Price transparency: the limit of negative information in advertisements

Master Thesis

‘Price transparency: the limit of negative information in

advertisements’

Author:

Zeno van Heerwaarden (10203028)

Under supervision of:

Joris Demmers

MSc. in Business Administration - Marketing Track Amsterdam Business School

Statement of originality

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

Table of contents

1

Introduction ... 4

1.1 The Age of Transparency ... 4

1.2 Price transparency ... 6 1.3 Two-sided advertising ... 7 1.4 Research contribution ... 7

2

Theoretical framework ... 9

2.1 Transparency ... 9 2.2 How to be transparent? ... 11 2.3 Price transparency ... 12 2.4 Effects of transparency ... 13 2.4.1 Willingness-to-pay ...14 2.4.2 Trustworthiness ...15

2.4.3 Negative attitude towards the advertisement ...17

2.4.4 Consumers’ involvement ...18

3

Study 1 ... 20

4

Study 2 ... 21

4.1 Method ... 21 4.2 Design ... 22 4.3 Procedure ... 23 4.4 Variables ... 23 4.5 Sample ... 24 4.6 Testing hypotheses ... 25

5

Results ... 26

5.1 Multiple mediation analysis ... 26

5.2 Moderation analysis ... 26

6

Discussion and conclusion ... 28

7

Future research ... 29

8

References ... 31

1 Introduction

1.1 The Age of Transparency

We live in the Age of Transparency (Fournier & Avery, 2011). The rise of the internet drastically changed the availability and convenient access to information (Fournier & Avery, 2011; Granados, Gupta & Kauffman, 2010; Grewal, Krishnan & Sharma, 2003; Miao & Mattila, 2007; Sinha, 2000). The amount of information from now until 2020 will double every two years (Gantz & Reinsel, 2012). This is also due to the fact that access of information has increased even more with internet technology changing from Web 1.0 to Web 2.0. This is because Web 1.0 was mainly a one-way communication tool where people remained passive receivers of information. Web 2.0, however, enabled the creation of online microcontent and social media (Alexander, 2006) and thereby allowed people to be more actively involved in creating online content. Together with an increasing flood of data, consumers are able to share their thoughts, opinions, and experiences about companies, brands, and products (Alexander, 2006; Austin & Upton, 2016; Fournier & Avery, 2011; Kirby, 2012). Most importantly about Web 2.0 is that people are able to ‘bypass’ companies. This means that companies are not able to control information or where it is going. Years ago, marketers were the ones that determined what information was revealed and what not, but also how it was revealed. Today, however, consumers exchange information on social media and review sites (Alexander, 2006). Social web is thereby changing branding, since companies are not able to completely set the agenda anymore (Fournier & Avery, 2011).

The online environment is thus not only owned by companies, but by the social collective. This environment becomes even more important since consumers use online information more and more when purchasing products (Granados et al. 2010; Pan & Fesenmaier, 2006). Hence companies have to deal with the enormous availability of information and opinions about their company and/or products. This forces companies to consider if and how they should proactively disclose information. It makes it even more difficult for companies to operate in the online environment, as it is characterized by exposure and criticism (Fournier & Avery, 2011). Next to that, consumers are more conscious about their society and environment and therefore demand more transparent products and practices. These trends force companies to openly communicate in order to maintain a good reputation (Bhaduri & Ha-Brookshire, 2011). Altogether, these

developments make it risky for companies to keep corporate wrongdoings a secret (Carter & Rogers, 2008).

Sharing information is also made a lot easier because of smartphones, which allow people to take pictures, make videos and create other content easily. Content can be send to others very fast. 75% of all digital data is now created by consumers, much of it via mobile devices people carry around with them every day (Austin & Upton, 2016). This enables people to share information within a short space of time, with a lot of other people. Sharing information via this way can also be described as stories ‘going viral’. Today’s consumers have a need-to-know mentality and are demanding information on product and business practices (Feitelberg, 2010). Web 2.0 and smartphones are making this need-to-know mentality possible (Alexander, 2006; Austin & Upton, 2016; Fournier & Avery, 2011; Kirby, 2012). Because of these information-sharing developments have formed so recently, most people and also companies have a lot of trouble figuring out what they should do with all this information and content (Austin & Upton, 2016).

These developments have led to a new age of information sharing, called the Age of Super-Transparency (Austin & Upton, 2016) or the Age of Super-Transparency (Fournier & Avery, 2011; Kirby, 2012). The Age of Super-Transparency is an overarching term that states that companies should adapt themselves to these new (online) developments. Austin and Upton (2016) provided guidelines for companies in order to let them deal with super transparency. First of all, companies should not assume that you can defend all information boundaries. Secondly, companies should look at questionable behaviors and actions within their company because it is more likely than ever that they might be leaked. Thirdly, a company should be able to respond quickly (Austin & Upton, 2016). It is thus very important for a company to lookout for shits in consumers’ perceptions of what is reasonable or acceptable. Because of the fast availability of information and the possibility to instantly share information with millions of people, the public can easily detect something they do not like about an organization. This can lead to severe reputational damage for companies. Hence it is very important for managers to prepare themselves for shits in what consumers consider fair or reasonable in regards to how a company operates (Austin & Upton, 2016). These shits will require companies to make changes. Fournier and Avery (2011) and Kirby (2012) support this view and call this phenomenon the Age of Transparency. This age is described as: everything that can be exposed, will be exposed. Fournier and Avery (2011) also state that companies have no choice but to adopt proactive positions of full disclosure and thereby be ready to make changes or at least be prepared to make them. The Age of Transparency is very much in line with the Age of Super-Transparency. Previous research already concluded that it is better for companies’ reputation to publish information themselves, rather than allowing a third party (e.g. journalists or bloggers) to expose that same information (Arpan & Roskos-Ewoldsen,

2005; Demmers, 2014). The move from secrecy to transparency can, however, be a hard one where mistakes are often made (Fournier & Avery, 2011). Some companies have stepped up to become proactively transparent, but a lot still struggle with questions like what, when, and how much information they should disclose (Austin & Upton, 2016; Fournier & Avery, 2011; Miao & Mattila, 2007).

