The effects of monetary incentives and type of argument on the perceived credibility of online consumers

(1)

The effects of monetary incentives and type

of argument on the perceived credibility of

online consumers

Master thesis Msc Marketing

(2)

1

Online customer reviews:

The effects of monetary incentives and type of

argument on the perceived credibility of online

consumers

Author: Marcel Moesker Student nr. 1795864 Eerste Hunzestraat 14 9715 BK Groningen Tel. +31 (0)6 267 111 32 E-mail: moeskerm@hotmail.com University: University of Groningen

Faculty of Economics and Business MSc Marketing

1st supervisor: Dr. J.A. Voerman

2nd supervisor: Dr. M.C. Leliveld

(3)

2

Abstract

The advent of the internet resulted in an increasing interest in a powerful instrument within

marketing: electronic word-of-mouth (e-WOM). Along with the increasing interest among marketing managers, the interest among scholars increased as well. With the result that different aspects have been investigated. This study examines the effects of monetary incentives. This is a construct which is neglected in the studies with regard to online customer reviews up till now. Although, the use of monetary incentives have been examined in terms of other constructs like transmission probability, the effects on the credibility perceived by the receiver has been under shadowed.

Along with the use of monetary incentives and the effects on perceived credibility, are in this research the effects examined of type of arguments. Valence is a popular construct within e-WOM studies, with positive and negative arguments being addressed. However, this study goes beyond those studies by investigating OCRs where both, positive and negative, arguments are combined, with regard to the perceived credibility.

From this study it became clear that monetary incentives can be considered as a useful tool when considering the perceived credibility. There were no significant results for this variable. However, this instrument should be used with caution. It also became clear that OCRs with two sides being

addressed (positive and negative) are perceived more credible than OCRs with one side being addressed (either positive or negative). Also, it appears from this study that heuristic style of processing affects the relation between type of argument and perceived credibility, with someone high on heuristic processing the effects of two sided arguments being even stronger, in terms of perceived credibility. Finally, an additional finding in this study is the direct effect of consumer criticism in terms of perceived credibility.

Keywords: word-of-mouth, online customer review, electronic-word-of-mouth, e-WOM, type of

argument, e-WOM credibility, monetary incentives, heuristic processing.

Acknowledgements: This master thesis is my final stage of the MSc Marketing programme at the

(4)

3

1. Introduction

Word-of-mouth (WOM) is a concept which is considered among marketing managers as a powerful tool to increase sales. In 1995, Amazon.com was the first online platform which brought this instrument into practice in an online environment (Yubo and Jinhong, 2008). Consumers was given the ability to share their opinions and ratings about the product, on this website. With the rise of new channels such as the internet and its related possibilities, more and more marketing activities have shift to these platforms, along with the usage of these channels (Gupta and Harris, 2005). Those platforms, refer to the multiple online platforms, which facilitate the opportunity for consumers and/or potential consumers to share their opinions or comments, such as weblogs, discussion forums, review websites, e-bulletin board systems, news groups, and social networking websites (Cheung and Lee, 2012). Consumers like to share their thoughts about their experiences with products by spreading electronic word-of-mouth (e-WOM). E-WOM communication refers to any positive or negative statement made by potential, actual, or former customers about a product or company, which is made available to a multitude of people and institutions via the Internet (Hennig-Thurau et al., 2004).

1.1.1Traditional WOM

The managers of Amazon.com might be considered as the pioneers of e-WOM, being the first in creating a platform for consumers to convey their opinions (Yubo and Jinhong, 2008). The managers of Amazon.com seemed to have considered the several benefits of WOM and turned this into

practice, online. This section will discuss traditional WOM since traditional WOM is similar to e-WOM to a certain extent. Moreover, traditional WOM forms the base for multiple studies and theories which will be set out in more detail in the next chapter.

WOM is within the context of marketing often defined as informal communications directed at other consumers about ownership, usage, or characteristics of particular goods and services and/or their sellers (Westbrook, 1987). The first studies covering the phenomenon of WOM can be found in the mid-1950s (Whyte, 1954; Katz and Lazarsfeld, 1955). Ever since, many studies followed about the effects of WOM. In most cases it can be considered as more effective than traditional marketing tools like advertising or sales (Katz en Lazarfeld 1955; Engel et al. 1969). It appeared to be a powerful tool for business due to its influence on several outcomes, such as awareness, perceptions and behaviour patterns (Buttle, 1998). Even if the information spread by the customer has negative characteristics, it is still very likely that the process of WOM will have a positive outcome: any form of publicity will have positive effects (Berger et al, 2010). According to Duan et al (2008) this has to do with creating awareness, by talking about the product it conveys its existence.

1.1.2. Differences between WOM and e-WOM

Although, e-WOM might show some resemblance to traditional WOM, there are also considerable differences between traditional WOM and e-WOM. First, traditional WOM is exchanged in private conversations to everyone who is present at that particular place and time, while with e-WOM it appears that the exchange of information can be on discontinuous moments (Hung and Li, 2007). As such, the receiver of the information doesn’t have to be present at the time the message is

(6)

5

longer and comes in larger quantities (Chatterjee, 2001). It allows managers to better analyse the e-WOM communications, compared to traditional e-WOM (Lee et al, 2008; Park & Kim, 2008). However, whereas with traditional WOM it is an obstacle to even obtain data, with e-WOM there is often an overload of data. This has the effect of managers facing the challenge to get the proper data (Cheung & Lee, 2012). In addition, for the consumer does this increase in creating and sharing information via the internet results in a significant increase in consumers’ search costs (Chen, Wu and Yoon, 2004; Tapscott and Williams, 2008). However, for the consumer it is more convenient to retrieve

information for e-WOM communications since traditional WOM is in oral form which complicates the retrieval of information. Finally, traditional WOM can be considered more frequently as a social process of people with strong ties. People know one another because they are good friends or family, and for that reason they feel concern for the other person’s welfare when it comes to communal relationships (Clark, 1984). In contradiction, with e-WOM the linkages between people can be considered weak, as of its anonymous setting on online platforms. Therefore, people in e-WOM settings feel less responsible for the other person’s welfare (Clark, 1984; Ryu and Feick, 2007). As a result of the listed differences in this section, managers have to approach e-WOM in a different way. Especially, knowing that as more e-WOM is being generated online, a significant amount of power and influence over enterprises has shifted to consumers (Chen, Wu and Yoon, 2004; Tapscott and Williams, 2008).

