variable asked if the respondent to indicate if he/she has visited the festival before. As one may expect, when someone has paid a visit in the past, he could hold certain expectations and this could influence his evaluation.
TABLE 04 Descriptive statistics (N=170)
Min Max Mean Std. deviation
Gender (0=male, 1=female) 0 1 0.62 0.49 Age (in years) 18 59 24.68 5.88 Education (0=mbo, 1=havo/vwo, 2=hbo/wo) 0 2 1.78 0.55 Employment status (0=studying, 1=working, 2=both, 3=NA) 0 3 0.69 0.79
Program multiple days (0=yes, 1=no, 2=unknown) 0 2 0.29 0.47 Visited before (0=yes, 1=no) 0 1 0.55 0.50 Evaluation Appeal (0=low, 7=high) 1.83 7.00 5.58 1.07 Evaluation Performance (0=unlikely, 7=likely) 1 7 5.89 1.28 Evaluation Grade (0=inferior, 10=excellent) 2 10 7.84 1.00 Consumption motivation Music motivated (0=low, 5=high) Enjoyment motivated (0=low, 5=high) 1 1 5 5 3.56 3.17 1.04 0.68 Number of genres 1 12 3.63 2.43
TABLE 05 Correlation table (N=170)
** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed).
Variables Min Max M SD α 1 2 3 4 5 6 7 8 9 10a 10b 11
1. Gender (0=male, 1=female) 0 1 0.62 0.49 - 2. Age (in years) 18 59 24.68 5.88 -0.54 - 3. Education
(0=mbo, 1=havo/vwo, 2=hbo/wo)
0 2 1.78 0.55 0.06 0.06 -
4. Employment status
(0=studying, 1=working, 2=both, 3=NA)
0 3 0.69 0.79 0.10 0.26** -0.05 -
5. Program multiple days (0=yes, 1=no, 2=unknown)
0 2 0.29 0.47 -0.05 -0.04 0.07 -0.12 - 6. Visited before (0=yes, 1=no) 0 1 0.55 0.50 0.06 -0.11 -0.17* -0.11 0.22** - 7. Evaluation Appeal (0=low, 7=high) 1.83 7.00 5.58 1.07 0.839 -0.03 0.00 0.04 0.10 -0.18* -0.19* - 8. Evaluation Performance (0=unlikely, 7=likely) 1 7 5.89 1.28 0.09 -0.04 0.01 0.11 -0.27** -0.23** 0.57** - 9. Evaluation Grade (0=inferior, 10=excellent) 2 10 7.84 1.00 0.02 -0.02 0.02 -0.03 -0.20** -0.20** 0.52** 0.62** - 10. Consumption motivation a. Music motivated (0=low, 5=high) b. Enjoyment motivated (0=low, 5=high) 1 1 5 5 3.56 3.17 1.04 0.68 0.627 0.501 -0.04 0.05 -0.01 -0.02 -0.14 -0.04 0.02 0.01 -0.19* -0.28** 0.09 -0.21** 0.10 0.06 0.26** 0.21* 0.18* 0.13 - 0.10 - 11. Number of genres 1 12 3.16 2.43 -0.05 0.15* 0.11 -0.01 0.17* 0.05 0.02 0.00 0.01 -0.16* -0.11 -
4. Results
4.1. Descriptive statistics
The results show that the majority of the respondents: were female (61%) with an average age of 25 years, were holding a Bachelor’s degree (84%) and were having the employment status of mainly studying (50%; see table 04).
