Confused consumers and inconsistent CSR : the effect of inconsistent CSR information on consumer confusion and buying behavior of customers that are socially involved

(1)

University of Amsterdam Amsterdam Business School

Confused Consumers and

Inconsistent CSR

The effect of inconsistent CSR information on consumer confusion and buying behavior of customers that are socially involved

Deja Duraku 6120423

dejaduraku@hotmail.com

MSc. In Business Administration- Marketing Universiteit van Amsterdam

Supervisor: Prof. Dr. M. Vock Second reader: Prof. Dr. L. Moratis 31 January 2014

(2)

Abstract

This research aimed to test whether Inconsistent CSR Information about companies leads to consumer confusion and consequently, a lower intent to buy and whether a high score on a scale of Socially Responsible Consumer Behavior (SRCB) had an effect on the relationship between the confusion and the intent to buy. Empirical research revealed that the relationship between Inconsistent CSR information and consumer confusion exists, but it could not be ruled out that this was entirely due to the inconsistency itself. The SRCB did not have the expected effect, but even without it, the model is largely sound. More research is required to clarify the relationship.

(3)

1. Introduction

Corporate social responsibility (CSR) has become one of the most important topics of discussion in business literature in the 21st century (Halme, et al., 2009). Globally, firms are following the trend to become more socially responsible by taking into account the environment and the social world they are operating in, aside from only making profit (Elkington, 1997). Investment in socially responsible funds has increased by 82% between 1997 and 1999, some 2,16 trillion dollars (Scherreik, 2000).

1.1 Context

In the recent past, when firms were criticized for irresponsible behavior, such as poor working conditions for employees and suppliers or pollution, a company’s response would typically be to promise to improve, or to simply ignore the criticism. In the 1960’s consumer activists and professional organizations started to oppose the irresponsible way of doing business that companies could be engaged in.

Nowadays, there are many different protest groups like Greenpeace, Camp for Climate Action or Friends of the Earth, who are trying to bring environmental issues to the attention of the public (Carrigan & Attala, 2001). Due to this increasing demand for transparency by the media and pressure groups, firms can now almost be said to have an obligation to give the public information about their CSR activities. The internet serves as a means of spreading protest groups’ environmental objections and their ideas for a responsible world. The operations of these activists can be seen in Web sites such as, the homepage of the Boycott Nike campaign (www.saigon.com/nike) or the McDonalds boycott homepage (www.mcspotlight.org) (Carrigan & Attala, 2001). From the companies’ perspective, protests have become difficult to contain, because of a greater access to information by consumers. Traditional PR no longer achieves the desired results in the information age and companies must be proactive to a certain degree at least.

Successful businesses are now under a lot of pressure to "give back" to society, for example by giving money to charities, protecting the environment or sponsoring events (Mohr, et. al., 2001; Elving, 2010). CSR is no longer something that a few companies are involved in, but it has become a common activity that most companies pursue to some degree. A tactic has

(6)

developed for companies, where CSR activities have been used to improve a company’s image, in issues that have the consumers' social interest, but there is also a genuine commitment to working on a positive relationship with stakeholders (Yoon, Gürhan-Canli & Schwarz, 2006). Consequently, companies may now have obligations and responsibilities to several different stakeholder groups.

Most research on CSR and reputation focuses on consumers, the most interesting and largest stakeholder group (Pomering & Johnson, 2009). Consumers themselves play an especially important role in building a company’s reputation, it has become common for them to spend their money either punishing or rewarding a company for their responsibility on social issues (Roberts, 1995), the phenomenon is often colloquially described as “voting with your wallet”. Consumers can and do attempt to help the environment and society by considering their buying behavior before purchasing a product that is good or bad for the environment or society.