1.2 Price transparency

Price transparency is one way for companies to be more transparent (Carter & Curry, 2010; Sinha, 2000). This type of transparency provides information about how the share of a price of a product is allocated among all major supply-side agents that bring a product to market (Carter & Curry, 2010). The strongest effects of price transparency were measured when the price was allocated to a supply-side agent that was most important for consumers, called the sympathetic agent (Carter & Curry, 2010). An example of such an agent is a coffee farmer. Consumers wonder if the price that is allocated to the coffee farmer is in line with the final selling price of coffee. Therefore, price transparency triggers social components because consumers are forced to think about if they consider these price allocations are fair (Carter & Curry, 2010). Even if consumers perceive certain pieces of information as negative, this type of transparency can still be used in companies’ advantage. Rather than evoking negative reactions, being transparent in advertisements can add value for consumers. For example, price transparency can cause an increase in trustworthiness of the advertisement or the brand (Demmers, 2014; Eisend, 2006; Schnackenberg & Tomlinson, 2014) and consumers’ willingness-to-pay (Carter & Curry, 2010; Demmers, 2014). An increase in willingness-to-pay because of transparency, however, is not always occurring. For example, Carter and Curry (2010) found a ‘fair share boundary’ that indicated that the positive effects of price transparency have its limits. When the share allocation to the sympathetic agent was above 40%, participants considered the share too generous, which made it unfair. These findings raised the question if transparency is effective when information is not positive, but negative. Even though, this specific boundary is very clear in terms of which is considered too negative, previous research fails to elaborate more on what the boundaries of negative information are (Eisend, 2006; Fournier & Avery, 2011; Miao & Mattila, 2007). More specifically, it is important for managers to know what the boundaries of negative information are in which it can still be used to increase trustworthiness and willingness-to-pay. If the positive and negative effects of negative information are not clear, then it is hard for managers to consider what they should share.

1.3 Two-sided advertising

If advertisements contain both positive and negative information one speaks of two-sided advertising (Crowley & Hoyer, 1994; Eisend, 2006). Two-sided advertising can be a useful advertising technique since it can be more effective than advertising that contains only positive information (Crowley & Hoyer, 1994; Eisend, 2006). This effect is also shown in research where the use of a certain amount of negative information leads to a higher trustworthiness and willingness-to-pay than when negative information was excluded (Carter & Curry, 2010; Crowley & Hoyer, 1994; Demmers, 2014; Eisend, 2006; Schnackenberg & Tomlinson, 2014). These effects can be explained by a trade-off that takes place between trustworthiness and negative attitudes towards the advertisement (Crowley & Holey, 1994; Eisend, 2006). Separate effects of negative information of trustworthiness and advertisement attitudes, in turn, effect willingness-to-pay. Hence the effect of negative information on willingness-to-pay is separately mediated by trustworthiness and attitude towards the advertisement. Meaning that as long as the information is not too negative, the increase in trustworthiness leads to a higher willingness-to-pay. But to what extent is too negative? Previous research only states that if information is not too negative (Eisend, 2006) or if there is a limited amount of negative information (Demmers, 2014) positive outcomes – like a higher willingness-to-pay – are measured. The boundaries wherein negative information can be used remains unclear (Eisend, 2006; Fournier & Avery, 2011; Miao & Mattila, 2007). This leads to the following research question:

What are the boundaries wherein negative information, used in price transparency, lead to a higher willingness-to-pay among consumers?

1.4 Research contribution

This research will contribute to current literature on price transparency in three ways. The first is that the boundaries of negative information in previous research are not clear (Demmers, 2014; Eisend, 2006; Fournier & Avery, 2011; Miao & Mattila, 2007). The boundary, which is found when the effects of negative information are no longer leading to a higher trustworthiness and willingness-to-pay, is tested in this research. The second contribution is that, in general, more research on price transparency and its effects is needed (Carter & Curry, 2010; Mohan, Buell & John, 2016). The third is that previous research on price transparency mainly focuses on intentions, rather than measuring actual behavior in regards to willingness-to-pay (Carter & Curry, 2010). Previous research concluded that what consumers say and actually do is very different (Jacquemet, Joule, Luchini & Shogren, 2011). It is, therefore, more accurate to measure

actual behavior. Carter and Curry (2010) also addressed this issue by suggesting that future research should focus on measuring actual behavior.

I test my theorizing in two studies. Study 1 consists of a questionnaire that was used to determine what consumers considered a fair share allocation in terms of a percentage of the sales price that was allocated to the sympathetic agent. Study 2 follows the advice of Carter and Curry (2010) by using a more accurate way of measuring willingness-to-pay via an actual auction website called Veylinx. By using this website, this research will have the opportunity to measure willingness-to-pay in an environment that is very much like the real world. In this research, an online experiment is conducted to fill in the three research gaps that are previously discussed. In Study 2, price transparency is used in advertisements by sharing information about the price allocation to the sympathetic agent, since this agent is most important for consumers. The negativity of this information is based on how negative consumers perceive that share allocation (Carter & Curry, 2010).

With the outcomes of this research, managers are given a practical, very concrete example of which share of price allocation is considered too negative and thus eliminates the positive effect of transparency on willingness-to-pay. Boundaries on negative information are very important if companies want to ask a premium price for their products, since they might decide to not make information public when consumers perceive the share allocation as too low. Not disclosing information, however, is not in line with current developments in the Age of Transparency where every piece of information that can be exposed, will be exposed (Fournier & Avery, 2011). More in line with the characteristics and consequences of the Age of Transparency would be that companies use the outcomes of this research to change the actual share that is allocated to the sympathetic agent in order to ask a premium price for their product. This means companies can determine which share allocation is considered fair among consumers and allows them to act upon that information.

The following sections will, successively, contain a literature review that discusses previous research on transparency, price transparency and its effects on negative attitudes towards the advertisement, trustworthiness and willingness-to-pay, a conceptual model that will explain this research, its variables and their relationships to one another, a method section, a results section, and finally a conclusion and discussion section where limitations and recommendations for future research are discussed.

2 Theoretical framework

2.1 Transparency

Research on transparency focuses on several study domains, including financial markets (Bloomfield & O’Hara, 1999; Eijffinger & Geraats, 2006), organizational governance (Bushman, Piotroski, & Smith, 2004), negotiations (Vorauer & Claude, 1998), and electronic markets (Granados et al., 2010; Zhu, 2004). Even though a lot of practitioners have called for greater transparency, very little is actually known how organizations manage it (Schnackenberg & Tomlinson, 2014). This is also caused by the fact that there is currently very little consensus about the meaning of transparency on a conceptual level (Schnackenberg & Tomlinson, 2014). Interestingly, literature on transparency is ‘classified’ in two sections. These two ‘classifications’ also indicate why there is little consensus on the meaning of transparency on a conceptual level. Some conceptualizations of transparency focus more on descriptions stating that a market is transparent when information is available. Other conceptualizations focus more on descriptions stating that a company is more transparent when it provides information. To clarify this distinction in conceptualizations, Table 1 provides an overview of conceptualizations categorized in these two sections.

As shown in Table 1, the availability of information in the market (Bloomfield & O’Hara, 1999; Bushman et al., 2004; Ganados et al., 2010; Zhu, 2004), while others focus on the firm that is more transparent when it provides information (Eijffinger & Geraats, 2006; Vorauer & Claude, 1998). All these definitions, from all these different study domains, have in common that transparency is about information (Bhaduri & Ha-Brookshire, 2011; Bushman et al., 2004; Granados et al., 2010; Schnackenberg & Tomlinson, 2014; Zhu, 2004). This explains why one talks about Web 2.0 as a cause for more transparency, because this technology increases information availability and accessibility. It thereby outlines transparency’s alignment with information.

Table 1. Definitions of transparency classified in two sections

Market is transparent when

information is available

Company is transparent when it

provides information

‘’The real-time, public dissemination of trade and quote information.’’

