The Effect Of Different Types of Transparent Marketing
On Product Valuation
Master thesis
By Jet van Strijp (10544143) M.Sc. in Business Studies – Marketing
University of Amsterdam, Faculty of Business and Economics June 30th, 2014
Abstract i
Abstract
Marketing as a discipline has changed over the past few decades and is still undergoing changes. Now it seems like the next marketing era has dawned, namely the transparency era. This research aims to define transparent marketing and its effect on actual behaviour. More specifically, the objective of this study is to determine the effect of different types of transparent marketing on product valuation. Literature on two-sided messaging, information disclosure and ethical marketing served as a starting point for this research. In order to collect data, a field experiment in the form of an online second price auction was conducted in which participants indicated their maximum willingness-to-pay as a bid. This provided insights on the effect of different types and sources of transparent marketing on product valuation. Moreover, the reliability of a scale for the perceived importance of transparency was tested. The results showed that, firstly, no transparency results in a higher willingness-to-pay than disclosure of transparent information by the firm. Secondly, it was found that the combination of cost/price transparency and disclosure by a consumer leads to a significantly higher product valuation. Based on these findings implications to both academics and practitioners can be made. To begin with, academics could use the conceptualisation of transparent marketing that was established in this research as a foundation for further research. Secondly, practitioners should be aware of the scepticism that dominates among consumers and therefore not make the firm itself disclose transparent information. To conclude with, further research is required in order to provide more credible and comprehensive knowledge on the topic of transparent marketing.Table of contents ii
Table of contents
Abstract ... i
List of figures and tables ... iv
Introduction ... 1
Theoretical framework ... 4
Transparency ... 4
Two-sided messaging, information disclosure and ethical marketing ... 4
Trust, honesty and fairness ... 6
Transparent marketing ... 7
Requirements of transparent information ... 8
Types of transparency ... 8 Source of transparency ... 9 Consumer behaviour ... 11 Information avoidance ... 11 Actual behaviour ... 13 Product valuation ... 14 This study ... 15 Method ... 17 Experiment design ... 17 Pre-test ... 17 Operationalisation ... 22 Procedure ... 24 Veylinx ... 24 Product ... 25 Stimuli ... 25 Experiment ... 25
Table of contents iii
Results ... 26
Demographics ... 26
Tests of normality ... 27
Tobit regressions ... 28
Discussion and conclusion ... 36
General discussion ... 36
Implications ... 38
Limitations and suggestions for future research ... 39
References ... 41
Appendix ... 48
Appendix I. Pre-test ... 48
List of figures and tables iv
List of figures and tables
Figure 1. Conceptualisation of transparent marketing ... 8
Figure 2. Typology of transparency for marketing management research ... 9
Figure 3. Framework for understanding information avoidance decisions ... 13
Figure 4. Means plot positive/negative score per statement ... 20
Figure 5. Means plot transparency score per statement ... 21
Figure 6. Histogram of bid amounts ... 29
Table 1. ANOVA of relevance means per type of transparency ... 18
Table 2. Contrast tests of relevance means per type of transparency ... 19
Table 3. Relevance means per type of transparency ... 19
Table 4. Contrast test statement 1 and 21 on positive/negative score ... 22
Table 5. Contrast test statement 1 and 21 on transparency score ... 22
Table 6. Schematic representation of five conditions ... 22
Table 7. Frequencies of gender ... 26
Table 8. Descriptives of age ... 26
Table 9. Frequencies of higher education ... 26
Table 10. Frequencies of students ... 27
Table 11. Frequencies of familiarity with Hero ... 27
Table 12. Tests of normality original bids ... 28
Table 13. Tests of normality bids with algorithm applied ... 28
Table 14. Descriptives bid amount with algorithm applied ... 28
Table 15. Tobit regression bid amount and type and source of transparency ... 30
Table 16. Tobit regression bid amounts and true treatments ... 31
Table 17. Tobit regression bid amount and type and source of transparency with age, student and gender taken into account ... 32
Table 18. Tobit regression bid amount and true treatments with age, student and gender taken into account ... 33
Table 19. One-way ANOVA transparency scale ... 33
Table 20. Reliability of transparency scale ... 34
Table 21. Reliability of alternative transparency scales ... 34
Introduction 1
Introduction
“Soon it will be true that every activity of the corporation—from finance to sales to production—is aimed at satisfying the needs and desires of the consumer. … When that stage of development is reached, the marketing revolution will be complete.” This is what Keith (1960) concludes in his article on the revolution from production to sales to marketing. Although indeed different activities of firms are increasingly aimed at satisfying the needs and wants of customers, it can be stated that the marketing evolution is not completed yet – if it ever will be. Together with consumer advertising that was established in the sales era, consumer scepticism on marketing and advertising has grown over the past few decades. Survey data from the 1930’s onwards confirm that consumers believe that advertising is often deceiving and that it persuades them to buy products and services they do not want. In fact, from childhood onwards people are taught to be sceptic on advertising and marketing (Boush, Friestad, & Rose, 1994).In answer to the scepticism of consumers and due to the wider availability of two-way communication through Web 2.0, a new marketing era, focused on relationships, dawned in the late 1990’s (Grönroos, 1996, 2004; Mckinney & Benson, 2013; Sheth & Parvatiyar, 1995; Stone, Woodcock, & Wilson, 1996). Although many definitions of Relationship Marketing (RM) exist, the discipline essentially strives to meet the needs and wants of all stakeholders, including both customers and the firm (Grönroos, 1990). On the contrary, some academics argue consumers do not want a relationship with firms. Rather, they want firms to help them make the decision that is good for them (Spenner & Freeman, 2012). In fact, applying RM involves the risk that the relationship trust is neglected (Nguyen, 2011). This is because consumers may believe that firms take advantage of their vulnerability and act opportunistically. Moreover, although Web 2.0 allows two-way communication between firms and consumers, it also allows two-way communication between consumers and other stakeholders. Accordingly, information for consumers became more widely available and was more convenient to access (Fournier & Avery, 2011).
The consequences of this increased availability and access to information, hence transparency, are significant. Where in the past firms themselves used to determine
Introduction 2 which information was shared and which not, now the public decides on this issue. According to Peppers and Rogers (2012) this ensures the “simple fact of transparency”, namely that people will find things out, whether firms decide to disclose information or not. Consequently, internal and external pressure on companies to become more transparent has increased (Huffington Post, 2013). The demand for corporate transparency has further increased after the financial crisis of 2008. Accounting scandals caused society to belief that firms, whether or not financial, merely strive to obtain money from consumers by misled (Bernardi & LaCross, 2005; Chiang & Chia, 2005; Kirby, 2012). Accordingly, consumer scepticism is not limited to marketing and advertising anymore. In fact, companies have become distrusted institutions in general (Kirby, 2012).