Even though CSR itself has been researched extensively in the academic literature, there are still areas that have not been researched or have received little attention in this field. Practically, there has been little research about the effect of inconsistent CSR information on the buying behavior of consumers that are especially socially involved. Socially involved consumers care about the power of companies and have opinions about companies’ behaviour towards their employees and the environment. For instance, how does a socially involved consumer react when a coffee company that they buy from is mentioned negatively in the press, despite their CSR program, which the consumer previously admired? Does the consumer stop buying coffee from that company? Only very few studies have focused on the effect of inconsistent CSR information (Wagner, Lutz and Weitz, 2009; Skarmeas and Leonidou, 2013). Most studies to date have only studied the effect of one specific, mostly positive, CSR activity on the consumer. But how do different consumers respond if they are actually confused about a company’s contradictory CSR information? Does the “highly involved” consumer group respond differently to this? And what effect will this have on their buying behavior? Companies need to consider that this type of consumer may just be the vanguard of what all consumers will be like in the future. Consumers already have access to more information than companies can possibly control, and this is likely to only increase. This study will contribute to the field of CSR literature by investigating what effect inconsistent CSR information has on the buying behavior of socially involved Dutch consumers.

(7)

1.2 Problem statement

It should be obvious that inconsistent CSR information could be damaging for a company. Consumers respond to a company’s CSR policies positively if they are good or bad. But many companies begin their CSR policies in response to bad press, the more common CSR profile for a company is that they do something for their employees or the environment, while also engaging in activities that are detrimental to both or either. Consumer confusion seems like a good variable to measure how consumers can react to this. It is a variable that is meant to measure the complex thought processes of consumers, and is closely linked with intent to buy.

Wagner, Lutz and Weitz (2009) have urged that research needs to be done concerning consumer confusion in cases of inconsistent CSR information. The aim of this study is to examine the effect that inconsistent CSR has on consumer confusion and what kind of impact this has on the buying behavior of consumers. Aside from this, Walsh et al. (2007) stated that one of the limitations of the research into consumer confusion proneness is the lack of research on mediators in the relationship, for instance whether consumers can be divided into different groups (Walsh et al., 2007).

Despite previous research to measure the effect of inconsistent CSR information on consumers, there has been no research on what kind of effect this has on consumer confusion in particular, can it explain or mediate this relationship? This is interesting because, consumer confusion, though originally coined for a different type of marketing problem is a well-researched variable, closely linked to buying behavior. By researching consumer confusion, it should be possible to gain greater insight in the actual mechanism that leads to changes in buying behavior. Previous research only focused on the end result, but left little room for finding nuance in the consumer’s responses. The research was conducted primarily to prove or disprove the assertion that companies should be consistent in their CSR information, and it was less concerned with why consumers punish firms for inconsistent CSR information.

Also Mohr, Webb and Harris (2001) emphasize that more research is needed on the consumer awareness of the socially (ir) responsible behaviors of companies. To more fully address that dimension, the moderating factor SRCB will be taken into consideration. It is a measure of Socially Responsible Consumer Behavior. This measure was designed to find out if people with a series of opinions about ethical matters would buy certain products and how much they would pay for those products. But this is an ideal moderator to further elucidate the process

(8)

through which consumer arrive at the decision to buy in cases of inconsistent information, as it divides consumers into different groups based on how much information they already have.

1.3 Research question

Taking all the previously mentioned variables into account, a research question can be formulated:

What is the effect of inconsistent CSR information on consumer confusion and the buying behavior and are these effects moderated by consumers’ SRCB?

The research question is divided into the following sub- questions:

 What is CSR?

 What is inconsistent CSR information?

 What is consumer confusion?

 What is buying behavior?

 What is Socially responsible consumer behavior (SRCB)

 How do all these variables relate?

1.4 Academic and managerial relevance

This study aims to make a contribution to academia by elucidating how more and less socially involved consumers’ intent to buy when coping with Consumer Confusion, which is caused by inconsistent CSR information. Consumer confusion is an existing mechanism for measuring an attitude in consumers that may lead to a change in their intent to buy and linking it to inconsistent CSR could open up a new way of researching people’s attitudes towards companies.

Companies may also find it useful to know how to react and respond to socially involved consumers that get positive as well as negative CSR information about them. CSR was originally seen as a form of public relations damage control. Clearly, society as a whole stands to benefit if companies act in the interest of all their stakeholders, but the only group with the reach and inclination to do so, are those that can affect their bottom line. If it can be discovered how inconsistent CSR information can influence the buying behavior of the socially involved group of consumers, companies may begin to make CSR an inherent part of their business operations.