(Bloomfield & O’Hara, 1999)

‘’The extent to which central banks disclose information that is related to the policy-making process.’’

(Eijffinger & Geraats, 2006) ‘’ The availability of firm-specific information

to those outside publicly traded firms.’’ (Bushman, Piotroski, & Smith, 2004)

‘’ The degree to which an individual’s objectives are readily apparent to others.’’ (Vorauer & Claude, 1998)

‘’ The availability and accessibility of market information to interested parties.’’

(Granados et al., 2010)

‘’The degree of visibility and accessibility of information.’’

(Zhu, 2004)

In line with enhanced information availability and accessibility is the definition of Zhu (2004) that defines transparency as the availability, accessibility, and visibility of information to its participants. This definition is somewhat expanded by Bhaduri and Ha-Brookshire (2011) who state that transparency is defined as visibility and accessibility of information, especially concerning business practices. Granados et al. (2010) developed a very similar definition which states that transparency is the ‘’availability and access of market information to interested parties’’. Here, not only quantity of information, but also the quality of how this information is being presented is important. Transparency is thereby different from information availability since it says something about the intention of a company to disclose information (Granados et al., 2010). This is an important aspect, because a company can also conceal information if it feels that being transparent could harm the company. For example, this happened in the Volkswagen scandal where eleven million cars were affected by a defeat device, which is software that allowed cheating in emissions tests (Winston, 2015). As predicted by Fournier and Avery (2011), every piece of information will be revealed in the Age of Transparency, which is exactly what happened with the Volkswagen scandal. One could thus argue that in today’s online environment, it should always be the company’s intention to share information and thereby be transparent.

It is also important to address that in today’s environment information does not follow a one-way route. As stated, Web 2.0 allows consumers to play an important role in producing information,

thereby shifting the power over information towards consumers, rather than remaining in the hands of companies (Fournier & Avery, 2011). It is, therefore, important to understand how consumers perceive transparency (Walther, 2004). The smallest changes in how information is being displayed can have its impact on how consumers perceive it. For example, stating that food contains 10% fat is perceived very different than when it says: for 90% fat free (Thaler & Tucker, 2013). 10% fat may sound way more negative than 90% fat free. Therefore, perceived transparency is an important aspect that should be taken into consideration when defining transparency. Zhu (2004) addresses the importance of perceived transparency by talking about the visibility of information. Next to that, Busman et al. (2004) also address the importance of quality of information.

To summarize, transparency is about information that is intentionally shared, taking the perception and quality of that information (Schnackenberg & Tomlinson, 2014). The quality of information consists of three components: disclosure, clarity and accuracy. The way consumers perceive transparency is in line with the clarity component that is enhanced as stakeholders understand information better. Next to that, disclosure is enhanced as stakeholders perceive information as more relevant and timely, and accuracy is enhanced as stakeholders perceive information as more reliable. This leads to the following definition: ‘’transparency is the perceived quality of intentionally shared information from a sender’’ (Schnackenberg & Tomlinson, 2014).

2.2 How to be transparent?

There are several ways for companies to be more transparent. Granados et al. (2010) developed a framework, shown in Figure 1, focused on business to consumer transparency. This framework describes all components important in that area. Even though some components are defined differently in other

articles, the framework does provide a good overview of relevant aspects one needs to

consider when

conducting research in the area of transparency. For example, it

describes who sends the Figure 1. Framework Business-to-Consumer Transparency (Granados et al., 2010)

information (component A, Figure 2) and who is receiving it (component B, Figure 2). Next to that, actions that a company can undertake in the area of transparency are also mentioned (component D, Figure 2). For example, a company can choose to disclose information, which means it is fully transparent about product information and price information. But, companies can also choose to conduct a more adapted form of transparency by giving distorted, biased or concealed information. When a company chooses one of these options, it becomes less transparent by revealing only preferable price and product information (biased) or only product information (opaque). It can even choose to give old information which is outdated (distorted). In the end, all companies are forced to become more transparent, so in the long term the only option would be to fully disclose price and product information (Fournier & Avery, 2011). The final important component in this framework is the elements that information consist of (component C, Figure 2). Information can be about a lot of things, like the product itself, its price, the inventory that is available, how much it costs to produce a product and about the process needed to produce and/or transport the product. A more transparent company will be more transparent in one of these elements (Granados et al., 2010). For example, Demmers (2014) and Carter and Curry (2010) conducted a research on price transparency where information was about the share of the selling price of a product allocated to a supply-side agent. This research will focus on price transparency as well, since the demand for transparency is especially wanted in the global supply chain, because of increased awareness of the environment and ethical social issues, like proper labor (Bhaduri et al., 2011). The next section will elaborate more on price transparency.

2.3 Price transparency

This paper will focus on price transparency because price transparency is an important aspect nowadays due to the positive effects of a transparent supply chain on a company’s image, but also on customer loyalty (Strutnin, 2008). Consumers’ interest go much further than only the brand. As companies have discovered, it is useless to talk about ethical behavior when your supply-side agent acts unethically (New, 2010). The need for a transparent supply chain is also caused by the pressing needs for sustainability and sustainable supply chains (Bhaduri & Ha-Brookshire, 2011). Price transparency is defined as ‘’disclosing beneficiaries of a product's revenues; for example, by dividing a price into gross retail proceeds, royalties, and taxes’’ (Carter and Curry, 2010). More specifically, price transparency reveals information about the allocation of price among supply-side agents. An example of price transparency is that an advertisement about a music CD states that a share allocation of 24% of the sales price is provided to the artist (Carter & Curry, 2010). Demmers (2014) uses a similar example in his research where a share allocation of the sales price of a fruit juice went to the grower. In both examples, the artist and the

fruit-grower are the most important supply-side agent for consumers, since consumers can most relate to that agent when they think of CDs or fruit juice as final products. This type of agent is also called the sympathetic agent (Carter & Curry, 2010). Because effects of price transparency were found to be most strongly among the share allocation to the sympathetic agent, this research will also focus on this supply-side agent.

Other definitions of price transparency state that price transparency is about information of other prices within the market (Granados et al., 2010; Grewal et al., 2003), which is not in line with the definition of Carter and Curry (2010). Information of other prices within the market is an important aspect that is provided by the internet and addresses an important issue for companies as well. Because this information is widely available, consumers will compare prices more easily which makes it harder for companies to ask a premium price (Grewal et al., 2003; Sinha, 2000). However, this type of price transparency is not discussed in this research, since this research will build upon the results of Carter and Curry (2010). Also, price transparency is sometimes confused with cost transparency. However, cost transparency is about information that makes sellers’ costs more transparent to buyers (Sinha, 2000; Zhu, 2004). Those costs could be about manufacturing costs or costs of materials. Zhu (2004) defines cost transparency as ‘’the voluntary disclosure of the costs associated with producing a good or providing a service’’. The difference with price transparency defined by Carter and Curry (2010) is that cost transparency does not reveal any information about price allocations. Cost transparency only states what different materials and processes that are used to produce a certain product cost. It is then up to consumers to determine if the final price of the product, based on these costs, is fair (Sinha, 2000). In the next sections effects of transparency and, more specifically, price transparency are discussed.