A total of 45% of the study respondents stated they had visited the festival they reviewed more than once. The top three of most mentioned festivals in this survey is: Lowlands (12%), Zwarte Cross (10%) and Grasnapolsky (7%; see the appendix). 71% of the particular festivals had a duration of more than one day. Almost the majority of the respondents (49%) on multiple-days festivals visited these festivals all days. Table 05 presents the correlation matrix of all variables. First, the table shows there is significant small positive correlation, r(170) = 0.26, p < 0.01)1, between employment
status and age. This is logical since the higher the age, the higher the probability that someone has a job or has a job and studies at the same time. This is because, in general, education takes place during the younger years of an individual. Second, the table shows a significant small negative correlation, r(170) = -0.17, p < 0.05), between visited before and education. This means that the lower the level of someone’s education, the more he/she has visited the festival before. Third, the table shows there is a significant small positive correlation, r(170) = 0.22, p < 0.01), between visited before and program multiple days. As someone has visited a festival before, the more often this festival lasts for more
0.18, p < 0.05), between evaluation appeal and program multiple days. This means the more appeal someone’s feels to the festival, the less long the festival lasts. This means, there could be a tension curve of the audience. Or, the longer the festival lasts, the more acts it has programmed, the lower appeal someone feels with the festival since it spans multiple categories. Although this is just guessing and should be apparent from further analysis. Fifth, the table shows a significant small negative correlation, r(170) = -0.19, p < 0.05), between the evaluation appeal and visited before. The greater the appeal, the less often someone has visited the festival before. Sixth, the table shows a significant negative correlation, r(170) = -0.27, p < 0.01), between evaluation performance and program multiple days. The lower the evaluation of the performance, the less often the festival lasts for multiple days. This means that the likeability that someone recommends the festival to friends or family could depend on the length of his tension curve. Seventh, the table shows a significant negative relationship, r(170) = -0.23, p < 0.01), between the evaluation performance and visited before. The more positive the evaluation of the performance, the less often someone has visited the festival before. Eight, the table shows a significant positive relationship, r(170) = 0.57, p < 0.01), between evaluation performance and evaluation appeal. This means that both constructs refer to an equal evaluation and relate to each other. Ninth, the evaluation grade has a significant negative correlation with program multiple days, r(170) = -0.20, p < 0.01), and visited before, r(170) = -0.20, p < 0.01). The more positive the evaluation grade, the more often the festival was not held over several days and was not visited before. This is in line with former negative relationships between evaluation variables and the length of the festival and if the respondent has visited the festival before. Tenth, the evaluation grade, which should function as a figurative umbrella that includes evaluation appeal and evaluation performance, has a significant positive correlation with evaluation appeal, r(170) = 0.52,
p < 0.01), and evaluation performance, r(170) = 0.62, p < 0.01). Eleventh, the table shows a significant negative correlation, r(170) = -0.19, p < 0.05), between music motivated and program multiple days. The more someone is music motivated, the shorter the festivals lasts. Twelfth, the table shows a significant positive correlation, r(170) = 0.26, p < 0.01), between music motivated and evaluation performance, and a significant positive correlation, r(170) = 0.18, p < 0.05), between music motivated and evaluation grade. The more someone is music motivated, the more positive its evaluation grade of the festival is. Thirteenth, the table shows that enjoyment motivated has a significant negative correlation, r(170) = -0.28, p < 0.01), with program multiple days, a significant negative correlation, r(170) = -0.21, p < 0.01), with visited before, and a significant positive correlation, r(170) = 0.21, p < 0.05) with evaluation performance. The more often people are enjoyment motivated, the shorter the festival lasts, the less often they have visited the festival before and the more they are positive in their performance evaluation. Fourteenth, the table shows that the number of genres has a significant positive correlation, r(170) = 0.15, p < 0.05), with age, a significant positive correlation, r(170) = 0.17, p < 0.05), with program multiple days. The older someone is, are, the more often they go to festivals with multiple genres and festivals that last longer than one day. Finally, there was a significant negative correlation, r(170) = -0.16, p < 0.05), between number of genres and music motivated consumer motivations. The more categories a festival spans, the less likely someone is music motivated.
Additionally, the table shows that there is no significant correlation between the independent variable category spanning (i.e. number of genres) and the dependent variables evaluation appeal, evaluation performance and evaluation grade. There is no significant correlation as well between the dependent variable category spanning (i.e. number of genres) and the moderating variable enjoyment motivated.
Regarding the control variables gender and employment status, no significant correlations were found with the independent and dependent variable. 4.2. Testing the hypotheses Before testing the hypothesis by means of a hierarchical regression analysis, a scatter plot was run in order to test for heteroscedasticity or homoscedasticity. The scatter plot showed that there was little connection to trace visually (see appendix b). Nevertheless, the regression will be run with the available data. For every dependent variable, namely 1) evaluation appeal, 2) evaluation performance and 3) evaluation grade, individual regression analyses are conducted. Tables 06, 07 and 08 present these regression models that predict the dependent variables for our sample (N=170). The following text will discuss each hierarchical regression analysis per dependent variable separately. Table 06 shows the regression of evaluation appeal, Table 07 shows the regression of evaluation performance and Table 08 shows the regression of evaluation grade.