In order to overcome distrust and gain trust again, transparency is crucial. In fact, the more transparent, the more trusted (Jahansoozi, 2006; Kanagaretnam, Mestelman, Nainar, & Shehata, 2010; Kirby, 2012). Unfortunately, it is not that simple as transparency brings challenges as well. For example, when a firm discloses on its operations to consumers the information becomes available to competitors as well. This makes imitation easier and could mean loss of competitive advantage.
All in all, it seems like the next marketing era has dawned again. In this transparency era the task of marketing has become to communicate more information, resulting in more transparency from organisations towards their stakeholders (Christensen, 2002). Although attention for the issue of transparency, both in literature and practice, is increasing, the field is relatively young and unexplored. Only little is known about transparency within marketing and that what is known often comes from other fields of study. In fact, even a comprehensive definition of transparent marketing and knowledge on the effect of transparent marketing on consumer behaviour is lacking. Therefore, this study investigates the effect of different types of transparent marketing on product valuation. The aim of this research is to contribute to the literature on transparent marketing in order to establish the discipline further. More specifically, firstly, this research strives to establish both a comprehensive definition and concept of transparent marketing. Secondly, this study seeks to define, the most relevant, types of transparent marketing that have the highest willingness-to-pay. Thirdly, this
Introduction 3 research aims to contribute to the literature on transparency marketing on actual consumer behaviour specifically. Therefore, the research objective is to find out what type of transparent marketing adds most value to the product for consumers and thereby for the firm. A field experiment will be conducted to empirically test how different types of transparent marketing affect willingness-to-pay, as this is product valuation eventually.
To summarise, this study aims to find out on the effect of different types of transparent marketing on product valuation, hence willingness-to-pay. Therefore, the main research question is as follows:
• What is the effect of different types of transparent marketing on product valuation?
Theoretical framework 4
Theoretical framework
This chapter deliberates on the key constructs relevant for this study, including the following main topics: transparency, consumer behaviour and product valuation. Definitions and theories are provided and discussed. Finally, a conceptual model for transparent marketing and the hypotheses of the current research are presented. TransparencyTransparency is not only demanded within marketing. Rather, it is a characteristic that is relevant for companies as a whole, called corporate transparency. The Cambridge dictionary for business English (2014) defines this concept as “a situation in which business and financial activities are done in an open way without secrets, so that people can trust that they are fair and honest”. Other definitions of corporate transparency stress the disclosure of corporate information on governance, financing, and politics (R. Bushman & Smith, 2003; R. M. Bushman, Piotroski, & Smith, 2004; Habib, 2008; Leuz & Oberholzer-Gee, 2006; Miller, 2004).
As mentioned before, transparent marketing has not been comprehensively defined yet. However, there are certain disciplines within marketing, communication and psychology that relate to transparent marketing. These include: two-sided messaging, information disclosure and ethical marketing.
Two-sided messaging, information disclosure and ethical marketing
Within the context of transparent marketing, two-sided messaging and information disclosure are rather similar, namely: both imply that negative information is being shared in addition to a positive marketing message (Arpan & Roskos-Ewoldsen, 2005; Ein-Gar, Shiv, & Tormala, 2012; Johar & Simmons, 2000). This positive marketing message still stems from the sales era in which firms began to advertise products and services with positive messaging (Keith, 1960). Still today, marketing messages mainly contain positive information. Even when topics such as negative ecological and social impact are communicated in Corporate Social Responsibility (CSR), positive messaging predominates in practice. On the contrary, continuous consumer scepticism proves that consumers are aware of the negative aspects of products, services and organisations (Boush et al., 1994; Calfee & Ringold, 1994).
Theoretical framework 5 Therefore, scholars have paid attention to the effects of two-sided messaging and information disclosure. In general, the outcomes of these studies prove that two-sided messaging or disclosure of negative information can have a positive effect, under specifiable circumstances and mediated and moderated by multiple variables. Ein-Gar et al. (2012) argue that negative information can have a positive effect on the disposition of consumer towards a product, depending on processing effort and presentation order of the information. This counter intuitive effect of negative information is called the blemishing effect. In addition, Trifts and Häubl (2003) prove that price transparency can positively affect preference for an online retailer, depending on the price attractiveness and trustworthiness of the retailer. Related to this, multiple studies have found that two-sided messaging increases the credibility of the selling party (Bohner, Einwiller, Erb, & Siebler, 2003; Crowley & Hoyer, 1994). Thus, one could say disclosure of negative information leads to an openness that evokes trust and a sense of honesty at consumers.
The importance of trust is stressed by Audi (2008) who states that “without consumer trust, marketing fails” and that the discipline therefore has an ethical obligation towards consumers. This ethical obligation has been defined in ethical marketing. Klein, Laczniak, and Murphy (2006) state this discipline involves “practices that emphasize transparent, trustworthy, and responsible policies and actions that exhibit integrity and fairness to customers and other stakeholders”. In a critical note they pose the question what transparent marketing policies exactly are and assume that they exceed the minimum that is required by a code of conduct. However, ethics is concerned with the question of right and wrong (Oxford Dictionaries, 2014) and different theories exist to define this (Laczniak & Murphy, 1993). The issue with ethical marketing is that firms determine what is right and wrong, since their own code of conduct serves as a guideline, instead of letting consumers decide on this. Still, it can be stated the discipline strives to fairness at least.
In short, two-sided messaging, information disclosure and ethical marketing aim for trust, honesty and fairness. These three principles were used in the definition of corporate transparency in order to express the purpose of transparency as well. Therefore, the following paragraph elaborates on these three concepts.