2.4 Effects of transparency

Operating in the Age of Transparency can be beneficial for companies. For example, ‘going viral’ can be used for marketing purposes, since information can be spread easily and fast. This comes, however, with risks, since companies do not have control over where the information is going and how people will react upon it. Moreover, if companies are transparent about prices, operations and other components, this information could contain negative information as well. Companies need to deal with these kinds of issues. It is, therefore, important for companies to know which effects transparent advertising has on consumers, especially in regards to negative information.

2.4.1 Willingness-to-pay

Transparency has positive effects on willingness-to-pay (Carter & Curry, 2010; Demmers, 2014). Willingness-to-pay is defined as “the denotation of the maximum price a buyer is willing to pay for a given quantity of a good” (Wertenbroch & Skiera, 2002). Willingness-to-pay is an important measurement to estimate consumers’ demand and is also used to construct optimal pricing schedules. Previous research, however, failed to measure actual behavior in regards to willingness-to-pay. Rather, consumers were asked what they intended to pay by measuring purchase intention (Carter & Curry, 2010). Unfortunately, this is a poor measurement because, most of the time, consumers act differently than they say they will. Therefore, measuring consumers’ behavior by asking how they will intent to behave is not desirable (Jacquemet et al., 2011). Carter and Curry (2010) also addressed this issue by suggesting future research on price transparency and its effects should focus on measuring actual behavior. To do this, demand revealing auctions are suggested (Noussair, Robin & Ruffieux, 2004). The idea behind these auctions is that consumers truly commit real money when they bid for a product. Hence by involving real money, more truthful information about consumers’ willingness-to-pay is measured. The bid that an individual proposes is a measurement that directly displays the limit he or she is willing to offer. One of the most used auctions is the Vickrey auction (Vickrey, 1961). This auction is described as a sealed-bid auction, meaning that individuals propose a bid without knowing what other individuals bid. The individual that wins the auction is the one that offered the highest bid, but pays for the second-highest bid. Another auction, called the Becker-DeGroot-Marschak (BDM) (Becker, DeGroot & Becker-DeGroot-Marschak, 1964), is also commonly used, but research concluded that a Vickrey auction is a more effective way to measure willingness-to-pay than the BDM auction (Noussair et al., 2004).

Transparency and price transparency lead to an increase in willingness-to-pay, to a certain limit at least (Carter & Curry, 2010; Demmers, 2014). Being transparent can also mean that negative information is revealed. Interestingly, even when messaging was transparent and negative information was included, an increase in willingness-to-pay was still measured. This increase in willingness-to-pay stopped when information was considered too negative (Carter & Curry, 2010). Findings in other research show that this U-shaped relationship is also caused by two other variables – trustworthiness and negative attitude towards the advertisement - that have a mediating effect on the relationship between transparency and willingness-to-pay. In the next section, trustworthiness, negative attitude towards the advertisement and their relationship with willingness-to-pay are discussed.

2.4.2 Trustworthiness

It is already concluded that it is better for the public image of a company to publish information themselves rather than allowing a third party, like journalists or consumer reviews, to reveal that same information (Arpan & Roskos-Ewoldsen, 2005). This has to do with trustworthiness or trust in the company or brand that is enhanced because of more transparent messaging (Demmers, 2014; Eisend, 2006; Schnackenberg & Tomlinson, 2014). Companies that are transparent are perceived as more trustworthy. Trustworthiness or trust is defined as ‘’the mutual confidence that no party to an exchange will exploit another’s vulnerabilities’’ (Barney & Hansen, 1994). This definition already provides insight into why trustworthiness and transparency are so intertwined, because transparency allows consumers to look into the company’s ‘black box’ and thereby makes the consumer less vulnerable than when a company keeps secrets (Kirby, 2012). The relationship between trustworthiness and transparency is shown in other research domains as well. For example, research that is focused on public administration. It is widely recognized that transparency is considered a key value for trustworthy governments (Grimmelikhuijsen, Porumbescu, Hong & Im, 2013). As pointed out by Grimmelikhuijsenet et al. (2013) trust or trustworthiness is defined in so many ways and is used in so many disciplines, that a more interdisciplinary definition is preferable: ‘’trust is a psychological state comprising the intention to accept vulnerability based upon positive expectations of the intentions or behavior of another’’ (Rousseau, Sitkin, Burt & Camerer, 1998). All definitions of trustworthiness have in common that there is some kind of positive expectation in terms of intentions of the one that is provoking trust (Grimmelikhuijsen et al., 2013).

The Age of Transparency and transparency are intertwined with trustworthiness. For example, Peppers and Rogers (2013) state that in today’s transparent environment, companies should ‘go to extremes’ to show that they care for their consumers. In order for companies to do that, they must achieve trustworthiness. In their book, Peppers and Rogers (2012) describe the relationship between transparency and trustworthiness. Their concern is that more transparency will make it harder for companies to be trusted. This is in line with the difficulty companies face when being more transparent (Fournier & Avery, 2011). Even though Peppers and Rogers (2012) contribute to explain how interrelated transparency and trustworthiness are, they contradict other authors who state that greater transparency leads to greater trustworthiness (Crowley & Holey, 1994; Eisend, 2006; Schnackenberg & Tomlinson, 2014). Schnackenberg and Tomlinson (2014) state that transparency is essential to reassure stakeholder trust in a company. Moreover, it is an essential component for the trust that stakeholders place in an organization and transparency is also seen as an antecedent of trustworthiness (Schnackenberg & Tomlinson, 2014). Other research states that if a company is more transparent it will enhance its credibility (Crowley & Holey,

1994; Eisend, 2006). The increase in credibility is in line with enhanced trustworthiness, since trustworthiness can be seen as a component of credibility (Ohanian, 1990). It is also stated that customers do not trust companies unless they are transparent (Kirby, 2012). This research follows the findings of authors that have concluded that greater transparency leads to greater trustworthiness.

Taken all these findings together, it is suggested that transparency, trustworthiness and willingness-to-pay are interrelated. It is namely found that greater trustworthiness leads to a higher willingness-to-pay (Carter & Curry, 2010; Crowley & Hoyer, 1994; Demmers, 2014; Eisend, 2006; Schnackenberg & Tomlinson, 2014). In this research, being more transparent means that a company is using negative information in its advertising as well. When negative information is used in its advertising, consumers are likely to perceive that company as more trustworthy (Crowley & Hoyer, 1994). The attribution theory explains why messages that include negative information are perceived as more truthful than messages that do not. The theory describes the process that individuals undergo in order to explain the causes of certain behavior or events (Kelley, 1967). ‘’When consumers buy a product they ‘’attribute’’ claims to the advertiser’s desire to sell the product (one-sided advertising) or to actual characteristics of the product communicated by an honest advertiser (two-sided advertising)’’ (Crowley & Hoyer, 1994). This means that when negative information is included in messaging, consumers often have the idea that the advertiser is telling the truth. When consumers have this idea, this increases the advertisers’ trustworthiness.