(9)

This study builds on the existing literature and aims to contribute to the academic knowledge and closing the existing literature gap. First, there is a literature gap between researching what kind of effect inconsistent CSR information has on consumer confusion. Until now, research has focused on researching the how inconsistent CSR information should be dealt with by companies (Wagner, Lutz and Weitz, 2009) or on constructing Consumer Confusion Proneness, i.e. creating a measuring scale for different constructs (Walsh and Mitchell, 2007; Walsh and Mitchell, 2010). The link between inconsistent CSR and Consumer Confusion has never been investigated before. Therefore, this study aims to investigate this relation. It would be interesting from an academic perspective to be able to construct a way of measuring the effect on consumers of inconsistent CSR, it is a common problem among companies that customers may be confused about their products or company, but this has not been directly linked with CSR information as the source of that confusion. Linking more abstract information about companies with confusion can lead to new ways of using it to model what happens to consumers when companies are in some way ambiguous.

Secondly, it may be interesting to include the personal trait SRCB as a moderator, which adds to the existing literature about the difference between high/low socially involved consumers when coping with conflicting CSR information. Most studies have researched SRCB in a purchase setting (Mohr, Webb and Harris, 2001; Mohr and Webb, 2005), but never as a moderator between consumer confusion and buying behavior. Both studies used the variable as a way of modeling different kinds of consumers and explaining different responses between those that scored highly and those that didn’t. The studies were not concerned with whether or not the consumers would actually buy the products they were asking about, they focused mainly on whether a company had a good reputation with these consumers. Using SRCB more directly may lead to it elucidating more of the relationship.

1.5 Structure

The structure of the thesis is as follows. Chapter 2 describes gives an overview of the most important relevant articles from the existing literature. It gives a definition of CSR and explains the problem concerning inconsistent CSR information. Then consumer confusion will be explained and the possible relation between conflicting CSR information and consumer confusion. The next section will focus on what kind of effect this will have on the buying behavior of the consumer. The chapter ends by describing the moderating variable SRCB and

(10)

the conceptual model. In chapter 3, the research method will be outlined. In chapter 4, the results will be presented, and in chapter 5, conclusions will be drawn and discussed.

(11)

2. Literature review

This chapter gives an overview of the relevant articles from existing literature and analyzes them to gain insights on the research topic. This chapter will focus on various aspects, including CSR, inconsistent CSR, consumer confusion, buying behavior and Socially Responsible Consumer Behavior, in turn. These concepts will be defined and prepared for their use in the methodology chapter.

2.1 CSR: Responsibility and irresponsibility

A lot of different definitions exist when we try to find a suitable definition for CSR. Dahlsrud (2006) dedicated a whole paper to defining CSR by analyzing the 37 definitions, which were used most often, according to Google. He examined the definitions based on five dimensions (environmental, social, economic, stakeholder and voluntariness). Even though there are many different definitions, they all have in common the obligations companies have towards People, Planet, Profit. These three elements together are often called the “triple P bottom line” which has to be in balance (Elkington, 1997).

The triple P bottom line reflects the duties companies have when doing business. The way a company does business has an effect economically, socially and environmentally. Profit is about striving towards a long-lasting and sustainable way of making profit. The dimension Planet stands for companies doing their business in a eco-friendly way, taking the environment into consideration. The last P, People, involves the effect companies have on people in and outside the firm, for example ensuring good working conditions, or making sure that the products which are provided by the companies are made according to health and safety rulesand regulations (Willard, 2002). Bob Willard (2002) names these three elements, the three E’s, economic, environment and equity. Essentially, both theories cover the same important aspects of CSR. The concept of CSR can be defined as having more than just the triple P dimensions, it can also includes labor, community relations and business surroundings (Crane et al., 2008; Kacperczyk, 2009). But labor and community can just as easily be seen as sub-concepts of people, and business surroundings are more of a corporate governance issue, which might just as well be adequately assigned to profit. For general discussion purposes, the three dimensions are sufficiently complex, and therefore this is the model that will be used for discussion in this research. CSR can therefore be said to exist as any number of activities by companies that seek to balance the three dimensions of people, planet and profit.