Theoretical framework 6
Trust, honesty and fairness
Although trust, honesty and fairness are rather abstract concepts, they have been defined in many ways in the academic literature. Firstly, trust can be defined as the tendency to believe an informational source without independent evidence of accuracy (Audi, 2008) or as “a generalised expectancy held by an individual that the word of another ... can be relied on” (Berger & Milkman, 2009 as cited by Mckinney & Benson, 2013). Both these definitions share the two elements that are stressed in the following definition by Darsono (2009): “trust is the willingness to rely on another party in the face of risk”. It goes without saying, that in every exchange consumers face the risk of not receiving the value they were expecting (Blank, 2013). In order for a firm to gain trust it needs to be both able and willing to provide consumers with at least the value they are expecting. Interestingly enough, research found that trust could grow in repeated interactions, especially when these are not fully transparent. Namely, a decrease in reciprocity over the rounds of a repeated investment game was found, with or without transparency. Therefore, a not fully transparent situation, hence no insight in the decay of the reciprocity of the other, might limit this effect. Moreover, trust is positively related to firm growth and profit (Kanagaretnam et al., 2010). Secondly, honesty is when one is “free of deceit; truthful and sincere” (Oxford Dictionaries, 2014). Forehand and Grier (2003) propose a firm is considered honest when it acts according to its expressed motive. In other words, consumers accept firms to make profits, if those would express this as one of their motives. This is supported by Ellen, Webb, and Mohr (2006) who found that consumers responded positively to strategic efforts of firms, which are self-centred motives, and negatively to stakeholder-driven efforts that are other-centred motives. Thirdly, fairness is considered as one of the goals of ethics. Although the Oxford Dictionaries (2014) define fair as “treating people equally without favouritism or discrimination”, the context of this paper requires another definition. This is because that definition still allows bad treatment of consumers, as long as all are treated badly. Probably, in the current context the definition of fairness as the Golden Rule suits best. This ethical law holds that one should treat others, as one would like to be treated oneself (Laczniak & Murphy, 1993).
To conclude, trust, honesty and fairness can be defined as the essence of transparency, which relates the discipline to two-sided messaging, information disclosure and
Theoretical framework 7 ethical marketing. Trust is the voluntary reliance of one on another party in the face of risk, which requires ability and willingness from the other party. Honesty is when a firm acts upon its expressed motives and when a firm expresses the motives upon which it acts. This requires openness. Lastly, the treatment of others is considered fair when it is how one would treat oneself too, which requires moral consciousness. Since the essence of transparency in general has been defined, the next paragraph will elaborate on transparency within marketing more specifically.
Transparent marketing
Although one might intuitively understand the concept of transparent marketing, there clearly is a need for a definition in the literature. To begin with, marketing itself is often defined as “the activity, set of institutions, and processes for creating, communicating, delivering, and exchanging offerings that have value for customers, clients, partners, and society at large” (American Marketing Association, 2014). Hence, marketing is concerned with the creation of value for and the exchange of it with the stakeholders of a firm, by fulfilling needs and wants of different parties. Moreover, trust, honesty and fairness were defined as the core concepts of transparency, hence its essence. In addition to the essence of transparency, an act of transparency applies that includes the behaviour necessary in order to be transparent. This implies the actual disclosure of information, both positive and negative.
Taken into consideration both the definition of marketing and the essence and act of transparency, this paper defines transparent marketing as:
a situation in which a company is open and without secrets about how it creates and exchanges value by providing information to all of its stakeholders, so that those can trust that the company is honest and fair
Theoretical framework 8
Figure 1. Conceptualisation of transparent marketing
Requirements of transparent information
Vishwanath and Kaufmann (2001) have defined five features of transparent information, namely: access, comprehensiveness, relevance, quality and reliability. Based on these, five dimensions of transparent information can be defined. Firstly, accessibility is whether the information can be reached easily (Oxford Dictionaries, 2014). Secondly, comprehensibility is the completeness of the information; hence whether all elements are included (Oxford Dictionaries, 2014). Thirdly, relevance is the appropriateness of the formation for the matter at hand (Oxford Dictionaries, 2014). Moreover, quality tells about the general level of excellence of the information (Oxford Dictionaries, 2014). Lastly, reliability holds whether the information can be trusted (Oxford Dictionaries, 2014). So, when information is disclosed with the purpose of being transparent, it should at least meet these five requirements.
Types of transparency
As one might expect, multiple types of transparency exist since information can be disclosed on many topics. Hultman and Axelsson (2007) have defined four types of transparency that are relevant within marketing management in Business-to-Business (B2B) relationships. These are: cost/price, supply, organisational and technological transparency. In order to make the types fit a B2C context the following changes apply to the scope of the four different types. Firstly, cost/price transparency provides
Theoretical framework 9 insights in what part of the cost price goes to which party that is involved in the production of the product or service. Secondly, supply transparency discloses information about how, where and by whom the product or service is made or provided. Thirdly, organisational transparency includes the disclosure of information on the organisation that sells the product or service. This can involve for example structure or details on employees. Lastly, technological transparency is extended to attribute transparency, as this is more widely applicable to different B2C products and services. This type is concerned with the quality and features of the product or service itself (Oxford Dictionaries, 2014).
In addition, the same authors have defined three facets that do apply to transparency in general. Firstly, the degree of transparency holds the spectrum from complete transparency to no shared information at all. Secondly, the direction of transparency includes either one or two-way transparency and whether the demander or supplier is the stronger party. Lastly, the network of relationships between buyers and sellers determines the distribution of transparency because the network structure can cause transparency to be distributed directly or indirectly and horizontally or vertically (figure 2).
Figure 2. Typology of transparency for marketing management research
Source of transparency
In order to reach the situation of transparent marketing, disclosure of information is required. In other words, the act of transparency needs to be executed. This can be done by multiple parties, which could possibly make or break the information disclosure strategy of a firm.
Theoretical framework 10 To begin with, disclosure of information might be a mandate by governmental institutions (Loewenstein, Sunstein, & Golman, 2013) However, information could be disclosed voluntary by firms themselves as well (Stone et al., 1996). According to Spenner and Freeman (2012) this is all consumers want; they want firms to provide them with simple information that is easy to navigate through, trust and outweigh. Moreover, Bar‐Isaac, Caruana, and Cuñat (2010) state that marketing is able to make it easier, or harder, for consumers to find more detailed information about products and services. The voluntary publication of information by firms, both positive and negative, would allow consumers to gather and evaluate information themselves in order to make a decision on their own. Hence, the firm facilitates free choice by consumers. In the literature, actual effectiveness of self-disclosure is discussed as well. For example, Trifts and Häubl (2003) found that when an online retailer provides access to uncensored competitor price information, this can lead to a greater long-term preference for the retailer. Moreover, Arpan and Roskos-Ewoldsen (2005) argue that use of the self-disclosure strategy “stealing thunder” during a crisis results in higher credibility ratings, less severe perceptions of the crisis at hand and a higher purchase intent. Two main reasons for this exist: firstly, disclosure of negative information about oneself enhances the credibility of that party. Secondly, people might change their meaning about the negative information disclosed, for example because they think the crisis could not be that severe when a firm discloses the information itself.
However, self-disclosure is rare because of three main reasons. At first, the costs associated minimise self-disclosure of information. Secondly, the risk of losing the advantage of possessing more information that benefits innovation often prevents firms from disclosure. Lastly, external influences might limit the disclosure of information by firms themselves (Vishwanath & Kaufmann, 2001). One of the most important external influences on firms is the scepticism on marketing and advertising that dominates in the minds of consumers nowadays. This sentiment could cause consumers to distrust trustworthy, honest and fair information. Even when firms disclose with the best intentions, consumers might still be sceptic since firms have become distrusted institutions in general (Bernardi & LaCross, 2005; Boush et al., 1994; Chiang & Chia, 2005). Therefore, in contrast to self-disclosure, other
Theoretical framework 11 stakeholders, like consumers, employees, NGO’s and press, could disclose information of the firm as well.