So, using negative information or revealing it can actually be beneficial for companies, since it has stronger effects than when negative information is excluded. If an advertisement contains both positive and negative information, one speaks of two-sided advertising (Crowley & Hoyer, 1994; Eisend, 2006). Companies that use two-sided advertising are better off than companies who use merely positive messages in their advertisements. This is because two-sided messages are more effective since it increases trustworthiness and willingness-to-pay more than when negative information was excluded (Carter & Curry, 2010; Crowley & Hoyer, 1994; Demmers, 2014; Eisend, 2006; Schnackenberg & Tomlinson, 2014). Including negative information within an advertisement shows that a company is willingly vulnerable and therefore enhances positive expectations among consumers (Schackenberg & Tomlinson, 2014). It also means that a company is more transparent when it includes more information, even negative information. This research will focus on including negative information as a form of transparency. By this, previous research clearly illustrates the relationship between transparency, trustworthiness and willingness-to-pay. So, it is suggested that there is an indirect effect of negative information (a form of transparency

in this research) on willingness-to-pay via trustworthiness. The following hypothesis is formulated:

H1: There is an indirect effect of negative information on willingness-to-pay via trustworthiness

2.4.3 Negative attitude towards the advertisement

Including negative information does, however, not always lead to a higher willingness-to-pay (Carter & Curry, 2010). The negative attitude towards the advertisement is said to increase when negative information is included in an advertisement (Crowley & Hoyer, 1994; Eisend, 2006). This will cause a negative effect on willingness-to-pay. So, when a two-sided advertisement increases trustworthiness (Eisend, 2006; Schnackenberg & Tomlinson, 2014) it, on the other hand, increases negative attitudes towards the advertisement since some information can be perceived as negative (Crowley & Holey, 1994; Eisend, 2006). This describes a trade-off between trustworthiness and negative attitude towards the advertisement. As long as information is not too negative, two-sided messaging will have a positive effect on trustworthiness and on willingness-to-pay (Carter & Curry, 2010; Demmers, 2014; Eisend, 2006). As stated, negative information in this research is based on the share allocation of the sales price to the sympathetic supply-agent. Moreover, negativity of that information is based on what consumers consider a fair share (Carter & Curry, 2010; Demmers, 2014). Including negative information within an advertisement is a form of being more transparent and therefore relates to the effects transparency has on trustworthiness and willingness-to-pay. The following hypothesis is formulated:

H2: There is an indirect effect of negative information on willingness-to-pay via trustworthiness via negative attitude towards the advertisement

Previous research fails to answer the question: what is too negative? Eisend (2006) describes this by: ‘’It is not unlikely that extreme amount of negative information could also lead to low credibility. However, it is unlikely that an advertiser would include more negative information than was absolutely necessary…’’. It is further stated that the disclosure of negative information leads to higher source credibility, less negative cognitions, and more favorable attitudes. But, this is only caused if negative information meets specific requirements. For example, the amount of negative information should not be too large and it should not be about something important. The question ‘what is too negative?’ remains unclear and this research will therefore try to find the specific boundaries of negative information. The boundary is found when the positive effects of two-sided advertising on trustworthiness and willingness-to-pay do not occur anymore and thereby lower trustworthiness and willingness-to-pay.

2.4.4 Consumers’ involvement

As stated, negative information will be based on the share of the selling price allocated to the sympathetic supply-agent. This will have a direct effect on negative attitude towards the advertisement, but this relationship is influenced by consumers’ involvement. Previous research found that consumers’ involvement is an important, moderating variable, when it comes to consumer behavior and attitudes (Bloch, 1981). Involvement is defined by Day (1970) as ‘’the general level of interest in the object or the centrality of the object to the person’s ego-structure’’. Previous research already stated that high involvement could change attitude more strongly than low involvement (Apsler & Sears, 1968). It is thus likely to assume that consumers’ involvement will have a moderating effect on the negative attitude towards the advertisement. The following hypothesis is formulated.

Hypothesis 3: the relationship between negativity of information and negative attitude towards the advertisement is moderated by consumers’ involvement

Figure 2 displays a conceptual model that illustrates the relationship between all variables. Also, Table 2 provides an overview of all hypotheses.

Table 2. Hypotheses

Hypotheses

H1: There is an indirect effect of negative information on willingness-to-pay via trustworthiness

H2: There is an indirect effect of negative information on willingness-to-pay via trustworthiness via negative attitude towards the advertisement

Hypothesis 3: the relationship between negativity of information and negative attitude towards the advertisement is moderated by consumers’ involvement

3 Study 1

This study was conducted in order to measure what consumers considered a fair share allocation to the sympathetic agent. A face-to-face, on-site data collection with pencil and paper was used in order to collect data to see what consumers consider a fair share allocation to the supply-side agent, with tea as end product, since tea is used in this research as end-product. This research method was chosen because of the limited interest that students have to participate in online surveys (Lefever & Matthiasdottir, 2007). The data collection took place in the city center of The Hague on the 22nd _{and 23}rd _{of July 2017. Respondents were randomly approached and asked if} they had time for a few short questions for this research. Their anonymity was discussed and confirmed. When they agreed upon participation, they were asked: ‘’What would you consider a fair percentage of the selling price that goes to the tea farmer, when you buy tea?’’ (or in Dutch: ‘’Wat vindt u een eerlijk percentage van de verkoopprijs van thee dat naar de theeboer gaat?’’). Sometimes respondents did not completely understand the question, so it was sometimes rephrased by saying: ‘’you have the selling price of tea; what percentage of that selling price, do you think, should go to the tea farmer?’’. 79 respondents participated in this research (N = 79). The average percentage of the selling price allocated to the tea farmer was set on 21,32% with a standard deviation of 21,16% (M = 21,32; SD = 21,16). The minimum percentage was 0% and the maximum percentage was 100%. The results are shown in Table 3.

Table 3. Descriptives Study 1

N Minimum Maximum Mean Std. Dev.

Percentage 79 0,00% 100% 21,32% 21,17%

Even though, Study 1 concludes that consumers find 21,32% share allocation to the tea farmer fair (when buying tea), Study 2 uses slightly different percentages. This is due to the fact that consumers differ greatly in estimating what share of the selling price should be allocated to the supply-side agent, with shares ranging from 0% to 60% (Carter & Curry, 2010). Therefore, Carter and Curry (2010) used external sources to determine that a 24% share allocation is considered fair for CDs. Because of these external sources and because of the big variances in percentages measured in Study 1, it is determined that even though this product group is different from tea, Study 2 will take 24% share allocations as what is considered ‘normal’ by consumers. This is because 24% is found to be close enough to 21,32% that measured in Study 1.