4.2.1. Hypothesis 1
The first hypothesis expects a negative relation between category spanning and evaluation appeal. A Pearson correlation analysis was used to test the presence of a negative relation between the variables category spanning and evaluation appeal. As shown in table 05, no significant correlation was found between the two variables. Category spanning and evaluation appeal are positive correlated (r = 0.02) and not significant.
TABLE 06 Hierarchical regression analysis of evaluation appeal
Dependent variable: Evaluation appeal Model 1 Model 2 Model 3 Model 4
β SE β SE β SE β SE
Gender (0=male, 1=female) -0.04 0.17 -0.04 0.17 -0.03 0.17 -0.03 0.17
Age (in years) -0.05 0.02 -0.06 0.02 -0.06 0.02 -0.06 0.02
Education (0=mbo, 1=havo/vwo, 2=hbo/wo) 0.03 0.16 0.02 0.16 0.02 0.16 0.02 0.16
Employment status (constant=studying)
Employment status working 0.05 0.22 0.04 0.22 0.04 0.22 0.04 0.22
Employment status both 0.08 0.23 0.08 0.24 0.08 0.24 0.08 0.24
Employment status NA 0.04 0.82 0.04 0.83 0.04 0.83 0.04 0.83
Program multiple days (constant=yes)
Program multiple days: no -0.15 0.19 -0.16* 0.19 -0.16 0.19 -0.16 0.19
Program multiple days: unknown -0.01 1.07 -0.01 1.08 -0.01 1.08 -0.01 1.08
Visited before (0=yes, 1=no) -0.14 0.18 -0.15 0.18 -0.15 0.18 -0.15 0.18
Number of genres -0.07 0.00 0.08 0.00 0.08 0.00 Music motivated 0.09 0.08 0.09 0.08 Enjoyment motivated -0.03 0.08 R Square 0.06 0.07 0.07 0.07 Adjusted R Square 0.01 0.01 0.00 0.00 N 170 170 170 170 *p < 0.05
TABLE 07 Hierarchical regression analysis of evaluation performance
Dependent variable: Evaluation performance Model 1 Model 2 Model 3 Model 4
β SE β SE β SE β SE
Gender (0=male, 1=female) 0.08 0.20 0.08 0.20 0.08 0.20 0.08 0.20
Age (in years) -0.07 0.02 -0.08 0.02 -0.09 0.02 -0.09 0.02
Education (0=mbo, 1=havo/vwo, 2=hbo/wo) 0.00 0.18 0.00 0.19 -0.01 0.19 -0.01 0.19
Employment status (constant=studying)
Employment status working 0.07 0.52 0.06 0.25 0.06 0.25 0.06 0.25
Employment status both 0.08 0.57 0.09 0.27 0.09 0.27 0.09 0.27
Employment status NA -0.04 0.95 -0.04 0.95 -0.03 0.95 -0.04 0.95
Program multiple days (constant=yes)
Program multiple days: no 0.18* 0.22 -0.19* 0.22 -0.19* 0.22 -0.19* 0.22
Program multiple days: unknown -0.17* 1.23 -0.17* 1.24 -0.17* 1.24 -0.17* 1.24
Visited before (0=yes, 1=no) -0.19* 0.20 -0.19* 0.20 -0.19* 0.20 -0.19* 0.21
Number of genres 0.06 0.01 0.08 0.00 0.12 0.00 Music motivated 0.05 0.09 0.02 0.09 Enjoyment motivated 0.10 0.09 R Square 0.14 0.11 0.15 0.15 Adjusted R Square 0.09 0.09 0.09 0.09 N 170 170 170 170 *p < 0.05
TABLE 08 Hierarchical regression analysis of evaluation grade
Dependent variable: Evaluation grade Model 1 Model 2 Model 3 Model 4
β SE β SE β SE β SE
Gender (0=male, 1=female) 0.02 0.16 0.02 0.16 0.03 0.16 0.01 0.16
Age (in years) -0.04 0.02 -0.05 0.02 -0.06 0.02 -0.06 0.02
Education (0=mbo, 1=havo/vwo, 2=hbo/wo) 0.01 0.15 0.00 0.15 0.00 0.15 0.00 0.15
Employment status (constant=studying)
Employment status working 0.18* 0.20 -0.01* 0.21 -0.02* 0.21 -0.02* 0.21
Employment status both -0.07 0.22 -0.07 0.22 -0.07 0.22 -0.06 0.22
Employment status NA -0.02 0.77 -0.02 0.77 -0.02 0.77 -0.03 0.77
Program multiple days (constant=yes)
Program multiple days: no -0.16 0.18 -0.17 0.18 -0.16 0.18 -0.16 0.18
Program multiple days: unknown -0.07 1.00 -0.06 1.00 -0.07 1.00 -0.07 1.00
Visited before (0=yes, 1=no) -0.16* 0.16 -0.16* 0.17 -0.16* 0.17 -0.16* 0.17
Number of genres 0.05 0.00 0.08 0.00 0.13 0.00 Music motivated 0.09 0.07 0.05 0.08 Enjoyment motivated 0.13 0.07 R Square 0.07 0.07 0.08 0.09 Adjusted R Square 0.02 0.02 0.02 0.02 N 170 170 170 170 *p < 0.05