Consumer behaviour
Within the scientific discipline of marketing an extensive amount of research has been conducted on consumer behaviour. Much of this research is focused on attitudes, preferences and intentions, which are assumed to precede behaviour in both the theory of planned behaviour and theory of reasoned action (Ajzen, 1985; Ajzen & Fishbein, 1980). For example, Burch et. al. (2012) found that there is a need for transparency among consumers. Namely, 68% out of 1,000 UK consumers agree that transparency has become more important for them when they want to buy something. In fact, after price and quality, it is the most important factor for them in the decision whether they want to buy something or not. This can be labelled as an attitude since these consumers have merely indicated that they value transparency in a survey. Whether they also act upon this attitude, or even preference or intention, is yet unknown. This discrepancy between how consumers intend to behave and how they actually behave has been researched in many disciplines. In the discipline of CSR it is called “the green gap” (Fowler III & Close, 2012). More specifically, this implies the gap between what consumers believe and what consumers actually do when it comes to protection and improvement of the environment. Unfortunately, this topic has remained unexplored in any transparency related discipline. Yet, it can be assumed that such a discrepancy also exits for transparency related believes and behaviours. However, since transparency is all about disclosure of information by another party, it is not a matter of what the consumer actually does by itself. Rather, a similar gap for transparency would be concerned with consumers accepting or avoiding the information that is disclosed by another party. Therefore, information avoidance is discussed next.
Information avoidance
Sweeny, Melnyk, Miller, and Shepperd (2010) define information avoidance as “any behavior intended to prevent or delay the acquisition of available but potentially un- wanted information”. Note that it concerns information that is available as is the case with transparent information. Although avoidance of negative information seems most logic, also positive information can be avoided, for example to not spoil a
Theoretical framework 12 surprise. Information avoidance can be either active, when one does ask another not to disclose information, or passive, when one does fail to ask for information. In addition to this definition, these same authors also established a framework for understanding information avoidance decisions according to the who, individual factors, when, situational factors, and why, motivations, of information avoidance (figure 3). Especially interesting for transparent marketing are the three main reasons why, in this case, consumers would avoid information. This could be because either the information demands a change in beliefs, demands an undesired action or the information evokes negative emotions. Such an undesired action could for example be consumers who, based on their beliefs, are restricted to obtain a product. Still, they want to and therefore avoid the information that is disclosed. Moreover, four main factors indicate when information is avoided, ergo the when of information avoidance. Firstly, when the perceived control over the outcome of the information is low, one is more likely to avoid the information. This applies to transparent information per definition since someone else than the self discloses it. Secondly, the fewer coping resources are available that can help one handle the information the more one will avoid the information. Thirdly, the expectation whether the information will be positive or negative affects information avoidance, namely when one expects negative information avoidance is more likely. Lastly, ease of obtaining and interpreting information does not apply in the case of transparent information. This is because it was determined, based on Vishwanath and Kaufmann (2001), that transparent information is accessible, comprehensive, relevant, qualitative and reliable by nature. Finally, who is more likely to avoid the information is mainly assigned to personal characteristics and circumstances.
In short, often a gap between believes and actual behaviour exists. Although consumers have indicated the need for transparency (Burch et.al., 2012), they might not per se behave according to it. Similarly, the literature stresses that although attitudes, preferences and intentions precede behaviour, those are not behaviour yet. Still, the actual performed behaviour is what counts for the firm. For example, one can have a positive and strong attitude towards a product, and even might have the intention to buy it. However, if one does not actually purchase the product, the effect
Theoretical framework 13 for the firm in terms of sales is still zero. Therefore, this research focuses on actual behaviour, which is discussed more in depth next.
Figure 3. Framework for understanding information avoidance decisions
Actual behaviour
The focus on actual behaviour while not taking into account attitudes, preferences or intentions of consumers into consideration is in line with the Behaviour Modification Perspective (BMP) by Skinner (as cited by Nord & Peter, 1980). This approach to behaviour modification focuses on “environmental factors that influence behaviour and avoids processes that are assumed to occur within the individual.” In other words, mediators such as attitudes, preferences and intentions are ignored in order to study the direct relationship between the environment and its effect on behaviour.
Typically, two types of environmental manipulations do exist, namely: classical or respondent conditioning, and instrumental or operant conditioning. The former, implies the process through which “a previously neutral stimulus, by being paired with an unconditioned stimulus, comes to elicit a response very similar to the response originally elicited by the unconditioned stimulus” (Nord & Peter, 1980). Pavlov, who found this effect, paired the sound of a bell with the food of a dog being served. After repeating it so many times, the dog started to slobber only hearing the bell, even when the food was not even served yet. So, the response to the food, the slobbering, was transferred to the sound of the bell. The latter, instrumental conditioning, assumes that behaviour is controlled by the individual, but that it can be conditioned by the consequences that occur after the behaviour. Hence, instrumental conditioning has occurred when the probability that an individual will behave a
Theoretical framework 14 certain way is altered by the events that follow certain behaviour. A well-known principle of instrumental conditioning are discriminating stimuli, which change the probability of a certain behaviour by their presence, which is the case with signs that states “50% off” for example (Nord & Peter, 1980).
To conclude, actual behaviour is only one of many parts that make up consumer behaviour. Still, this study focuses on actual behaviour specifically as this actually affects results in terms of sales for firms. This approach is in line with the BMP that focuses on the effect of environmental factors on, in this case, consumer behaviour without mediation of attitudes, preferences, and intentions.
Besides an actual purchase another example of actual consumer behaviour is an actual bid on a product or service. In that case, consumers valuate the product or service themselves instead of an agreement with a previously determined price. Since this research strives to find out whether transparency affects the value of a product in the eyes of the consumer, product valuation will be discussed next.
Product valuation
According to the Oxford Dictionaries (2014) value is “the regard that something is held to deserve; the importance, worth, or usefulness of something”. Therefore, one could say that, in the context of consumer products, value is what a good is worth to a consumer. The maximum amount an individual is ready to spend to a product or service is called willingness-to-pay (Chan, Kadiyali, & Park, 2007; Koschate-Fischer, Diamantopoulos, & Oldenkotte, 2012). Hence, willingness-to-pay is the amount of money that a consumer is willing to spend to a product. However, when the price would exceed this amount the consumer would rather not obtain the product at all than for a larger amount of money. Therefore, this study uses willingness-to-pay as a measure for product valuation.