4 Study 2

4.1 Method

In order to measure actual behavior, that is suggested by previous research (Carter & Curry, 2010), because it is more reliable than asking consumers for intentions (Jacquemet et al., 2002), this research will use an actual auction website called Velinx. This website enables an auction of a product on which consumers can offer a sealed-bid. No other bids are shown and consumers can only offer one bid. The consumer with the highest bid wins the auction and has to pay the second-highest offer. Wertenbroch and Skiera (2002) describe this sort of auction as a Vickrey auction (Vickrey, 1961). It is a more truthful experiment since consumers actually have to buy the product.

4.2 Design

By using the auction website Veylinx, I will conduct one online experiment. This is a simple between-subjects design with five conditions. These conditions are based on previous research conducted by Carter and Curry (2010). The authors used the share of the selling price allocated to the supply-agent. These shares are displayed in percentages and are based upon previous research that already identified what consumers consider a ‘fair’ share, namely 24% (Carter & Curry, 2010). Since the goal of this research is to investigate how negative information can be, the percentages

drops, making the five conditions 24%, 16%, 8%, 4% and 2%. These are the five treatments that will be used in this experiment. Study 1 measured a percentage of 21,32%, but since the goal of this research is to find the boundary of negative information, the starting point of 24% or 21,32% should not make that much of a difference. It is assumed that the more interesting results will arise for the lower percentages. The share is allocated to a tea farmer, using a tea box with six different flavors of tea as the product that is advertised. Figure 4 displays an example of an advertisement used in the experiment. The advertisement is in Dutch since most consumers that use Veylinx are Dutch. All advertisements that are used can be found in the Appendix. Nothing but a short description is used as text with as little as possible visual effects because this might influence consumers’ perceptions. Tea is specifically chosen, because it is easy to relate with a tea farmer as the sympathetic agent for a product like tea. Also, it is a common good which is in line with previous research that used fruit juice whereby the share went to the fruit-grower (Demmers, 2014). Next to measuring willingness-to-pay based on these five conditions, Veylinx offers the ability to include a few survey questions. These questions were used to measure other dependent, mediating and moderating variables. These variables are discussed in the next section.

Figure 4. Example of advertisement used in experiment (in Dutch)

4.3 Procedure

In order to participate in an auction organized by Veylinx, one should register on the website. A few general details are requested, like name, age, gender and e-mail. Veylinx already has a substantial amount of people registered on their website, but in order for this research to be conducted 100 new subscribers were to be found. Via social media and direct contacts 102 people subscribed to the website. Participants are then invited for an auction via e-mail. They can decide whether they participate or not. When people participate they are shown one of the five advertisements (which contain one of the conditions) and are requested to offer a bid in Eurocents. Participants have five minutes to offer a bid and are also able to offer €0, -. After they placed a bid, they are asked a few survey questions. Those questions measured the variables discussed in section 3.3. The highest bid wins the auction and has to pay the second highest bid.

4.4 Variables

The independent variable is this research is negativity of information. This variable is, as stated before, is based on the share of the selling price that is allocated to the tea farmer. The variable will consist of five treatments: 24%, 16%, 8%, 4% and 2%.

Three dependent variables were measured in this research: willingness-to-pay, negative attitude towards the advertisement and trustworthiness. Negative attitude towards the advertisement and trustworthiness are also expected to mediate the relationship between negativity of information and willingness-to-pay. Since the auction website Veylinx offers limited space for survey questions, the variables trustworthiness and negative attitude towards the advertisement were measured by asking only one question. Trustworthiness was measured by asking: ‘I found the advertisement…(1) very untrustworthy, (2) untrustworthy, (3) neutral, (4) trustworthy, (5) very trustworthy’. Negative attitude towards the advertisment was measured in a similar way by asking: ‘I found the advertisement…(1) very negative, (2) negative, (3) neutral, (4) positive, (5) very positive’. Here, trustworthiness and negative attitude towards the advertisement are used as single-item measures. Single item measures are recommended by several authors (Bergkvist & Rossiter, 2007). They state that marketing researchers should use single-item measures for constructs that are ‘’doubly concrete’’. These are constructs with a clear objective, like attitude towards the advertisement (Bergkvist & Rossiter, 2009). Multi-item measures are sometimes accurate since respondents can become bored or tired by filling in all these questions. Besides, it is argued that multi-item measures are costly and time consuming (Bergkvist, 2015).

Also, one moderating variable was included in this research, namely consumers’ involvement. It is expected to moderate the effect of negativity of information on negative attitude towards the advertisement. Again, there was only room for one question to measure consumers’ involvement, so it was measured by asking: ‘To which extent do you consider it important that a certain amount of money is allocated to the tea farmer?’ The options to answer are: (1) extremely important, (2) very important, (3) moderately important, (4) slightly important, (5) not important at all. Again, the arguments that acknowledge the power of single-item measures is taken into account.

4.5 Sample

Veylinx is a Dutch auction website with mostly Dutch participants. A total of 851 people participated in the online auction but 28 participants were removed since they failed to complete the survey question. This left a total of 823 participants (N = 823) that participate in the online auction. 402 of them are male (51.2%) and 421 are female (48.8%). Ages differed between 18 and 79 years with an average age of 44 years (M = 44, SD = 13.8). The average bid was €5.19 (M = 5.19, SD = 5.82). The division of participants among the five conditions is quite equally. Table 2 provides an overview of the descriptive statistics of all variables used in this paper. Also, Table 2 provides an overview of Pearson’s correlations between all variables. It is shown that attitude towards the advertisement weak to moderate predictor of willingness-to-pay with a Pearson correlation coefficient of r = -0.341 with a significance value less than 0.001 (p < 0.001). It is also shown that trustworthiness is a weak to moderate predictor of willingness-to-pay with a Pearson correlation coefficient of r = -0.301 with a significance value less than 0.001 (p < 0.001). Trustworthiness is also a moderate to strong predictor of attitude towards the advertisement (r = 0.634, p < 0.001). Finally, consumers’ involvement is a weak to moderate predictor of attitude towards the advertisement (r = 0.342, p < 0.001).

Table 4. Means, Standard Deviations and Correlations Variable M SD 1 2 3 4 5 6 7 1. Gender .49 .50 - 2. Age 43.46 13.77 -.014 - 3. WTP (€) 5.20 5.82 -.080* .014 - 4. Treatment (Negativity of info) 10.77% 8.2% -.006 -.019 .001 - 5. Attitude towards ad 2.44 .78 -.063 .094** -.341** -.061 - 6. Trustworthiness 2.54 .69 -.035 .062 -.301** -.063 .634** - 7. Involvement 2.16 .81 -.089* .131** -.133** -.021 .342** .290** -

** Correlation is significant at the 0.01 level (2-tailed) * Correlation is significant at the 0.05 level (2-tailed)

4.6 Testing hypotheses

To test H1 and H2, a multiple mediation analysis was conducted with negativity of information as the independent variable (X), willingness-to-pay as the dependent variable (Y) and negative attitude towards the advertisement (M1) and trustworthiness (M2) as mediating variables. The mediation analysis is done by Process Hayes, using Model 7 (Hayes, 2013). To test H3, a moderation analysis was conducted with negativity of information as the independent variable (X), negative attitude towards the advertisement as dependent (Y) and consumers’ involvement as the moderating variable (M) which is done by Process Hayes, using Model 1 (Hayes, 2013).