(12)

In terms of public perception, this model is a bit different, though. A company may theoretically be balancing three dimensions, but that may not be how they are perceived. This is in part because the “profit” dimension is a given. A company that does not pursue profit is a charity and those are organizations that are completely dedicated to the other two dimensions, i.e. a non-profit organization. Firms can therefore be said to engage in positive CSR when they undertake activities that are not directly profitable, but are beneficial for the environment or for people, i.e., they aim to treat people ethically or limit the effect on the environment of their business practices. Firms are defined as engaging in negative CSR activities when they have a weak or unethical performance (Alniacik et al., 2010). Or if they make a decision based on the profit dimension that has a negative effect on one of the other two dimensions, for instance, a sweatshop, which is a sound economic decision because the labor is cheap. But the people working in the sweatshop are generally underpaid and are usually working in an unsafe manner. These types of behaviors are deemed unethical by the public, and companies how engage in them are therefore being irresponsible.

Publicity of a firm’s ethical actions has an effect on the consumers’ attitude to the firm (Skowronski & Carlston, 1987). Boulstridge and Carrigan (2000) concluded that, for consumers, that a firm’s level of social responsibility was not the most important reason to buy a product. They showed that consumers still bought products of unethical companies if they were aware of the unethical activities. Consumers that bought a product on regular basis found it hard to boycott a company because of unethical behavior, because their personal reasons for buying the product weighed heavier than societal reasons against buying. But according to Mascarenhas (1995) companies are even being stimulated to act in an ethical way, because the information about a firm’s ethical behavior can affect the sales and image of the company. Harmon and McKenna-Harmon’s (1994) research also supports the assumption that negative information has a stronger effect on attitudes than positive information does. They state that customers are more likely to share negative than positive information with each other about the products and services of a company. This also applies to information from the Mass media, as researched by Branch and Tkacik (2003) and Sheban (2006).

Consumers are therefore said to have a “negativity bias” in measuring a company’s behavior, they weigh negative behaviors more heavily than positive behaviors (Reeder & Brewer, 1979; Skowronki & Carlsong, 1987). Earlier studies have stated that negative information about a firm can influence a consumer’s attitude towards the firm, more than

(13)

positive information does (Herr et al., 1991; Amine, 1996). This means that if a company has been involved in an unethical way of doing business, this is a bigger indication of the firm’s characteristics for the consumer, than when a firm behaves in an ethical way. For instance, a firm that is doing business in an unethical way by causing pollution with their production, provides stronger evidence that the company is a unethical firm, compared to a company that avoids pollution in their way of doing business, which though positive, goes largely unnoticed (Folkes, 1988; Folkes & Kamins, 1999). Typically, when a consumer is conscious of information about a firm’s unethical behavior, he/she would sooner boycott the firm’s products. But at the same time it does not mean that when a firm behaves ethically, a consumer will be more likely to buy their products. Ingram, Skinner and Taylor (2005) found that consumers that were highly committed could forgive companies for their unethical behavior when the damage was relatively low. The higher the damage of the company, the less forgiving consumers were. There is also the issue of companies focusing on one CSR issue to the detriment of others. As previously discussed, there are three sides to the CSR issue: People, planet, profit. (Elkington, 1997). A company might focus on environmental effects to the detriment of the effects on employees or suppliers. In some cases it may even be impossible for a company to achieve anything in both main themes at the same time. For example, to give coffee farmers a fair price, they have to produce an ecologically unsound amount of beans from their land if the price of the final product is to remain competitive (Martínez-Torres, 2008). Companies may suffer from various catch 22’s that cannot be explained properly to the public, but can be exploited by activists as a stick to beat the company with. Aside from this issue, some companies are in traditionally “bad” industries (Yoon, Gürhan-Canli and Schwarz, 2006). Examples are oil and tobacco. The companies may be exemplary within their industry, may do everything better than their competitors, but still make a product which is toxic to the environment or consumers. Ultimately, for these companies, it is not about being perceived as good, but about being perceived as the lesser of two evils.