To summarise, the need for transparency by firms among consumers exists, both on a corporate and marketing level. Two-sided messaging, information disclosure and ethical marketing are disciplines that overlap in some way with transparent marketing. This is mainly on the topic of the essence of transparent marketing, which is the aim for trust, honesty and fairness. In addition to the essence of transparent marketing, an
Theoretical framework 15 act of transparent marketing has been defined, which implies the disclosure of information, both positive and negative. In order to be transparent, information needs to be accessible, comprehensive, relevant, qualitative and reliable. Moreover, four types of transparent marketing for a B2C context have been defined, including cost/price, organisational, supply and attribute transparency. Also, different sources of transparent marketing have been specified. In general, the distinction between two sources can be made, namely: the firm and other parties. The main reason for a firm to not disclose information itself is the growing scepticism of consumers towards firms in general. In addition, consumer behaviour has been discussed. It is likely that a discrepancy between what consumers believe and what consumers do exists for transparency issues. Consequently, information avoidance can occur in which consumers avoid information that is disclosed because it requires a change in believes, an undesired action or evokes unpleasant emotions. Due to the gap between believes and behaviour, actual behaviour modification has been discussed more into depth. More specifically, product valuation as actual behaviour was examined in which was found that willingness-to-pay could serve as a measure.
This study
This study strives to find out if and how transparent marketing affects product valuation in order to determine which type of transparent marketing adds most value to the product for consumers and thereby for the firm. A field experiment tests which type of transparency and what source of transparency creates most value for the consumer and thereby for the firm. Since this experiment takes place in a B2C context the four adjusted types of transparency based on those of Hultman and Axelsson (2007) will be used. These include: cost/price transparency, organisational transparency, supply transparency and attribute transparency. In addition, the disclosure source will be manipulated because this research strives to find out whether transparency by the firm increases product valuation. When the disclosure source would not be manipulated, the experiment would merely examine whether the availability of more information increases willingness-to-pay. An advantage of this manipulation is that, additionally, the effect of different disclosure sources can be tested. Therefore two conditions apply to the source of transparency, namely: disclosure by the firm and disclosure by a consumer. Moreover, the degree, direction and distribution of transparency are not considered in the current research as the
Theoretical framework 16 online auction tool with which data will be collected predetermines these. The degree of transparency will be kept equal in all conditions, direction is uni-directional and downstream, while distribution is direct and vertical. Finally, this study will consider actual consumer behaviour in the form of actual auction bids, in which the willingness-to-pay of consumers is used as a measure for product valuation.
Based on the literature discussed before two hypotheses have been established. Firstly, since consumers have indicated the need for transparency it is assumed that:
• Transparency on how value is created and exchanged will lead to a higher willingness-to-pay.
Secondly, as literature has demonstrated the increasing scepticism of consumers towards firms it is expected that:
• When information is disclosed by a consumer this will lead to a higher willingness-to-pay since the level of scepticism towards the firm is minimised. The next chapter discusses the method that is used for the experiment with which the answer to the main research question is to be found empirically.
Method 17
Method
This chapter discusses the method used for the experiment that is conducted to investigate the effect of different types of marketing transparency on product valuation. Firstly, the experimental design of this study is presented. Secondly, the pre-test and the final manipulations that will be used in the experiment are discussed. Thirdly, the operationalisation, or measurement, of the independent and dependent variables is presented. Lastly, a description of the procedure of the experiment is provided, including the research tool Veylinx, the product that is auctioned, the stimuli that will be used and the experiment itself.Experiment design
The present research is conducted according to a 3x2 factorial design between subjects. Firstly, the source of transparency is expressed in three conditions, namely: (1) no disclosure, (2) disclosure by the firm, and (3) disclosure by the consumer. Secondly, the two most relevant types of transparency will be selected out of the following four: (1) cost/price, (2) supply, (3) organisational and (4) attribute transparency.
Pre-test
In order to select the two most relevant types of transparent marketing and to make sure all statements disclose information that is equally positive or negative and transparent, a pre-test was conducted. This pre-test included three statements for each cost/price, supply and organisational transparency. Since the pre-test was combined with the pre-test of another study, 22 statements on attribute transparency were tested. This makes 31 statements in total.
Twenty-eight respondents were asked to indicate how positive or negative the disclosed information appeared to them according to a seven-point Likert scale. The scale ranged from 1) very negative to 7) very positive, with 4) indicated as neutral. In addition, all statements were tested on perceived degree of transparency. Respondents were asked to indicate whether they disagreed or agreed with the statements being relevant, comprehensive, qualitative and reliable (Vishwanath & Kaufmann, 2001) on a Likert scale, again ranging from 1 to 7. The mean scores of these four criteria were
Method 18 used as a score for perceived degree of transparency. Accessibility was not used as an indicator because the information was presented to the respondents directly and they did not need to search for it themselves (Appendix I. Pre-test).
In order to determine the two most relevant types of transparent marketing for this study, a one-way ANOVA was conducted, which appeared to be significant, F (3, 108) = 12,841, p = .000 (table 1). This means that significant differences in terms of relevance for this study do exist between the four types of transparency. In order to determine the exact significant differences between the four types, additional contrast tests were conducted (table 2). With these contrast tests each of the types of transparency was compared to the other three types. The order was as follows: (1) cost/price, (2) organisational, (3) supply and (4) attribute transparency. The four contrast tests point out that both organisational and attribute transparency differ significantly from the, for each, other three types of transparency. However, both types do not differ the same way. In fact, attribute transparency appears to be significantly more relevant for this study, where organisational transparency appears to be significantly less relevant for this study. This is reflected in the relevance means of each type of transparency (table 3). Consequently, the two most relevant types of transparency, attribute (M = 5.008) and cost/price (M = 4.023) were selected to be tested during this research.