1.1.3 Monetary incentives

By discussing traditional WOM and e-WOM it can be concluded that these are useful and powerful tools. Therefore, consumers and companies use this instrument more and more; companies to gain customers, customers to obtain information (Gupta & Harris, 2005). This increase is mainly driven by the target ability and cost effectiveness of this tool compared to traditional promotional tools (Mummert, 2000). However, not every consumer is motivated to step in the process of word-of-mouth (Dichter, 1966). Especially with the multiple channels available. To raise the quality and the volume of consumer contributions, a strategy often used is to offer an incentive (Hsieh and Kraut, 2010). Every company can incorporate an online customer review system which enables the

customer to give their valuable opinion. Before the consumer is willing to provide his or her review of the product, there has to be a certain incentive to do so (Guo et al, 2009). As such, Guo et al (2009) distinguishes six different types of incentives with their respective motives, which are derived from several studies:

- Social relations: a sense of belonging and recognition; Socialization needs.

- Altruistic factors: suggestions for fellow consumers; suggestions for companies and brands - Self enhancement: self-assertion, and enhancement of self-esteem; indication of social

status; the acquisition reputation

- Expectations of returns: exchange of information; mutual benefits; acquisition of more resources and power.

- Self-interests: love for the products; satisfaction in sharing; way of entertainment - Economic incentives: discounts and material incentives; monetary incentives

(7)

6

difficult to change since they are embedded in the beliefs of the consumer (Dichter, 1966). Also, it is an instrument which is manageable and it enables management to monitor the costs of a campaign. Also the effects are easy to be observed. Due to the listed reasons of selecting this type of incentive, can this instrument be considered as the most popular one(Ryu and Feick, 2007).. As such, the focus of this research will be on monetary incentives. Not only because it is a popular instrument and therefore in the interest of many companies, moreover, because it enables this research to come near to a real life setting. Also, it enables to manipulate the incentive variable and monitor the results.

1.1.4 Type of argument

In initial efforts to stimulate the consumer with incentives to generate customer opinions, the focus was mainly on satisfying the customer. This was done hoping that the consumer would generate positive rather than negative information (Rosen, 2000). For the same reason, influential consumers, like opinion leaders, were targeted hoping they would generate positive customer reviews. However, more and more customers started questioning the amount of positively grounded messages,

especially in terms of the credibility (Ryu and Feick, 2007). 1.1.5 Perceived credibility

E-WOM credibility is defined as the extent to which one perceives the recommendation as

believable, true, or factual (Tseng and Fogg, 1999). The credibility of an online customer review (OCR) has been under increasing attention since the advent of e-WOM studies. This is because many consumers use this as a criterion when evaluating an OCR. The consumer review will be part of the decision process. Obviously, the consumer want to make the right decision and needs credible input, as such. The way the receiver evaluates the message can be for personal reasons. However, it appears to be affected by the level of sidedness of the message: whether positive and/or negative arguments are provided. In addition, messages are evaluated on valence, whether the stimulus is positive, neutral or negative (Cheung, 2009). These will be discussed in more detail in the next chapter.

Credibility of e-WOM is not only a criterion used by consumers to evaluate OCRs. Different studies noticed that credibility affects other concepts, as well. E-WOM credibility is proven to be positively related to e-WOM adaption (Cheung et al, 2009). The e-WOM adaptation, on its turn, is proven to have a positive relationship with the purchase intention which can be considered as a good indicator of the actual purchase (Fishbein and Ajzen, 1980; Ajzen, 1991). As such, credibility of e-WOM is an important concept and certifies other important concepts.

In this study the perceived credibility is split up into source credibility and content credibility. This is done because two concepts will be central in this research, namely the type of argument used and the use of a monetary incentive. Type of argument is believed to be mainly related to the content credibility since it says something about the content. Source credibility is more related to the use of monetary incentives since the writer of the review is the one who receives the incentive, eventually.

1.2 Problem statement

(8)

7

difficult to be affected by a company. A type of extrinsic incentive often used by companies to attract consumers in the process of writing a review is an economic (monetary) reward. One of the

implications of using incentives to stimulate online customer reviews (OCRs), is that the spontaneity of the OCR decreases, more importantly, it might affect the credibility of the OCR. There are many studies conducted concerning e-WOM credibility. However, the use and effects of incentives are neglected in those studies. Although, e-WOM has been around since 1995, it really emerged during the last decade (Gupta and Harris, 2005). That is probably the reason why many studies focused on the outcome which is the most relevant to companies: does it result in more generated content by consumers and how can this tool be used effectively. What the effects are on the receiving end, has been neglected to the information of the researcher. However, in traditional word-of-mouth studies it can be seen that paying for referrals have a negative effect on sender credibility (Helm, 2009). As opposed to that study, will this research take place in an e-WOM setting. As can be seen in section 1.1.2. there are considerable difference between traditional WOM and e-WOM. One of these differences is the difference in tie strength which is normally weaker in e-WOM settings. This can be considered as one of the reasons why these effects should be re-examined. It will be investigated if there is an effect on the credibility perceived by the receiver, when economic incentives are used. Furthermore, in many studies the valence of the message is considered. However, those studies mainly focus on just positive or just negative valenced messages, while in online customer reviews (OCRs) there is the possibility that positive and negative arguments are combined. The effects of the combined arguments will be investigated in this research, as well. Finally, it is expected that some of the characteristics of the consumers might influence the possible effects. For this study the selected moderators are style of processing, proneness for monetary rewards and consumer scepticism towards OCRs which will be discussed in more detail in the next chapter. Subsequently, the following research question is composed:

What are the effects of type of argument, and the use of monetary incentives, on the perceived credibility of the receiver, and to what extent are these relations moderated by style of processing,

proneness for monetary rewards, and consumer scepticism towards OCRs?

1.3 Research questions

To answer the main question, the following research questions are derived and will be investigated in this research:

RQ1: What type of argument has a stronger effect on the perceived credibility?

RQ2: What are the effects of the use of monetary incentives on the perceived credibility? RQ3: What are the moderating effects of the use of monetary incentives in relation with type of argument and perceived credibility?

RQ4: What are the moderating effects of the receivers’ style of processing on the way the receiver evaluates the message in terms of perceived credibility?

RQ5: How does the moderator “consumer scepticism towards OCRs”, relate to perceived credibility? RQ6: How does the moderator “proneness for monetary rewards”, relate to perceived credibility?