The reputation of a company is linked to how ethically or unethically consumers view the business that has been carried out (Brunk, 2010). Brunk’s study shows that reputation is how outsiders see the organization and that this is beyond the company’s control. Many companies have experienced a setback in their reputation, because consumers found their way of doing business unethical and several companies’ reputations have been damaged because of unethical business practices, for instance Nike, Gap, Nestlé, Shell and Enron (Brunk, 2010). When

(14)

consumers join forces and boycott a company for making unethical business decisions, they can obtain significant publicity and consequently, power. In order to avoid such negative situations, companies want to be seen as socially responsible (Sen, Gurhan-Calli & Morwitz, 2001; Klein, Smith & John, 2004). The sweatshop incident involving Nike ultimately made the company change in a positive way, they revealed potentially harmful details about their manufacturing operations and started to be more transparent about their business towards consumers (Carter, 2005). But there is no indication that the company would have changed anything about it’s business practices had there been no negative publicity.

The question arises, why a company is doing CSR in the first place. Yoon, Gürhan-Canli and Schwarz (2006) examined the different types of motives for a company to engage in CSR and how knowledge of the motives can improve a company’s image, but can also backfire: namely, when the sincerity of motives is ambiguous. When consumers suspect a company’s motivation for CSR is only to improve its image, they believe that the motives may not be honest. This can cause a backfire-effect, making the CSR activity ineffective and also damaging the company’s image even more than it already was. Interestingly enough, Yoon et al.’s research focused on the effects of CSR activities by companies with a bad reputation. They concluded that even if the company has a bad or good reputation, it is still very important in which CSR activities a company is involved, making the salient benefit an important factor. Also, if a company has a good or a bad reputation and they are supporting a good cause, it can backfire, if it becomes known that the advertising-related expenses exceed the CSR contribution itself.

The definition of CSR that will be used for this paper is a broad one, any activity a company can engage that is not directly for profit, but is beneficial to one of the other two dimensions, i.e. people or planet. However, the actual CSR practices of companies are not under investigation here. Most companies provide CSR information themselves, for public consumption. This information is likely to be somewhat skewed towards the company’s point of view. Since CSR has become very fashionable recently, many companies provide some information about their CSR efforts, usually on their websites. Companies are unlikely to be the source of negative CSR information about their own activities, so this information usually comes from outside the company, from the media or an activist group. The possibility arises that this information can be conflicting, some information will claim the company is doing well, and has various activities that support the environment and its people, other information will be negative, i.e. claiming the opposite. The main thrust of this research will be on whether

(15)

or not this “inconsistent” CSR information is likely to lead to consumer confusion, and whether this also has an effect on buying behavior. But what is inconsistent CSR?

2.2 Inconsistent CSR

Communication about the CSR initiatives of a company may come through a firm’s marketing communication efforts, like marketing programs, but also from the media, non-government organizations, etc. But often the given information is seen as inconsistent, because of the contradictions in CSR information from several different sources, which can lead to a reduction in credibility (Schultz and Morsing, 2003). This could lead to consumers getting confused and beginning to view the company more critically. Reports of incidents of unethical corporate behavior are a regular fixture in the media. Frequently people doubt the investment in CSR activities and whether the company actually does what it says it does.

The study of Wagner et al. (2009) is the only research that has investigated how consumers react to inconsistent CSR information, in their definition. Other existing literature has investigated the impact of either positive or negative CSR information. In Wagner et al. (2009) the term ‘corporate hypocrisy’ is used to describe the claim of a firm to be something that it is not. There are several cases of firms representing the perception of being socially responsible, but behaving in a socially irresponsible way. This makes the consumer question the firm’s actual behavior, which will result in damaging the firm’s reputation and their actual sales. Therefore overall consumers have high expectations of companies “giving back” to the environment and the people. But, as seen in previous research communicating CSR activities can also provoke a skeptic reaction in consumers (Skarmeas & Leonidou, 2013), which, according to Wagner et al. leads directly to lower sales.