In the first model of the hierarchical regression all the control variables are included, these are: gender (male, female), age (in years), education (divided in the Dutch education levels of mbo, havo/vwo, hbo/wo). Although education is an ordinal variable, errors in the analysis appeared after the inclusion of its dummy variables. Therefore, it is chosen to use education as a continuous variable in further analysis. The other control variables, used as dummy variables, are: employment status (studying, working, both, NA), if the festival lasts for multiple days (yes, no, unknown) and if the respondent had visited the festival before (yes, no). Regarding table 06, concerning the dependent variable evaluation appeal, Model 1 shows no significant associations. Additionally, Model 1 explained 1% of the variance but was statistically not significant F(9, 160) = 1.21, p = 0.30. Model 2 contains the main effect. Therefore, the variable category spanning was added. The explained variance by Model 2 contained 1%, although not significant. The model shows that if the festivals last for only one day, the audience will be lesser appealed (β = -0.16; p < 0.05). Since this is a dummy variable, and the constant is that the festival lasts for multiple days, this means that the evaluation appeal will increase if the duration of the festival is multiple days compared to one day. Nevertheless, model 2 was statistically not significant F(1, 159) = 0.72, p = 0.40.
Concluding, hypothesis 1 that expects a negative effect of category spanning on evaluation appeal will be rejected.
4.2.2. Hypothesis 2
The second hypothesis expects a negative relation between category spanning and evaluation performance. A Pearson correlation analysis was used as well to test the presence of a negative relation between the variables category spanning and evaluation
performance. As shown in table 05, no significant correlation was found between the two variables (r = 0.00).
Table 07 shows the hierarchical regression analysis of the dependent variable evaluation performance. The same variables are added in the same way as was done in table 06. In order to avoid repetition, only the results will be discussed. Model 1 shows that the duration of the festival has an influence on the evaluation performance. Although the wording is a bit counterintuitive, the model shows that if the duration of the festival changes from multiple days to one day the evaluation performance will increase (β = 0.18; p < 0.05). Additionally, if the duration of the festival changes from multiple days to a duration of which the respondent does not know, the evaluation performance will decrease (β = -0.17; p < 0.05). Last, if the respondent has not visited the festival before, compared to a respondent who has visited the festival before, the evaluation performance will decrease (β = -0.19; p < 0.05). On top of that, Model 1 explained 9% of the variance and was statistically significant F(9,160) = 2.88, p < 0.05.
Model 2, which included the main effect, shows again associations between the control variables and the dependent variable evaluation performance. Compared to a festival duration of multiple days, a festival that lasts only one day has a negative influence on the evaluation performance (β = -0.19; p < 0.05), and a festival with a duration of which the respondent does not know has a negative influence on the evaluation performance as well (β = -0.17; p < 0.05). Compared to a respondent who has visited the festival before, a respondent who visited the festival for the first time will have a negative influence on the evaluation performance (β = -0.19; p < 0.05). Model 2 explained 9% of the variance as well, although this was not significant F(1,159) = 0.60, p = 0.44.
Concluding, hypothesis 2 that expects a negative effect of category spanning on evaluation performance will be rejected.