1.4 Social communication

(9)

8

brand, a product, or a service offered for sale. This definition also implies for electronic word-of-mouth, with the difference that it a)isn’t oral and b) that the communication is via an internet platform, and c) from one-to-many persons. Considering these differences, the elements from social communication still holds. As such, in describing the social processes during this report, elements will be named which are derived from Hovland (1948). The following elements can be distinguished in social communications:

- The communicator (source): the person who starts and conveys the communication message, which is in the case of e-WOM the online customer review.

- The stimulus (content): refers to the message, or review, transmitted by the communicator. - The receiver (audience): is the individual who conceives/observes/interpret the

communication message.

- The response (main effect): is made to the communicator by the receiver.

These terms will be used when discussing the different concepts in this research. Except stimulus will be replaced by using OCR: online customer review.

1.5 Academical relevance

This article will be an addition to the written literature about e-WOM. E-WOM has been under increasing attention over the last few years, with the focus on several effects of e-WOM. However, the use of incentives, especially monetary incentives, have been neglected in most literature about e-WOM, while it can be considered as a significant tool to help increasing the word-of-mouth. Even in the traditional literature about WOM little attention is paid on the use of incentives and the effects on credibility. Furthermore, valence of OCRs is a frequently studied construct within the literature of e-WOM. However, not much attention has been given to reviews were both type of arguments are combined.

1.6 Managerial relevance

The increasing attention of eWOM comes along with opportunities and threats for managers as well (Strauss, 2002). As described earlier, the use of incentives is a good tool to stimulate eWOM but it is important to know in what way it will affect the customer reviews. If a review is perceived as credible, it is proven to be a good predictor of the purchase intentions.

1.7 Thesis structure

(10)

9

Chapter 2 Theoretical framework

2.1 Introduction to the chapter

In this chapter different concepts and relations are discussed. First, the independent variables will be discussed and the possible relation towards the dependent variable, perceived e-WOM credibility. The discussed connection between the independent and dependent variable is captured in a hypothesis. Subsequently, the covariates will be discussed. This chapter will be concluded with a conceptual model and an overview of the hypotheses.

2.2 Type of argument

The communicator can have different goals to participate in the process of e-WOM. This person can participate in order to pick up an incentive (Ryu and Feick, 2007), want to share its expertise and inform others, or want to make a recommendation. Either way, when communicating the message different arguments are used to convince the receiver. In this chapter the focus will be on the type of argument. The main issue central to this discussion is the level of viewpoints or sidedness of the message. There are OCRs where mainly one point of view is addressed, either positive grounded or negative grounded. This has been investigated in multiple studies where valence of the message has been addressed(Halstead, 2002; Harrison-Walker, 2001; Heitmann et al., 2007). In contrast, there are OCRs where two sides are addressed: positive and negative. This has been underexposed in the literature of e-WOM. The next section commence with explaining the differences between one sided OCRs, in terms of perceived credibility. When studying the types of sidedness, it should be

mentioned it is about one OCR on itself, not in comparison with other OCRs.

2.2.1. One sided OCRs: Positive word-of-mouth and its effects

Positive WOM is described as encouraging the choice for a particular brand (Sundaram, Mitra and Webster, 1998; East, Hammond and Lomax, 2008). Positive e-WOM has many persuasive effects, with all the positive points of the product being addressed (Dellarocas et al. 2007; Duan et al, 2008). Positive WOM is strongly correlated with cognitive evaluations. These cognitive evaluations are often considered before one spreads the positive WOM, which implies that someone has well thought about the content of the message, and therefore, can be considered credible (Sweeney et al., 2005). In a study from Sun-Jae and Jang-Sun (2009) the participants were exposed to positive and negative eWOM messages. The positive sets of stimuli showed higher scores than the negative messages in most situations, like attitude towards the product and website, and purchase intention. However, in terms of credibility this was not the case. Negative messages scored higher. Positive messages don’t have a negative effect on credibility per se. However, when there are only positive messages presented, it will harm the credibility in the long run (Sun-Jae and Jang-Sun, 2009).

2.2.2. One sided OCRs: Negative word-of mouth and its effects

Negative WOM is described as the discouragement of the choice for a particular brand (East,

(11)

10

and mental stress experienced by an individual who holds two or more contradictory beliefs, ideas, and/or values at the same time (Festinger, 1957). Other reasons are: to let the emotions out, to warn others or seeking revenge, which are issues which could harm the credibility of the message

(Sweeney et al., 2005; Wangenheim, 2005)

Like mentioned in the previous paragraph, do negative messages have greater impact on credibility (Sun-Jae and Jang-Sun, 2009). This can be because it is well established that negative information usually has more impact on judgment than positive information (Skowronski & Carlston, 1989). People like to address the negative side in things, which is considered to be something human (Baumeister et al., 2001). Fiske (1980) observed that there is usually more positive information than negative information. Therefore, negative information is perceived as more useful and credible. For example, when a negatively framed message is communicated about the reliability of the product, it is perceived as more credible, because a reliable product is assumed to be the default condition for established products. Laczniak, DeCarlo and Ramaswami (2001) found that people in some cases even reacted against all advice and got even more involved with a brand which was subject of negative WOM. This odd reaction can be explained by the reactance theory by Brehm (1966). According this theory someone will be aroused when his/her opinions are threatened. As a reaction will someone be more determined and won’t deviate from his opinion.

Combining section 2.2.1 and 2.2.2 it is expected that in a situation with an OCR, where one point of view is expressed, a negative message is perceived more credible than a positive OCR. Therefore, the following hypothesis can be formed:

H1a: A negative OCR is perceived more credible than a positive OCR. 2.2.3 Message sidedness: positive and negative valenced OCRs

The concept of message sidedness is somewhat related to the previous discussed concept, in the sense that it also has to do with valence of the message. However, instead of whether the focus is on a positively or negatively valenced message, is it also possible that both points of view are addressed in the stimulus. With one-sided messages, a positive or negative point is addressed, while with two-sided messages both perspectives are being presented (Cheung, 2012). In traditional marketing studies it has already been proven that campaigns with both sided arguments being presented tend to be more thrustworthy (Hovland et al, 1953; Lumsdaine & Janis, 1953). Also in other marketing studies, it is suggested that two-sided information makes the reference frame of the consumer more complete, and is therefore perceived as more credible (Kamins & Assael, 1987).