According to Wagner et al., the two types of strategy that companies use for this are reactive and proactive strategies, i.e. the company implements CSR to ‘fix’ an issue that has received negative publicity, or a company is always involving itself in CSR activities. Consumers can react with skepticism to either strategy, and respond negatively to insincerity. The motivation cannot be transparently to the company’s own benefit, or to the benefit of its stakeholders, but must be at least somewhat value driven, even idealistic. Wagner et al. use the construct of perceived corporate hypocrisy, but for this research, their way of modeling inconsistent CSR as a set of two conflicting statements is more useful than the construct of perceived corporate hypocrisy.

(16)

Therefore, inconsistent CSR information will be defined in this study as contradictory, i.e. ethical and unethical, information about a company’s social and environmental responsibility. In practice, this means that respondents in this empirical study will be exposed to at least two pieces of CSR information, one of which will be positive in nature, and one negative, and these will contradict each other.

2.3 Consumer confusion

After discussing inconsistent CSR from the companies’ perspective, it is now time to focus on the consumer. To describe the effect of inconsistent CSR, this sub-chapter will discuss consumer confusion.

Consumer confusion happens when people do not exactly understand the product or service they are dealing with. This has been seen to have an effect on the buying behavior of consumers, which causes the sales to decrease and reduces satisfaction. It can also cause negative word of mouth, cognitive dissonance, decision postponement, decreased trust, and decreases in loyalty (Walsh & Mitchell, 2007). Confusion can be defined in many ways. Turnbull et al. (2000, p, 144) define confusion as: “error that arises when a consumer is not able to interpret the facets of a product/service in the right way, while processing the information”. Pin et al. (1995, p.108) mention the “concept of consumer confusion”, that says that consumers want to know exactly what they get, when and how, without getting more options to choose from. Consumers are provided with an ever-expanding amount of information about more products that are sold through more channels and that are advertised in more ways (Mitchell et al., 2005). Confusion among consumers has become a larger problem, because of the increased amount of information that is available in their purchasing environments. Walsh and Mitchell (2007) state that “consumer sovereignty” is necessary, the consumer must have enough information about the product and fully understand the information, so that they can make a well thought-out choice. However, this is not the case when the consumer is confused. The propensity for consumer confusion is known in the literature as consumer confusion proneness (CCP). Walsh & Mitchell (2007) define consumer confusion proneness as a condition where consumers have to process similarity of products, overload of information and ambiguity of information, which negatively influence the processing of the information and the making of an informed choice. According to Walsh et al. (2007), confusion has a conscious and unconscious aspect, which makes awareness an important factor. This is because consciousness can make

(17)

consumers aware of the fact that they are confused and gives them the possibility to reduce it. The opinion of the authors is that consumers are conscious of the fact that they are confused and that they have to find a solution to handle it.

In the existing literature, five types of confusion are described: lack of consistency,

information overload, choice overload, similarity and lack of information. The first aspect is similarity. This is when products or services show a resemblance in quality, characteristics or

appearance. The products or services are alike, because they lack different features, which help make the consumer’s decision easier, but they are not identical. Similarity of products and services has a negative effect on the decision making prowess of the consumer, reducing their satisfaction when purchasing the product or service (Walsh & Mitchell, 2007). The second aspect is choice overload, which occurs when the purchasing options of a product or service while choosing or selecting, are too broad. It then becomes overpowering and uncontrollable for the consumer to make a well-considered decision (Broniarczyk, 2008). Overload of information can occur when there is too much information about a product or service that a consumer has to process before making a decision to purchase it or not. The information provided makes the buying process complex and time consuming (Mitchel et al., 2005). This overload of information can have a negative effect, it distracts the consumers from the products themselves (Carrigan & Attala, 2001). Lack of information is when the information provided is insufficient or is absent when it is required. The information is too little to develop an understanding of the product or service, and a decision to buy is unlikely to occur. The Dragon International study (1991) shows that only 26% of the respondents could name a company that is socially responsible and only 18% could name a company that is socially irresponsible. Carrigan & Attala (2001) and Boulstridge and Carrigan (2000) concluded in their respective researches that most consumers lacked information to distinguish whether a company’s way of doing business was unethical.