5 Results

5.1 Multiple mediation analysis

The multiple mediation analysis showed three significant results. First of all, there was a significant effect of negative attitude towards the advertisement on willingnesstopay (b1 = -0.322, p < 0.001). Meaning that two consumers who differ one unit in the negative attitude towards the advertisement are estimated to differ a = -0.332 on willingness-to-pay. This relationship is negatively significant (p < 0.001). Second of all, there was a significant effect of trustworthiness on willingness-to-pay (b2 = -0.207, p < 0.001). Two consumers who differ one unit in trustworthiness are estimated to differ by a = -2.07 units on willingness-to-pay. This relationship is negatively significant (p < 0.001). Third of all, there is a positive significant effect of negative attitude towards the advertisement on trustworthiness (a3 = 0.555, p < 0.001). Two consumers who differ one unit on negative attitude towards the advertisement are estimated to differ a = 0.555 units on trustworthiness.

There were also some insignificant effects measured. The direct effect of negative information on willingness-to-pay is not significant, c’ = -0.003, t(819) = -0.703, p = 0.482. The relationship between negative information and trustworthiness is also found to be insignificant (a2 = -0.002, p = 0.372). No indirect on negative information on willingness-to-pay, via trustworthiness was found. H1 is therefore not supported. Finally, the relationship between negative information and negative attitude towards the advertisement was not significant (a1 = -0.006, p = 0.081). No indirect effect of negative information on willingness-to-pay, via negative attitude towards the advertisement was found. This means that H2 is not supported. All results are summarized in Table 5.

5.2 Moderation analysis

The moderation analysis (results summarized in Table 6) shows that the moderating effect on consumers’ involvement on the relationship between negative information and negative attitude towards the advertisement is not significant (c3 = 0.002, p = 0.388). H3 is thereby not supported.

Table 5 . M u lt ip le med ia ti o n anal ys is Cons eq u en t M 1 (N ega tive att itud e tow ar d s the ad.) M 2 (Tr u stw or thin ess) (Y ) Will in gn ess -to -p ay A nte ce d ent C oef f. SE p C oef f. SE p C oef f. SE p N egat ivi ty o f i n fo a1 -. 006 0 .003 0 .081 a2 -0 .002 0. 002 0. 372 c' -0. 003 0. 004 0. 482 M1 a3 0. 555 0. 024 <0. 001 b1 -0. 322 0. 054 <0. 001 M2 b2 -0. 207 0 .061 <0. 001 C onst ant iM1 2. 501 0 .045 <0 .001 iM2 1. 214 0. 067 <0. 001 iY 1. 342 0. 139 <0. 001 R 2 = 0. 004 F(1, 821 ) = 3. 048, p = 0. 08 1 R 2 = 0. 4 02 F(2, 820 ) = 2 75.901, p < 0. 001 R 2 = 0. 129 F(3, 819 ) = 4 0. 372, p < 0. 0 01 Table 6 . Mo der a ti on ana lys is C oef f. SE t p Int er ce p t i1 N egat ivi ty o f in fo ( X ) c1 -0.009 0.011 -0.818 0.414 Invol v em ent (M ) c2 0.310 0.065 4.788 < 0.001 XM -inte rac ti on c3 0.002 0.005 0.388 0.698 R 2 = 0.120 F(3, 819 ) = 2 4.811, p < 0.0 01

6 Discussion and conclusion

This research was intended to contribute to current literature in three ways: (1) to specify the boundaries of negative information. Therefore, the main question in is this research is: What are the boundaries wherein negative information, used in price transparency, lead to a higher willingness-to-pay among consumers? By manipulating the negativity of information, it was hoped to gain more insights into when consumers found information to be too negative. The second goal (2) of this research was to gain more insights into price transparency and its effects. Thirdly (3), this research contributes to previous literature on price transparency by using measurements for actual behavior, rather than (purchase) intentions. No support was found for H1, H2 and H3.

Unfortunately, this research did not find a boundary of negative information, because there was no significant, direct effect found of negative information on willingness-to-pay. That no direct effect was found is not in line with previous research that did find a positive relationship between negative information and willingness-to-pay (Carter & Curry, 2010; Crowley & Hoyer, 1994; Demmers, 2014; Eisend, 2006; Schnackenberg & Tomlinson, 2014). The findings in this research can be explained by several factors. First of all, Carter and Curry (2010) specifically addressed in their research that a higher willingness-to-pay was also moderated by characteristics of the decision-maker. If consumers have different ideas about what is considered negative or fair, then other effects may have been found. For example, some cultures are more focused on fitting in and others are much more individual and thereby differ in what is considered fair (Mattila and Patterson, 2004). This is also supported by other authors (Grimmelikhuijsen et al., 2013). It may be that the Dutch respondents in this research differ in what they perceive as fair and how that influences them, compared to the American respondents used in Carter and Curry’s (2010) research. However, Study 1 did ask Dutch consumers, specifically, what they considered a fair share. Hence cultural differences may have its influence, but it seems that other factors might explain the difference in findings of this research compared to previous research better.

This research did find three significant effects, which might help to explain why no direct relationship between negative information and willingness-to-pay was found. Even though H1 and H2 were not supported, there is still a significant relationship found between trustworthiness and willingness-to-pay and negative attitude towards the advertisement and willingness-to-pay. These findings support previous findings that also confirm this relationship (Crowley & Holey, 1994; Eisend, 2006). This is one contribution that this research provides, because it has proven this relationship by using a Vickrey auction and thereby measuring actual behavior, rather than intentions. Since Vickrey auctions are said to have much more realistic measurements than intentions (Jacquemet et al., 2002; Noussair et al., 2004), this research concludes that the previous found relationship between trustworthiness, negative attitude towards the advertisement and

willingness-to-pay still exists. These findings also indicate that the variable negative information explains why previous research found opposite results.

An explanation could be that respondents in Study 1 gave socially desirable answers. This can be caused by the fact that Study 1 was conducted via an on-site data collection questionnaire. Because respondents were asked about what they considered a fair share allocation to a tea farmer, social components were triggered (Carter & Curry, 2010). Results of this study may be influenced by social desirability, which is more at hand when face-to-face interviews are conducted (Heerwegh, 2009). People might feel embarrassed when they are asked about a fair share allocation and may propose a higher share than when they are asked online, via a survey. This might also explain why two respondents in Study 1 proposed 100% as share allocation. One respondent did even elaborate on this percentage stating that she really feels big companies do not have the right to take the income of the tea farmer by taking unfair shares. These reactions may indicate that a different percentage could have been measured when another survey method was used. Thus, the negative information may not have been measured the way people really perceive it.