In advertisements in traditional marketing campaigns it is normally the case that the communication is one-sided and positively framed (Knowles and Linn, 2004). On the contrary, communications from the consumer are also one-sided with the exception that the communication is frequently negatively framed. Two-sided messages are more effective when the product is unfamiliar or unfavourable to the consumer (Kamins & Assael, 1987). Fiske's (1980) explanation may be expressed in terms of the gap between the position implied by the message and the position held by the receiver. Information that restates what the receiver believes may increase certainty but is unlikely to change other aspects of a receiver's judgment. In contrast, information that differs from the receiver's position may change their judgment.

(12)

11

presented, it will affect the credibility of e-WOM messages negatively (Sun-Jae and Jang-Sun, 2009). Positive information results in more positive outcomes in terms of attitudes and intentions, but the presence of negative information in the reviews enhance the perceived credibility of an e-WOM message (Doh & Hwang, 2009).

In summary, it can be expected that the credibility of the message will be increased when both positive and negative arguments are provided. However, this appears to be one of the few reasons why consumers prefer mixed eWOM. Mixed reviews have limited value to the consumer in terms of making purchase decisions or changing attitudes (Edwards and Edwards, 2012; Hartman et al., 2012). Based on the previous studies, the following hypothesis can be formed:

H1b: A two sided OCR is perceived more credible than an one sided OCR.

2.3 Monetary incentives for writing an OCR

One of the core elements of this master thesis is the usage of incentives in relation with OCRs. The scarce amount of studies which paid attention to the effects of the usage of incentives were directed to the willingness to spread e-WOM due to the use of incentives (Guo et al, 2009). However, the effects of economic incentives on other aspects have been neglected in most e-WOM studies. Due to the limited attention to this subject, it is hard to make an assumption in terms of what the effect will be on the perceived credibility. Therefore, traditional literature about WOM has been consulted. In addition to that, behavioural aspects related to incentives will be considered to make an assumption about the possible effects of incentives on the perceived credibility. The use of incentives might not affect the credibility at all, it is how the receiver of the message perceives it. This matter is also been taken into account in this section.

Incentives and perceived credibility in traditional WOM

Considering traditional WOM it can be claimed that paying for referrals have a negative effect on the perceived credibility by the recipient (Helm, 2009). However, traditional WOM is significantly

different from eWOM. In traditional WOM several aspects are involved than just the message and the fact that an incentive has been given. It is beyond the scope of this research but issues like tie strength plays a huge role in traditional WOM and its related credibility due to its live setting. From the study of Helm (2009) it became clear that paying for the referral has a negative effect on the perceived credibility no matter how strong the relationship is. However, strong ties still can be seen as more credible. This information can help us a bit more in predicting the relationship in this research between incentives and perceived credibility because in this research, and in most

situations regarding eWOM there is a weak link. Weak ties are more responsive to referral incentives, but at the same time are less influential (Helm, 2009).

(13)

12

Social exchange theory

As it can be seen in this thesis sharing e-WOM is a social process between two or in most cases multiple individuals. The person who is providing the (e-)WOM, the communicator, has to suffer certain costs (Gatignon and Robertsen, 1986). These costs can be the effort and the time spent in communicating the message. The theory which goes more in depth on this matter is the social exchange theory. The decision to engage in the process depends for a great part on the subjective costs and benefits of the exchange and the comparison with alternatives (Frenzen and Nakamoto, 1993). However, in the case of e-WOM the social process is less visible and the relationship between the communicator and receiver can be considered as a weak tie. With weak ties the exchange relationship is mainly driven by self-interest. The communicator does not feel any responsibility for the receiver of the message. The goal of the communicator in these weak ties is to maximize their own outcome and minimize the costs made. The main objective is to achieve an equilibrium by increasing what they receive and reduce what they give (Walster, Berscheid and Walster, 1973; Walster, Walster, and Berscheid, 1978). Therefore, it is expected that in the case of incentives the communicators only seek for opportunities to increase their own benefits and reduce as much the cost by putting effort in a reliable customer review (Ryu & Feick, 2007)

Equity theory

Another theory which comes close to the theory discussed in the previous section is the equity theory. The equity theory entails that people try to maintain equity towards each other in terms of input they put in a certain job compared to the outcomes they receive. The equity theory is applied in many studies concerning employer- employee relationships. The ratio between inputs and outputs should be in an equilibrium in order to stay motivated and maintain fairness. The employer pays wages and requires a certain output from the employee. The employee will put in the effort which (s)he perceives to be equal to the received payment (Adams, 1965). Thus, in the case of an incentive being provided yields inequity due to the increase in input for the communicator. It is expected the communicator feels obliged towards the company to bring back the equity by providing a certain output in favour for the company (Walster, Walster, and Berscheid, 1978). It is expected that the communicator is more influenced by the economic incentive and being in equity with the company, than being concerned about the psychological or social risk of not telling the truth to the receiver of the message about the product (Frenzen and Nakamoto, 1993; Tӧrnblom and Nilsson, 1993) Although, the theory which is outlined in the previous section in this chapter is mainly applicable in situations of traditional WOM, it is expected that it might be appropriate in situations concerning eWOM. Combining the different theories it is expected that the use of incentives have a negative effect on the perceived credibility. Therefore, the following hypothesis can be formed:

H2a: The use of an economic incentive negatively affects the credibility of OCRs perceived by the receiver.

2.4.2 Moderating role of incentives

(14)

13

the receiver of the communication will perceive the message as credible when the receiver attributes the communicator’s review to product related reasons. However, when the receiver of the message attribute the review to non-product related reasons, they suspect underlying motives of writing such a review, like in this study, receiving incentives, and perceive it as less credible (Sen and Lerman, 2007; Lee and Youn, 2009). This theory can be complemented with the expectancy theory. The expectancy theory proposes that an individual will decide to behave in a certain way because they are motivated to select a specific behavior over other behaviors due to what they expect the result of that selected behavior will be (Oliver, 1974). Combining these possible theories with the theory which is conveyed in the rest of this section concerning incentives, the following proposition can be made: it is expected that the receiver of the message might think that the communicator is in debt, due to the incentive, with the company who provided the incentive. Therefore the communicator will provide a positive, one sided message because this ‘behavior’ is expected to be desired by the company. Due this increase in the possibility of a positive message, the perceived credibility will decrease, as opposed to no monetary incentive involved.

The following hypothesis can be formed:

H2b: When monetary incentives are used it is expected to moderate negatively the relation between type of argument and perceived credibility.