(18)

1. Five aspects of consumer confusion

Lack of consistency happens when the information provided for the service, product or

company is not in accordance with the consumer’s beliefs. There is a shortage of agreement between what the consumer’s perception is and the information provided about the service and product (Walsh et al., 2007). Based on the five types of consumer confusion, lack of consistency is probably the most useful one to combine with research on the effects of inconsistent CSR. It matches up well with the concept of “Confusion marketing”, which can be described as the process in which companies use their marketing techniques to make the product/service so complex that it confuses the consumer into making the “right” decision (Drummund & Rule, 2005). It is not so much that companies spread false information on purpose, but that they try to willfully obfuscate the issues that may concern consumers. An example would be something like the term “low-fat”. The term is not exactly unambiguous, it only refers to itself, and is in no way indicative of the actual fat content. The amount of fat is deemed low, but in comparison to what exactly? CSR information could be similarly misrepresented, especially when contradictory (i.e. positive and negative) CSR information is presented to the consumer.

Consumer confusion has been linked to CSR before. Chen & Chang’s (2012), their definition is that it is a product-specific marketing problem and that consumers would have to pick which product is “greener”, but that companies deliberately make it difficult to find out which product this is. But companies’ CSR initiatives are even more complicated than that. For instance Shell, unable to mitigate the environmental consequences of oil production, focused on other aspects of CSR, and developed initiatives in other countries or areas of their business, or even completely unrelated to their business, such as primary schools not even in the area. This could be seen as “greenwashing”, the process whereby a company represents itself as more

Similarity Consumer confusion Lack of consistency Choice overload Information overload Lack of information

(19)

environmentally friendly than it actually is. In Chen and Chang (2012) the relationship between greenwashing, “green trust”, “green perceived risk” and “green consumer confusion” is outlined. Two of these variables are defined in similar ways, but from a different perspective. Green trust argues from the point of view of a consumer who is willing to believe in a company, green perceived risk from the more cynical perspective of a consumer who already expects that the product will impact the environment in a negative way. Green Consumer Confusion as defined in Chen & Chang (2012) is mostly drawn from one of the five dimensions of consumer confusion, namely similarity, i.e. a product that resembles another. The idea behind this is that it is difficult to distinguish which product is “greener” than another product, making it a similar problem for a consumer as when buying, for instance, a generic product that has similar packaging to a name brand. Greenwashing is defined by them as a company seeking to present itself as greener than it actually is through confusing information. Chen & Chang see this as one of the main avenues through which companies confuse consumers for their own interests.

But Chen & Chang’s research is focused on a product and not on the company itself, but the questions in their surveys are comparable to Wagner et. al. (2009), both being based on Walsh et.al. (2007). But where Wagner focused on companies, Chen and Chang focus on products. This research uses Chen and Chang’s concept of consumer confusion, but applies it to Wagner’s view of the company. Taking the relationship between “greenwashing” and “green customer confusion” outlined by Chang et. al. (2012) and applying them to companies as a whole, using the same methods as in Walsh et al.(2007) creates a framework for a hypothesis about a relationship. The reasoning for this hypothesis is as follows; Inconsistent CSR can be easily described as ambiguous information, which corresponds clearly to one of the existing aspects of consumer confusion, “lack of consistency”, which will be used in this empirical study.

Hypothesis 1: Inconsistent CSR information will lead to more consumer confusion than

consistent CSR information.

2.4 Buying behavior

Buying behavior, which will be discussed next, is often used as a variable when surveying consumers about ethical issues, and has also been linked with confusion.