Table 1. ANOVA of relevance means per type of transparency
Sum of Squares df Mean Square F Sig. Between Groups 48.266 3 16.089 12.841 .000 Within Groups 135.315 108 1.253 Total 183.582 111
Method 19
Table 2. Contrast tests of relevance means per type of transparency
Contrast Value of Contrast Std. Error t df Sig. (2-tailed) Relevance means Assume equal variances 1 .079 .732 .109 108 .914 2 3.555 .732 4.852 108 .000 3 .222 .732 .304 108 .762 4 -3.857 .732 -5.264 108 .000 Does not assume equal variances 1 .079 .759 .105 43.274 .917 2 3.555 .876 4.057 35.901 .000 3 .222 .725 .307 46.576 .760 4 -3.857 .527 -7.319 94.179 .000
Table 3. Relevance means per type of transparency
N Mean Std. Deviation Std. Error
Cost/price transparency 28 4.023 1.179 .222
Organisational transparency 28 3.154 1.435 .271
Supply transparency 28 3.988 1.101 .208
Attribute transparency 28 5.008 .588 .111
Total 112 4.043 1.286 .121
Next, one statement on attribute transparency and one statement on cost/price transparency needed to be selected from the remaining 25 statements. Of course, both statements were required to be equally positive or negative and equally transparent in order to avoid these factors influencing the results. Therefore, two additional one-way ANOVAs were conducted. Firstly, it was determined which statements were equally positive or negative, F (30, 837) = 3.681, p = .000 (figure 4, marked with an *). Secondly, out of the seven statements that were left, the two statements that were indicated as equally transparent were selected, F (30, 837) = 12.377, p = .000 (figure 5, marked with an *).
Method 20
Figure 4. Means plot positive/negative score per statement
These three criteria together, most relevant, equally positive or negative and equally transparent, determined that the following two statements would be used during the experiment:
• Statement 1: “20% of the price of Hero Kokoswater goes to the farmers that supply the coconut water for this drink” (20% van de prijs van Hero
Kokoswater komt terecht bij de boeren die het kokoswater voor deze drank leveren).
Method 21
Figure 5. Means plot transparency score per statement
In order to verify that no significant differences on the three criteria between these two statements existed, two more one-way ANOVAs with contrast tests were conducted. One to compare the positive/negative score of the two statements, F( 30, 837) = 12.370, p = .000, and one to compare the transparency score of the two statements, F (30, 837) = 3.681, p = .000. These tests proved that there was no significant difference (p = .754) on how positive or negative the content of the statements appeared (table 4). Also, no significant difference (p = .774) exists in the extent of transparency between these two statements (table 5). In other words, the two statements seemed equal to each other on the criteria that were set, which made them suitable to be used as stimuli in the experiment.
Method 22
Table 4. Contrast test statement 1 and 21 on positive/negative score
Contrast Value of Contrast Std. Error t df Sig. (2-tailed) Positive/ negative score Assume equal variances 1 -.1071 .341 -.314 837 .754
Does not assume equal variances
1 -.1071 .388 -.276 47.422 .784
Table 5. Contrast test statement 1 and 21 on transparency score
Contrast Value of Contrast Std. Error t df Sig. (2-tailed) Transpar ency score Assume equal variances 1 -.0714 .248 -.288 837 .774 Does not assume equal variances 1 -.0714 .277 -.258 50.673 .798
Finally, the pre-test led to the following five manipulations: no transparency, cost/price transparency by the firm, cost/price transparency by the consumer, attribute transparency by the firm and attribute transparency by the consumer (table 6).
Table 6. Schematic representation of five conditions
Operationalisation
Although the essence of transparent marketing is rather abstract, the independent variables, hence the types of transparency and the sources of transparency, require a manipulation and measurement. The former are operationalised through statements that each disclose information that is equally transparent. The topic depends on the type of transparency. For the latter, the distinction is made between disclosure by the firm and disclosure by a consumer. Disclosure by the firm is expressed as the
Source of transparency Type of transparency No disclosure Disclosure by firm Disclosure by consumer Cost/price NT CTF CTC
Method 23 information being disclosed by an expert of the firm, while a Facebook post from a consumer displays disclosure by a consumer. It goes without saying, that the control condition without transparency does not disclose any information at all.
The dependent variable for this experiment is product valuation. This will be measured with willingness-to-pay. This is the maximum amount an individual is ready to spend to a product or service (Chan et al., 2007; Koschate-Fischer et al., 2012). This variable will be measured in euro cents.
In addition to the measurement of willingness-to-pay, the research tool used allows to ask five complementary survey questions. Four out of these five questions will be used in order to check the reliability of the conceptualisation of transparency as it was established in the theoretical framework of this study. One question is dedicated to each of the items on the essence of transparency, thus trust, honesty and fairness. The last question concerns the act of transparency. Moreover, it is important to know whether respondents are familiar with the firm Hero and its products or not. Therefore, also one question will be used to obtain this information. Therefore, on the conceptualisation of marketing transparency presented in the theoretical framework the following five survey questions were established:
• I am familiar with Hero (Ik ben bekend met Hero).
• I think it is important to have confidence in firms before I decide to buy something at them (Ik vind het belangrijk om vertrouwen te hebben in
bedrijven voordat ik besluit iets bij hen te kopen).
• I think honesty and openness by firms is important (Ik vind eerlijkheid en
openheid door bedrijven belangrijk).
• I think equal treatment of consumers by firms is important (Ik vind gelijke
behandeling van consumenten door bedrijven belangrijk).
• I think it is important that firms disclose as much information as possible about themselves (Ik vind het belangrijk dat bedrijven zoveel mogelijk
informatie beschikbaar stellen over zichzelf).
The first question is answered with yes or no. The other four questions are answered with a 5-point Likert scale with the following labels: totally disagree, disagree, neutral, agree, and totally agree.
Method 24 Procedure
This last section discusses how the actual experiment was prepared and executed. To begin with, the research tool with which the experiment is conducted is discussed. Moreover, this section deliberates on the product that is auctioned. Next, the stimuli that will be used in order manipulate the conditions are presented. Lastly, the actual experiment is discussed briefly.
Veylinx
This study strives to determine which out of four types of transparent marketing, adds most value to a product for the consumer and thereby for the firm. In order to answer this research question empirically, an experiment will be conducted because it allows the use of a controlled setting. More specifically, a field experiment will be conducted, which means the experiment is conducted in a real-life setting (Bryman & Bell, 2007). In this case, the experiment is conducted within the online auction Veylinx. This is a research tool of the University of Amsterdam with which real people, place real bids on real products. People indicate their maximum willingness-to-pay for the product that is auctioned and the person with the highest bid gets the product for the second highest price (Veylinx, 2014). This type of auctions is called second price sealed bid, or Vickrey, auctions. Research has shown that second price sealed bid auctions are preferred as a method in order to obtain willingness-to-pay information (Noussair, Robin, & Ruffieux, 2004). The disadvantage of first bid auctions, auctions in which the highest bidder gets the product, is that participators to these auctions tend to estimate what the product is worth to others. Consequently, they make a bid slightly above this amount so that they can obtain the product for a bid as small as possible but still large enough to win the product. In second price auctions there is no need to pose a higher bid since one might be forced to pay a higher amount than what one thinks the product is worth. Also, there is no need to pose a lower bid since one might lose the product for no good reason. So in the end, participators to second bid auctions indicate their actual willingness-to-pay, which allows to determine how consumers valuate products. In the end, the person who values the product most wins (Amit, 2013).