2.4 Monetary reward proneness

With incentives being one of the main issues being addressed, attention will be given to the sensitivity of individuals with regard to monetary rewards. People might act differently when a monetary incentive is involved. In this section the sensitivity towards monetary rewards will be discussed, by first, explaining the underlying theory why people react on rewards, and second, discussing the possible behaviour on monetary rewards.

Reinforcement sensitivity theory

One of the theories behind the way people react on rewards is the reinforcement sensitivity theory (Gray, 1970). In the brain, certain neural structures can be found which underlie individual

differences in sensitivity to rewards, punishment and motivation (Corr, 2008). Within the behavioural approach system (BAS), behaviours can be found that are motivated to seek rewards. When this systems is activated, individuals crave excitement, show remarkable persistence, and they feel very delighted when getting rewards (Pickering & Gray, 1999).

The BAS can be considered as the ‘reward system’ within the brain-behaviour structures. This system has a mediating role and reinforces the way people behave. Rewards tend to act as a reinforcer. A reinforcer can be considered as something that, when served after a behaviour, causes the

probability of that behaviour’s occurrence to increase. However, it should be noted that just because something is labelled as a reward, it does not necessarily imply it is a reinforce. It can only be

considered as an reinforcer if it increases the probability of a behaviour. For some people a monetary reward can affect their behaviour while for others it will not change the outcome for certain actions (Corr, 2008).

The behaviour and psychology of incentives

(15)

14

It seems obvious that paying someone is the way to get someone doing a task. However, there are situations in which paying someone can backfire. For example when someone is already intrinsically motivated to do the task, the extrinsic motivation will crowd out the interest in the task (Deci, 1971; Deci et al. 1999). Also, when the task can be seen as a noble task, the introduction of monetary incentives can backfire. Studies on this matter investigated pro-social behaviour like blood

donations. It seems that offering payments for the donations resulted in a reduction of people willing to give blood (Titmuss, 1970; Mellstrom and Johannesson, 2008).

The size of the payment can also affect the sensitivity to monetary rewards. Paying the participant too much for successfully completing a task can result that this person gets too nervous to fulfil the task. This is called choking (Baumeister, 1984). The high incentives schemes will cause that the performance might decrease. It could be that some people choke due to the pressure which comes along with the high incentive schemes, however, paid professionals in sports this seems to be less a concern (Paserman, 2010). On the other hand, paying too little can backfire as well. Paying a very low wage might be counterproductive opposed to paying nothing (Gneezy and Rustichini, 2000). The main interpretation of this finding is simply that people find it degrading to work for a small amount of money (Ariely et al., 2008).

Moderating role of reward proneness

The sensitivity to monetary rewards is in this study not considered as an independent variable but is expected to have a moderating role. It is expected that people might take their own perceptions into consideration when making judgements in terms of the perceived credibility. The theory which underpins this hypotheses is the self-perception theory. This theory explains that people develop their attitudes by observing their own behaviour and conclude what attitudes must have caused it (Bem, 1967). The people attempts to explain others’ behaviours in the same way they interprets their own behaviours rationally (Robak et al. 2005).

If someone is more sensitive for monetary rewards, it is expected that this person knows better how it changes their behaviour and therefore negatively affects the relationship between incentives and perceived credibility (Kamenica, 2012). Therefore, the following hypotheses can be formed

H3: reward proneness strengthens the negative relationship between incentives and perceived credibility

2.6 Consumer criticism regarding OCRs

Consumer criticisms or scepticism is a construct which has been investigated in several studies concerning scepticism towards advertising (Calfee and Ringold, 1994). For marketing managers is advertising one of the main instruments in their strategies. The Internet as a platform for sharing opinions on products has become an important marketing instrument to compete for consumer attention and visits (Chatterjee, 2001). It is also considered as in instrument which can be managed (Rosen, 2000). Consumers thought to be found a trustworthy source of information for their purchase decisions. Previous studies found that consumers perceive WOM more persuasive than traditional media, such as advertising on tv, radio, print ads or personal selling(Arndt, 1967; Herr et al., 1991; King and Summers, 1970)

(16)

15

among consumers. Obermiller and Spangenberg (1998) once presented a paradox whereby they suggested that the characteristic of a free market encourages exaggerations in marketing messages which subsequently results in consumer scepticism. The fact that e-WOM induces easy access to information sources from different people causes people to be more sceptical, even if the messages are positive (Sher and Lee, 2009).

Another fact with the use of e-WOM is that the source of the information, the communicator, is often a stranger. This is a person with whom the consumer has a weak tie and whom the consumer will never meet. This raises doubt on the trustworthiness of the reviews (Ryu and Feick, 2007). The fact that online communicators even get paid for referrals or receive a portion of the turnover casts even more doubts (Chatterjee, 2001).

It is clear that there is some scepticism among consumers with regard to OCRs. However, it is not exactly clear what effect it will have. Sher and Lee (2009) found out that highly sceptical consumers base their opinions, believes and attitudes towards the review on intrinsic believes instead of situational factors. On the other hand, consumers with low scepticism look more at heuristics when forming an attitude, so they look at the quantity of the reviews and, in the case of this study, if there is an incentive involved for the communicator.

H4: OCR scepticism negatively interacts the relation between the use of incentives and perceived credibility.

2.7 Style of processing

Every individual is different in the way we behave and act, as well in the way one processes information. The style of processing has been investigated in multiple marketing and consumer behaviour studies. Consumers who have been faced with various situations, have been proven to have different processing capabilities, motivations and prior experiences (Bettman, 1979). When forming judgements, different studies found that individuals differ significantly in three areas. Namely, individuals differ in the way they acquire information, have different strategies during the acquisition of the information and finally, they differ on the way they utilize the information when its acquired (Henry, 1980). Different models about style of processing can be found in literature like the Elaboration Likelihood Model (ELM), but the focus in this study will be on the Heuristic-Systematic Model of information processing (HSM), which share many of the concepts developed by the ELM. The HSM is a model by Chaiken (1980) which attempts to explain how people process persuasive messages like advertisements. According to the model do individuals process messages in a heuristic way or they process it systematically. The HSM model was conceived as a persuasion model but has involved into a general model of information processing (Todorov et al, 1997)

Heuristic processing

(17)

16

Accessibility refers to the ability to retrieve the memory and use it. Applicability, refers to the relevance of the memory in relation to the given situation (Chen et al. 1999).

Marketers are aware of this phenomenon of heuristic processing and exploit these mental habits by priming them in advertisements. Heuristic provide shortcuts in the mind, which can result in fast conclusions. An example of an heuristic is that in a message a lot of arguments are presented which might imply that the author of the message is well informed and convincing, without even reading the full content of the message (Eagly and Chaiken, 1993).