Managers can conclude from this research that negative information is a very hard to measure variable, which is very dependent on cultural influences, what might explain socially desirable answers as well. However, in the Age of Transparency, nothing remains a secret, so they have to prepare to be exposed. In order to be prepared they must dive deeper into understanding what consumers perceive as negative and how this influences their attitude towards advertisements, the company itself and willingness-to-pay. Previous research did find positive effects of negative information used in advertising. However, this research shows how hard it can be to determine what consumers perceive as negative. Since there are also many different forms of negative information, managers truly need to be careful how they operate in the super transparent world. It is only sure that they have no choice but to prepare themselves.

7 Future research

This research also knows some limitations. It concluded that H1, H2 and H3 were not supported. Explanations for these findings could be in the fact that all constructs were measured by single-item measures. Even though authors do support the use of single-single-item measures (Bergkvist, 2015), there are also authors who state that using single-item measures should be discouraged. For example, Wanous and Reichers (1996) state that single-item measures are discouraged because their reliability cannot be tested. Other authors also presume single-item measure have low

reliability and even call the use of it in academic research a ‘’fatal error’’ (Wanous, Reichers & Hudy, 1997). It is a possibility that if multi-item measurements were used in this research, results would be different because reliability of the constructs could have been tested. However, articles that support the use of multi-item measurements are a bit outdated. Students are less willing to participate in online research (Lefever & Matthiasdottir, 2007), so it may be that the benefits of single-item measurements overrule the downsides. To test this, future research should combine the findings of these articles and conduct multiple studies containing multi-item and single-item measures in order to compare the effects of the two.

Next, the auction website Veylinx had limited room for questions. The results of this study may be influenced by the fact that consumers did not see the share allocation that went to the tea farmer. Even though text was limited in the advertisement, it could be that consumers did not read the text properly. To test this, a manipulation check could have been conducted. If consumers did not read the text properly, they might not have been influenced by the manipulation. A control condition could also have been added in order to test this. Future research that use an auction website like Veylinx should try to incorporate such a manipulation check. Also, to overcome the differences in responses in a study that is conducted via a face-to-face interview and an online study, future research should follow the advice of Heerwegh (2009) who states that mixed-mode survey designs may decline effects of social desirable answers by using both web surveys and surveys conducted by an interviewer. In that case, assumptions about whether or not social desirable answers did influence results are out ruled.

8 References

- Alexander, B. (2006). Web 2.0. A New Wave of Innovation for Teachning and learning, 32-44.

- Anderson, N. H. (1971). Integration theory and attitude change. Psychological review, 78(3), 171.

- Apsler, R., & Sears, D. O. (1968). Warning, personal involvement, and attitude change. Journal of personality and social psychology, 9(2p1), 162.

- Arpan, L. M., & Roskos-Ewoldsen, D. R. (2005). Stealing thunder: Analysis of the effects of proactive disclosure of crisis information. Public Relations Review, 31(3), 425-433.

- Barney, J. B., & Hansen, M. H. (1994). Trustworthiness as a source of competitive advantage. Strategic management journal, 15(S1), 175-190.

- Becker, G. M., DeGroot, M. H., & Marschak, J. (1964). Measuring utility by a single‐ response sequential method. Systems Research and Behavioral Science, 9(3), 226-232. - Bergkvist, L. (2015). Appropriate use of single-item measures is here to stay. Marketing

Letters, 26(3), 245-255.

- Bergkvist, L., & Rossiter, J. R. (2007). The predictive validity of multiple-item versus single-item measures of the same constructs. Journal of marketing research, 44(2), 175-184.

- Bergkvist, L., & Rossiter, J. R. (2009). Tailor-made single-item measures of doubly concrete constructs. International Journal of Advertising, 28(4), 607-621.

- Bhaduri, G., & Ha-Brookshire, J. E. (2011). Do transparent business practices pay? Exploration of transparency and consumer purchase intention. Clothing and Textiles Research Journal, 29(2), 135-149.

- Bloch, P. H. (1981). An exploration into the scaling of consumers' involvement with a product class. ACR North American Advances.

- Bloomfield, R., & O'Hara, M. (1999). Market transparency: Who wins and who loses?. The Review of Financial Studies, 12(1), 5-35.

- Bushman, R., Piotroski, J., & Smith, A. 2004. What determines corporate transparency? Journal of Accounting Research, 2: 207-252.

- Carter, C. R., & Rogers, D. S. (2008). A framework of sustainable supply chain management: Moving toward new theory. International Journal of Physical Distribution & Logistics Management, 38, 360-387.

- Carter, R. E., & Curry, D. J. (2010). Transparent pricing: theory, tests, and implications for marketing practice. Journal of the Academy of Marketing Science, 38(6), 759-774.

- Crowley, A. E., & Hoyer, W. D. (1994). An integrative framework for understanding two-sided persuasion. Journal of Consumer research, 20(4), 561-574.

- Day, G. S. (1970). Buyer attitudes and brand choice behavior. Free Pr.

- Demmers, J. (2014). Ook negatieve informatie is goed voor de verkoop. Tijdschrift voor Marketing, 66-68.

- Eijffinger, S. C., & Geraats, P. M. (2006). How transparent are central banks?. European Journal of Political Economy, 22(1), 1-21.

- Eisend, M. (2006). Two-sided advertising: A meta-analysis. International Journal of Research in Marketing, 23(2), 187-198.

- Feitelberg, R. (2010). Fashion’s new order: Transparency. Womens Wear Daily, 199, 16.

- Fournier, S., & Avery, J. (2011). The uninvited brand. Business horizons, 54(3), 193-207.

- Gantz, J., & Reinsel, D. (2012). The digital universe in 2020: Big data, bigger digital shadows, and biggest growth in the far east. IDC iView: IDC Analyze the future, 2007(2012), 1-16.

- Granados, N., Gupta, A., & Kauffman, R. J. (2005). Transparency strategy in internet-based selling. Advances in the economics of information systems, 80-112.

- Granados, N., Gupta, A., & Kauffman, R. J. (2008). Designing online selling mechanisms: Transparency levels and prices. Decision Support Systems, 45(4), 729-745.

- Granados, N., Gupta, A., & Kauffman, R. J. (2010). Research commentary-information transparency in business-to-consumer markets: Concepts, framework, and research agenda. Information Systems Research, 21(2), 207-226.

- Grewal, D., Iyer, G. R., Krishnan, R., & Sharma, A. (2003). The Internet and the price– value–loyalty chain. Journal of Business Research, 56(5), 391-398.

- Grimmelikhuijsen, S., Porumbescu, G., Hong, B., & Im, T. (2013). The effect of transparency on trust in government: A cross‐national comparative experiment. Public Administration Review, 73(4), 575-586.

- Hayes, A. F. (2013). Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-based.

- Jacquemet, N., Joule, R. V., Luchini, S., & Shogren, J. F. (2011). Do people always pay less than they say? Testbed laboratory experiments with IV and HG values. Journal of Public Economic Theory, 13(5), 857-882.

- Kelley, H. H. (1967). Attribution theory in social psychology. In Nebraska symposium on motivation. University of Nebraska Press.