Method 25
Product
The product that is auctioned via Veylinx is Hero Coconut water Neutral (Hero
Kokoswater Naturel). This beverage is produced by Hero. Since 1914, this Dutch firm
produces foods and drinks based on fruit. The Hero Coconut water Neutral includes 98% coconut water and 2% fruit sugars. According to Hero this last ingredient was consciously added for the following reason: “we do this because coconut water is that pure you would taste every difference between different harvests directly” (Dit doen
we omdat kokoswater zó puur is dat je ieder smaakverschil uit de diverse oogsten direct zou proeven) (Hero, 2014). Accordingly, this product was selected because it
discloses attribute information that might usually not be shared with consumers. Also, producer Hero is willing to share information concerning cost/price, its supply chain and the organisation itself. Lastly, coconut water is a relatively unknown product, which causes consumers not to know the average price of the product already. This secures the external validity of this study.
Stimuli
Since five conditions will be tested, five different stimuli have been developed. In all stimuli the product, Hero coconut water Neutral, is displayed as the main object. In addition, a title states: 2-pack Hero Coconut water (2-pack Hero Kokoswater). A subtitle disclosed the quantity of the product: 2x330ml. The conditions in which the information is disclosed by the firm make use of the character of fruit specialist Rianne that is used on the website of Hero already (Hero, 2014). A quotation of Rianne plus her picture discloses the statement on one of the transparency types in its exact form, left from the bottle of Hero coconut water. Disclosure of information by the consumer was replicated by developing a Facebook post of a consumer. The statements were altered to make them better suit the purpose of being disclosed by a consumer (Appendix II. Stimuli).
Experiment
The actual experiment of this study was conducted with Veylinx on the 28th of April, 2014. In the morning 1081 people who subscribed at Veylinx received an email with an invitation to the auction of that day. In total, 424 respondents made a bid and 391 completed the survey. This includes the bids of zero euro cents, or in other words, the people who did not want to obtain the product at any price.
Results 26
Results
This chapter presents the results of the field experiment that was conducted for this study. This includes demographics, tests of normality and Tobit regressions. All tests are conducted with either SPSS or Stata.Demographics
In total 424 respondents, of which 234 men and 190 women, made a bid (table 7). The ages of the respondents ranged from 16 to 103 years old, with an average age of 40 years old (table 8). Moreover, majority of the sample appeared to be higher educated, namely 198 respondents (N=301) (table 9). Also, 94 respondents seemed still to be studying as they indicated to be students (N=292) (table 10). Lastly, only 13 respondents appeared not to be familiar with the brand Hero in contrast to 374 of them who indicated they were (N=387) (table 11).
Table 7. Frequencies of gender
Frequency Percent Valid Percent Cumulative Percent Valid Men 234 55.2 55.2 55.2 Women 190 44.8 44.8 100.0 Total 424 100.0 100.0
Table 8. Descriptives of age
N Minimum Maximum Mean Std. Deviation
Age 424 16.00 103.00 40.245 15.899
Valid N (listwise) 424
Table 9. Frequencies of higher education
Frequency Percent Valid Percent Cumulative Percent Valid No 103 24.3 34.2 34.2 Higher education 198 46.7 65.8 100.0 Total 301 71.0 100.0 Missing System 123 29.0 Total 424 100.0
Results 27
Table 10. Frequencies of students
Frequency Percent Valid Percent Cumulative Percent Valid No 198 46.7 67.8 67.8 Student 94 22.2 32.2 100.0 Total 292 68.9 100.0 Missing System 132 31.1 Total 424 100.0
Table 11. Frequencies of familiarity with Hero
Frequency Percent Valid Percent Cumulative Percent Valid Not 13 3.1 3.4 3.4 Familiar 374 88.2 96.6 100.0 Total 387 91.3 100.0 Missing System 37 8.7 Total 424 100.0 Tests of normality
To begin with, Shapiro Wilk tests were conducted in order to find out whether the obtained data were normally distributed. Both the originals bids in euro cents (table 12) and the bids to which an algorithm was applied (table 13) appeared to be significant regarding normality, F = 250, p = .000. Hence, the data were not distributed normally. Rather, they were skewed to the left. This was expected, mainly because of the large amount of bids of zero euro cents. Due to the purpose of Veylinx being a research tool, the online auction encourages its respondents to always make a bid. In case respondents do not want the product at any price, they can enter €0,00 as their bid. This in contrast to other auctions where bidding normally starts from any monetary value except zero. Veylinx strives to obtain a bid from every respondent invited because each bid, also bids of zero euro cents, are considered to be research data. Another explanation for the non-normal distribution of the data is that people normally tend to make bids of round figures, such as €0,50, €1,00 and €1,50 etcetera. This tendency causes peeks around certain values that can explain not perfectly normally distributed data as well.
Results 28
Table 12. Tests of normality original bids
Kolmogorov-Smirnova Shapiro-Wilk
Statistic df Sig. Statistic df Sig.
Bid amount .279 205 .000 .319 205 .000
a. Lilliefors Significance Correction
Table 13. Tests of normality bids with algorithm applied
Kolmogorov-Smirnova Shapiro-Wilk
Statistic df Sig. Statistic df Sig. Bid amount
algorithm
.145 205 .000 .909 205 .000 a. Lilliefors Significance Correction
Tobit regressions
In order to conduct parametric tests, data are assumed to have a normal distribution. This is the case when the probability distribution of a variable is perfectly symmetrical and it not too flat or peaked. In other words, the probability distribution of the variable is supposed to have a skew and kurtosis of zero (Field, 2009).
On the contrary, the data of this study were not distributed normally. In fact, the skew and kurtosis values of the bid amount variable to which the algorithm already was applied still were -.235 for skewness and -1.758 for kurtosis (table 14). Hence, the non-normal distribution. Still, the data were not just skewed or leptokurtic. Rather, the data of this study are censored. This means that data cannot go below or above a certain value at one side of the distribution (Institute for digital research and education, 2014). In this case, the data are left-censored because the value of the bid amount cannot go lower than zero euro cents (figure 6).
Table 14. Descriptives bid amount with algorithm applied
Statistic Std. Error Bid amount algorithm Skewness -.235 .119
Results 29 Since the assumption of normality does not apply to the dataset of this study, parametric tests could not be used. Though, Stata, another type of data analyses software, can conduct the test that is required when data are censored, namely a Tobit regression. The Tobit model, also called a censored regression model, is “designed to estimate linear relationships between variables when there is either left- or right-censoring in the dependent variable (also known as right-censoring from below and above, respectively)” (Institute for digital research and education, 2014). Since the distribution probability of the bid amounts is left-censored a Tobit regressions does apply here.