H5a: Heuristic processing affects the relationship between the use of incentives and perceived credibility.

H5b: Heuristic processing affects the relationship between the type of argument and perceived credibility.

Systematic processing

Opposing to heuristic processing we think logically when encountering a message. In this case we process the information systematically. This style of processing is more analytic and involves more cognitive effort. The judgements resulting from systematic processing are derived on in-depth treatment of relevant information. With systematic processing does personal relevance play an important role, so when determining the validity of the message the reliability of the source and the content plays an important role in order to be persuaded (Chaiken, 1980). The recipient develops attitudes towards the message after (s)he spends considerable cognitive effort when reading the message and actively attempts to comprehend the message and evaluate the arguments. H5c: Systematic processing affects the relationship between the use of incentives and perceived credibility

H5d: Systematic processing affects the relationship between the type of argument and perceived credibility.

2.8 Conceptual model

Combining the expected relations in the previous chapters, the following conceptual model can be presented, which can be seen in figure 2.1.

(18)

17

Fig. 2.1: Conceptual model

2.9 overview of hypotheses

H1a: A negative OCR is perceived more credible than a positive OCR. H1b: A two sided OCR is perceived more credible than an one sided OCR.

H2a: The use of an economic incentive negatively affects the credibility of OCRs perceived by the receiver.

H2b: When monetary incentives are used it is expected to moderate negatively the relation between type of argument and perceived credibility.

H3: reward proneness strengthens the negative relationship between incentives and perceived credibility

H4: OCR scepticism negatively interacts the relation between the use of incentives and perceived credibility.

H5a: Heuristic processing affects the relationship between the use of incentives and perceived credibility.

H5b: Heuristic processing affects the relationship between the type of argument and perceived credibility.

H5c: Systematic processing affects the relationship between the use of incentives and perceived credibility

(19)

18

3. Methodology

In this section the setup of the research will be discussed. The operationalization of the different concepts will be discussed before the survey was conducted. However, some preliminary results and outcomes of pre-tests will be presented in this chapter, as well. This chapter starts with the design of the experiment and a description of the participants. Thereafter, the procedure will be discussed, followed with the operationalization of the independent variables. Subsequently, the dependent variables are discussed. Finally, before some preliminary tests are conducted, a plan of analysis will be addressed.

3.1 Participants and design

In order to gain insights in the perceived credibility a quantitative research in the form of an experiment was conducted. Two independent variables were the centre of attention in this study: use of monetary incentive and type of argument. Monetary incentive was measured on two levels: either there is a monetary incentive involved or not. Type of argument was measured on three levels: the type of arguments used in the OCR were either positive, negative or positive ánd negative. As such, this experiment can be considered as a 2 x 3 between subjects experimental design. This type of experiment was chosen to avoid any carry-over effects between subjects (Malhotra, 2010). The respondent was assigned to one of the conditions listed in table 3.1 and due to this type of research it will be assured that the respondent have limited information about the goal of the research which might bias the outcome (Maholtra, 2010)

Table 3.1 is presented to provide an overview of the different conditions and the respective amount of respondents eventually assigned to each condition.

No incentive Incentive

One sided, positive 28 respondents 28 respondents

One sided, negative 29 respondents 29 respondents

Two sided, positive and negative

29 respondents 30 respondents

Table 3.1 overview of conditions

The population in this research was composed of all individuals of interest to the researcher. The target population refers to all the possible participants who are expected to have the desired information an capabilities needed for this research (Malhotra and Birks, 2007). The target population were Dutch male and female individuals who are active online. The selection for this composition is because it was expected this group is familiar with buying products online and in that case might have used customer reviews in their buying decision.

In total, 185 people participated in the study by filling in the questionnaire. However, twelve

(20)

19 Variable Results Gender Age Mean: 33,21 std. dev.: 14,671 N = 173 Education level N=173

(21)

20

People were contacted via e-mail or social media with the question to fill in the questionnaire. The questionnaire was built with the help of Qualtrics. This programme enables the user to make a single questionnaire with questions applicable to every respondent, but more importantly, it enables to provide different manipulations to different respondents. It randomly assigns the different conditions to different respondents. Eventually, there was one single link, but the respondent got randomly assigned to a certain condition (which can be incentive/no incentive and the arguments in the OCR can be positive, negative or two sided (positive and negative)). The programme keeps count of how many respondents were distributed to every condition and it made sure that the amount of respondents was evenly distributed.

Due to the fact that in this study incentives have a central role, like proneness for monetary rewards, there have no incentives been rewarded to the participant for filling in the questionnaire. Providing a reward might possibly bias the outcomes.

Design of survey

The different conditions were left similar to each other as much as possible. In this way the bias concerning other external factors will be left to a minimum. A new web template was designed for this study which should resemble a web shop which exploits electronic products. This page has features like a ‘buy now’ button and a shopping cart and an area where consumers can leave there comments/reviews. From these web templates was made a print screen which is used for the

manipulations in the questionnaire. Figure 3.1 displays one of the print screens. The decision to make a new template and not edit an established website is because people might have formed an

(22)

21

Figure 3.1: example of print screen for the condition Monetary incentive/positive arguments.

3.2 Procedure

To approach the participants I chose to select the method which fits best to the experiment. The participants were approached in an online setting via e-mail or social media. This way of approaching people has a few advantages. First, the respondent doesn’t feel urged to fill in the questionnaire and can decide for him/herself when is the best time to fill in the questionnaire. Second, they can take as long as they want (even if they needed to close the questionnaire and open the link after a few hours, Qualtrics remembers where the respondent had left off). Third, the data is organized more properly and it is made sure that every individual is evenly randomised to a certain condition. Fourth and probably the most important one, by approaching the respondent to fill in the questionnaire online, it means the person is placed behind a computer screen which is ideal for displacing the respondent in the situation of actually reading an OCR. It was expected that the person can identify with the situation more easily (Maholtra, 2010).

The steps which the respondent had to take are listed in table 3.3 and will be discussed in this section. After a short welcome page, where the respondent is briefly informed what the

questionnaire is all about, the participant moves to the next page. On this page the respondent will be confronted with the only manipulation of the research. A print screen from a webpage is

(23)

22

respondent is instructed to take a good look at the following print screen en read the accompanying customer review.