Figure 6. Histogram of bid amounts
To begin with, the relationship between the bid amount in euro cents and the source of transparency and the type of transparency was tested (table 15). For this test N=424 of which 164 cases were left-censored, 260 cases were uncensored and zero cases were right-censored. The probability of Chi2 (p = .356) already indicates that there is no fit between the proposed model on source and type of transparency compared to an
Results 30 empty model. This is confirmed by the non-significant results. For this reason, this test indicates that there are no significant grand effects caused by solely the source or type of transparency.
Table 15. Tobit regression bid amount and type and source of transparency
LR chi2 (3) = 3.23
Prob > chi2 = 0.356 Pseudo R2 = 0.001 Bid amount Coef. Std. Err. t P>|t| [95% Conf. Interval] Source Consumer -37.415 34.945 -1.07 0.285 -106.104 31.273 Firm -74.171 46.666 -1.59 0.113 -165.89 17.555 Type Cost/price -44.299 34.983 -1.27 0.206 -113.062 24.464 Attribute 0 (omitted) _cons 69.941 31.835 2.20 0.029 7.365 132.518 /sigma 297.941 13.831 270.694 325.067
Next, the relationship between the bid amounts in euro cents and the five true treatments was analysed (table 16). Similar to the test above N=424, with 164 left-censored cases and 260 unleft-censored cases. Again no significant effects were found which was already expected because of the probability of Chi2 (p = .467) that was not significant as well. These results indicate that there is no significant effect of one of the five treatments on bid amount.
Thus, so far, no significant effects of either the source or type of transparency or the true treatments were found. Therefore, the previous tests were repeated, however, this time, age, gender and being a student or not were taken into account.
Results 31
Table 16. Tobit regression bid amounts and true treatments
LR chi2 (3) = 3.57
Prob > chi2 = 0.467 Pseudo R2 = 0.001 Bid amount Coef. Std. Err. t P>|t| [95% Conf. Interval] True treatments CTC 85.473 50.545 1.69 0.092 -13.881 184.827 ATC 20.989 47.614 0.44 0.660 -72.601 114.581 ATF 2.771 49.346 0.06 0.955 -94.226 99.768 CTF 26.983 48.734 0.55 0.580 -68.810 122.778 _cons -4.145 34.752 -0.12 0.905 -72.455 64.1654 /sigma 297.625 13.820 270.459 324.791
The first test analysed the relationship between bid amount and the source and type of transparency compared to the control condition, taken age, gender and being a student or not into consideration (table 17). This resulted in a test in which N=292 with 101 left-censored cases and 191 uncensored cases. Although the probability of Chi2 (p = .119) indicates that this model does not fit better than an empty model, some significant results were found. Namely, the firm as source of transparency appears to be significant opposed to no transparency, p < 0.05. The coefficient indicates that when the firm itself discloses information, there is a -131.603 point increase in the predicted value of bid amount compared to no disclosure at all. In other words, disclose of information by the firm decreases product valuation opposed to no transparency. In addition to the comparison of the sources of transparency to the control condition, a contrast test was conducted in order to test whether the two sources differed significantly from each other. However, this was not the case (p = .2111). In other words, no significant difference exists between disclosure by the firm or consumer. Moreover it was found that being a student increases the predicted value of bid amount with 112.718 points, p < 0.05. This means that students are more likely to valuate the product higher than non-students.
Results 32
Table 17. Tobit regression bid amount and type and source of transparency with age, student and gender taken into account
LR chi2 (3) = 10.13
Prob > chi2 = 0.119 Pseudo R2 = 0.004 Bid amount Coef. Std. Err. t P>|t| [95% Conf. Interval] Source Consumer -57.366 45.008 -1.27 0.203 -145.956 31.223 Firm -131.603 59.649 -2.21 0.028 -249.009 -14.196 Type Cost/price -55.233 44.736 -1.23 0.218 -143.2885 32.822 Attribute 0 (omitted) Birth year -3.052 1.646 -1.85 0.065 -6.293 .187 Student 112.718 56.622 1.99 0.047 1.269 224.167 Gender 24.980 40.306 0.62 0.536 -54.354 104.314 _cons 6087.764 3235.92 1.88 0.061 -281.475 12457 /sigma 318.850 17.092 285.206 352.493
The second Tobit regression tested the relationship between bid amount and the true treatments compared to the control condition, with age, gender and being a student or not taken into account (table 18). Also for this test N=292 of which 101 cases were left-censored and 191 cases were uncensored. Again the probability of Chi2 indicates that no significant results will be found (p = .159). On the contrary, the condition in which cost/price transparency was disclosed by the consumer appeared to be significant, p < 0.05. Hence, consumers tend to valuate the product higher when information on cost/price is provided by a consumer than when no information is disclosed at all. No such effect was found for any of the other conditions. Lastly, some contrast tests between all true treatments were conducted in order to verify whether these significantly differed from each other. This did not deliver any significant results. In other words, none of the five treatments significantly increases product valuation more than any other of the treatments.
Results 33
Table 18. Tobit regression bid amount and true treatments with age, student and gender taken into account
LR chi2 (3) = 10.56
Prob > chi2 = 0.159 Pseudo R2 = 0.004 Bid amount Coef. Std. Err. t P>|t| [95% Conf. Interval] True treatments CTC 147.485 64.311 2.29 0.023 20.899 274.072 ATC 63.112 61.240 1.03 0.304 -57.428 183.653 ATF 33.083 62.819 0.53 0.599 -90.565 156.731 CTF 58.901 63.656 0.93 0.356 -66.395 184.199 Birth year -3.082 1.643 -1.88 0.062 -6.316 .152 Student 109.784 56.678 194 0.054 -1.778 22.348 Gender 26.011 40.258 0.65 0.519 -53.229 105.252 _cons 6015.224 3230.646 1.86 0.064 -343.729 12374.18 /sigma 318.267 17.069 284.668 351.866 Transparency scale
The four out of the five survey questions that were posed to the respondents after they made their bid, were aimed at finding out how important these respondents think transparency is. The questions asked about the three pillars of the essence of transparency, trust, honesty and fairness, and about the act of transparency, providing information.
Firstly, a one-way ANOVA was conducted in order to find out whether the scale items significantly differed from each other. Fortunately, this was not the case, which means that not one item affects the outcome of the scale significantly, F (237856.638, 1818744659.460) = .317, p = .994 (table 19).
Table 19. One-way ANOVA transparency scale
Sum of Squares df Mean Square F Sig. Between
Groups
237856.638 15 15857.109 .317 .994 Within Groups 18744659.460 375 49985.759