After the respondent viewed the print screen, questions were presented which are related to the print screen. These questions are about the content credibility and the source credibility, resulting in a net result for the perceived credibility, being the DV. On the next page questions were asked about the IV’s. The questions presented here were all related to the manipulated condition. All that

followed were questions about the covariates. These questions are derived from literature and will be discussed later on. Each covariate is covered on a new page, starting with style of processing with questions related to systematic processing and questions with regard to heuristic processing. After this page the OCR scepticism is addressed, followed with questions with regard to proneness to monetary rewards. The questionnaire is rounded off with some descriptive questions like age, gender and education level. Finally, the participant is thanked for his/her participation to this research.

Order Element Number

of items

Variable

1. Introduction to the survey  Next page

2. Print screen of webpage shown with OCR

3. Source credibility 5 Dependent variable

4. Content credibility 4 Dependent variable

 Next page

5. e-WOM Valence 1 Manipulation check

(IV)

6. Sidedness 2 Manipulation check

(IV)  Next page

(7.) Seen the incentive (only in conditions with incentives involved)

1 Control question (IV)  Next page

7. (8.) Systematic processing 5 Covariate

8. (9.) Heuristic processing 4 Covariate  Next page

9. (10.) OCR scepticism 9 Covariate  Next page

10. (11.) Proneness for monetary rewards 8 Covariate  Next page

11. (12.) Age/gender/education level 3 Descriptive variables  Next page

12. (13.) Respondent was thanked for the contribution to the research

(24)

23

3.3 Operationalization of variables

The different variables need to be operationalized in order to measure it. Table 3.4 presents an overview of the different variables and the items used to measure it. Also, the internal reliability is presented for each item scale, by using Cronbach’s Alpha.

Concepts / variables Items Cronbach’s Alpha Scale Source Dependent variables Source credibility

Trust: un-trustworthy - trustworthy Credibility: not credible - credible Honest: dishonest - honest Unquestionable: unquestionable -

questionable

Conclusive: inconclusive – conclusive

0,890 7-point bipolar rating scale. Uday et al. (2006), cheung et al. (2009) and Ohanian (1991) Content credibility

Believable: unbelievable - believable Authentic: not authentic – authentic Reasonable: unreasonable - reasonable Convincing: not convincing – convincing

0,890 7-point bipolar rating scale. Uday et al. (2006), cheung et al. (2009) and Ohanian (1991) Independent variables (Man. Check) Type of argument: Valance

very negative – very positive -7-point Likert scale: Derived from Hartman et al., 2013 (Man. Check) Type of argument: sidedness Two items:

I think that this review presents primarily (a) ____ argument about the tablet”, and (b) “I think that the reviewer of this product _______ refute or addresses two points of view).

, ranging from 1 (one sided) to 7 (two sided) for

question (a), and from 1

(ineffectively) to 7 (effectively) for question (b). 0,870 Two 7-point bipolar rating scales were used Derived from Kim (2006) (Man. Check: ) Incentives

A control question was asked in the condition with monetary incentive involved if they have noticed that the communicator received an incentive.

Covariates

(25)

24

style of processing

about in a review, I am likely to stop and think about it.

2. If I need to make a purchase decision after reading a review, the more reviews I read, the better. 3. After thinking about the

information in the review, I have a broader understanding.

4. When I encounter information in a review, I read or listen to most of it, even though I may not agree with its perspective.

5. It is important for me to interpret information in a review in a way that applies directly to my life.

scale; strongly disagree – strongly agree (1980;1989), Griffin et al. (2002) Heuristic style of processing

1. When I encounter information in a review, I focus on only a few key points.

2. There is far more information in reviews than I personally need. 3. When I read a review, I rarely spend much time thinking about it. 4. If I need to make a purchase decision after reading a review, the advice of one expert is enough for me. 0,618 7-point Likert scale; strongly disagree – strongly agree Chaiken (1980;1989), Griffin et al. (2002) Consumer scepticismwi th regard to OCRs

1.We can depend on getting the truth in most

OCRs.

2. OCRs aim is to inform the consumer.

3. I believe OCRs are informative. 4. OCRs are generally truthful. 5. an OCR is a reliable source of information

about the quality and performance of products.

6. OCRs are truth well told. 7. In general, OCRs presents a true picture

of the product being reviewed 8. I feel I've been accurately informed after

viewing most OCRs

9. Most OCRs provides consumers with essential information. 0,927 7-point Likert scale; strongly disagree – strongly agree Derived from Obermiller & Spangenberg (1998) Proneness for monetary

1. Receiving incentives makes me feel good.

0,882 7-point Likert scale; strongly

(26)

25

incentives 2. I enjoy searching for options to receive an incentive

3. When I receive an incentive, I feel that I’m getting a good deal

4. I enjoy receiving incentives, regardless of the amount of money I receive by doing so

5. I am loyal to the person/company I receive an incentive from

6. I am more likely to buy brands for which I receive an incentive from 7. Incentives have caused me to do things I normally would do not. 8. Beyond the money I get,

receiving incentives give me a sense of joy

disagree – strongly agree

et al., 1990

Table 3.4: Operationalization overview

3.3.1 Operationalization of IV: Monetary incentive

The independent variable (IV) ‘monetary incentive’ is manipulated for the different conditions. The respondent might appointed a situation where the author of the review can receive a monetary incentive for giving their opinion in a review, or the respondent might come across the condition where no incentive is reserved for the author. Instead a neutral text is provided. In the situation with the incentive the following text is published on top of the webpage: “your opinion is worth money! Earn money by placing a review. 1 euro per review”. In this way it is subtle communicated to the reader of the OCR that the communicator is rewarded for his/her efforts. In the situation of no incentive there is a neutral text placed on the same spot where the incentive text is placed. The only difference is that the text reads: “you are only a few clicks away from the best shopping experience” the rest of the webpage layout remained the same. An overview of the different customer reviews and the manipulation concerning the monetary incentive can be found in table 3.5.

Type of argument

Monetary reward

OCR

Negative Yes Text on top of webpage

OCR I have bought this tablet a week ago en I have spent several hours on it already. There are two aspects which got my attention during the first times using this tablet. These aspects help me to form my first impression about the product which I would like to share with you.

The operating system of this tablet is totally not friendly. I experienced only bad things about the operating system during the short time I worked with it.

(27)

26

Negative No Text on top of webpage

The operating system of this tablet is totally not friendly. I experienced only bad things about the operating system during the short time I worked with it.

I’ve worked with this tablet for several hours but I had to recharge the battery for several times, I consider the battery life time to be very bad.

Positive Yes Text on top of webpage

The operating system of this tablet is very user friendly. I

experienced only good things about the operating system during the short time I worked with it.

I’ve worked with this tablet for several hours without recharging the battery, I consider the battery life time to be very good

Positive No Text on top of webpage

I’ve worked with this tablet for several hours without recharging the battery, I consider the battery life time to be very good

(28)

27

On the other hand, I’ve worked with this tablet for several hours but I had to recharge the battery for several times, I consider the battery life time to be very bad.

Positive and negative No Text on top of webpage

On the other hand, I’ve worked with this tablet for several hours but I had to recharge the battery for several times, I consider the battery life time to be very bad.

Table 3.5: Overview of conditions and reviews used

Manipulation check Monetary incentive IV

A control question, which can be considered as a manipulation check, is placed after the questions related to the DVs were asked. The question was along the lines if the respondent noticed that the author of the review can receive an monetary reward of 1 euro for their contribution. From the 87 respondents who were confronted with the incentive condition (author received incentive) 53 individuals noticed the fact that the author can receive 1 euro for the review, against 34 who did not noticed this. This means that the majority (60%) noticed the possible incentive. The manipulation check consist of one question, therefore, no reliability analysis is performed for any internal consistency. In addition, ANCOVA was performed with the group who noticed the incentive. No significant results were found however.

3.3.2 Operationalization of IV: Type of argument

(29)

28

(positive and negative) the second paragraph is the same as in the positive condition and the third paragraph is the same as in the negative condition with the addition of the text ‘on the other hand’ in the beginning of the sentence. This is done to prime the respondent that really a contradiction is taking place.

Anova to test differences between means for type of argument

A 2x3 Anova was conducted to see if there were differences in means between the six conditions in terms of the dependent variable. All assumptions to be made before conducting Anova were satisfied: The Levene’s test of equality of error variances showed for source credibility 0,458 > 0,05, and for content credibility this was 0,095 > 0,05. In table 3.6 an overview is presented for the means of source credibility.

Type of argument Monetary Incentive No monetary incentive

Positive M =3,83, SD = 1,18 M = 4,21, SD =1,33

Negative M = 3,73, SD =1,46 M = 4,55, SD =1,06

Positive and negative M = 4,89, SD = 1,24 M = 4,59, SD =1,32

LSD post-hoc tests (Type of argument)

Mean difference p-value

Positive vs negative -0,12 0,626

Positive vs positive and negative

-0,72 0,003

Negative vs positive and negative

-0,6 0,012

Table 3.6 Overview of means after Anova for source credibility. P= 0,05

LSD post-hoc tests were conducted to see whether there were significant differences between the groups. Table shows that there is a significant difference for the groups with positive versus positive and negative arguments: 0,003 < 0,05. As well, there is a significant difference for source credibility when comparing means for negative versus positve and negative arguments: 0,012 < 0,05. It can be concluded that for source credibility there are significant differences between one sided and negative sided arguments.

For content credibility similar results can be found when conducting the Anova analysis. The results are presented in table. Positve versus positive and negative show a p-value of 0,001. With 0,001 < 0,05 this difference between means can be considered significant. For negative versus positve and negative arguments 0,011 < 0,05. Between positive and negative arguments no significant

differences can be found 0,397

Type of argument Monetary Incentive No monetary incentive

Positive M =3,72, SD = 1,50 M = 4,05, SD = 1,44

Negative M = 3,78, SD =1,58 M = 4,44, SD =1,14

Positive and negative M = 4,80, SD = 1,21 M = 4,77, SD = 1,52

LSD post-hoc tests (Type of argument)

Mean difference p-value

Positive vs negative -0,22 0,397

Positive vs positive and negative

-0,90 0,001

(30)

29

negative

Table 3.7 Overview of means after Anova for content credibility. P= 0,05 Reliability analysis type of argument

Two items are used to test whether the respondent perceives the OCR as one or two sided (positive or negative vs positive and negative). The internal consistency is measured by using Cronbach’s Alpha. The Cronbach’s Alpha for sidedness is 0,870 > 0,6 which is needed for the items to be internal consistent (Maholtra, 2010)

3.3.3 Operationalization of DVs

In this research the dependent variable is the credibility perceived by the receiver of the OCR. When an individual judges the credibility of an OCR (s)he looks at different element which can roughly be divided in source credibility and the content credibility. To measure these types of credibility,

multiple items lists are tested and proven to be valid (Uday et al, 2006). The items used can be found in table 3.4. The source credibility is measured on a five item, 7-point Likert bipolar scale. The content credibility is measured on a four item, 7-point Likert bipolar scale.

Reliability analysis DVs

A reliability analysis is done by controlling for the Cronbach’s Alpha. For the items to be internal consistent it as to meet the following condition α > 0,6. The Cronbach’s Alpha for source credibility is 0,890. Therefore, the items satisfy the condition: 0,890 > 0,6. An analysis is conducted to test if the Cronbach’s alpha could be improved by deleting one of the items. It appears that this is not the case. The Cronbach’s Alpha for content credibility is 0,890. As such, the items for content credibility are reliable, with 0,890 > 0,6. For content credibility goes the same, the Cronbach’s Alpha could not be improved by deleting one of the items. All items used for source and content credibility can be used for further analyses.

3.3.4 Operationalization of covariate Style of processing

Style of processing is adopted in the model as a covariate. In this study style of processing entails systematic processing and heuristic processing. These constructs are used in multiple studies and the items used are tried and tested and turn to be valid (Chaiken, 1980;1989; Griffin et al., 2002). The items for systematic and heuristic processing are slightly modified at their discretion to be more applicable to OCRs. The adjusted items are displayed in table 3.4.

Reliability analysis covariate style of processing

As for the dependent and independent variables a reliability analysis is conducted to verify the internal consistency. For systematic processing a Cronbach’s Alpha of 0,754 was found. As such, the items for systematic processing can be considered reliable since, 0,754 > 0,6. The Cronbach’s Alpha could not be improved when items were deleted. Therefore, all the items can be used for further analysis. For heuristic processing a Cronbach’s Alpha of 0,618 can be found, which barely meets the condition for internal consistency: 0,618 > 0,6. However, the Cronbach’s Alpha could not be

improved by deleting any of the items. Therefore, for heuristic processing all the item can be used for further analysis.

3.3.5 Operationalization of covariate consumer scepticism towards OCRs