(20)

households buying goods and services for personal consumption (Kotler, 2004, p.189). The first definition of consumer behavior, given by Belch (1978, p. 21) is ‘the process and activities people engage in when searching for, selecting, purchasing, using, evaluating, and disposing of products and services so as to satisfy their needs and desires’. Schiffman’s research into consumer behavior attempted to explain how individuals make the decisions to spend their available resources (time, money, effort) on commodities. But consumer behavior is not limited to what consumers buy, but also focuses on why they buy, where they buy, how often they buy and how often they use it (Schiffman, 1997). For this research only the intention of buying will be researched.

Buying intention is an important aspect that has to be taken into account, because it’s a useful tool for measuring if inconsistent CSR information will make consumers delay or even cancel their purchase, thus investigating whether inconsistent CSR information and consumer confusion has an effect on sales. Customers are expected to delay purchase decisions or cancel purchases altogether if they are uncertain they are making the right choice. A variety of surveys have shown that consumers say that they are influenced by the CSR profile of a company in their buying intent (Smith, 2003). On the other hand, there is no certainty that consumers always buy the products of companies that do good. As mentioned in 2.1, there is a negativity bias, so in fact they may only not buy products from an unethical company. Castaldo and Perini (2004) report that there is also a possibility that consumer may not be aware of the bad CSR activities of a firm when buying their products. Creyer and Ross (1997) state that people find the ethicality of a firm’s behavior is important when consumers are considering buying something from the firm. Consumers expect companies to exhibit ethical behavior and will also reward ethical behavior by being willing to pay more for that firm’s product. But in Boulstridge and Carrigan’s (2000) study, most consumers claimed social responsibility was not an important factor in their intention to buy. Boulstridge and Carrigan (2000) did agree with Ceyer and Ross (1997) that consumers would still buy products at a lower price from less socially responsible firms, even knowing about the unethical activities. Ulrich and Sarasin (1995) found that consumers in general care more about price, quality and value than they do about the social issues. According to Carrigan & Attala’s (2001) study the factors brand image and fashion trends were more important components during the purchasing than the aspect of a firm’s social responsibility. Before investigating how this has an effect on the buying behavior, first consumers need to be aware of a firm’s level of social responsibility. Only then can this factor in their purchasing (Mohr et al., 2001).

(21)

Although Russell and Russell (2010) were actually trying to prove the importance of geographical proximity on CSR effectiveness and it only looked at positive CSR activities, to the exclusion of harmful firm activities, the paper yielded interesting results for this research. The study found that consumers’ awareness of CSR activities and their purchase intentions are linked positively. Another promising result was that a higher awareness level of corporate social contribution and local community contributions, both common CSR activities, were found to have more positive effects on consumers’ purchase intentions compared to no awareness by consumers. This could mean that consumers seek “good” CSR activities, and may be prone to buy products from “good” companies. But this effect was only found on the “people” dimension of CSR initiatives, environmental activities seemed to have little effect on consumer’s intention to buy. In “Uncertainty in ethical consumer choice” by Hassan et al. (2013), an attempt was made to do a qualitative analysis of consumer’s uncertainty. The conceptual model created explains the causes of the uncertainty to stem from the complexity, ambiguity, conflict and credibility of the company’s CSR information and information from other sources. This resembles the way inconsistent CSR information is conceived of for this paper. The result of conflicting information is consumers that have negative emotions about companies, have to compromise their beliefs and, most importantly, delay purchases. This is a promising result that should lead the way in this thesis, all the research confirms that this is likely to be the case. This leads to the second hypothesis:

Hypothesis 2a: Inconsistent CSR will result in a lower intent to buy than consistent CSR.

Consumer confusion is related to lower buying behavior, not only according to Hassan et al. (2013), but also according to Mitchell and Papavassilou (1997), whose research ranked delay or abandon purchase the third most likely thing for a confused consumer to do, only just behind asking for more information, or rethinking the purchase entirely. They modeled consumer behavior as “confusion reduction strategies”. In Mitchell and Papavassilou (1999) they summarize the effect as “When consumers abandon a specific purchase, they often reduce the importance of acquiring the product through dissonance mechanisms and/or transfer the purchasing goals to a product category with which they are more familiar and comfortable”. Interestingly enough, all of the other five generic strategies they propose mention abandoning or delaying purchase by involving other people or asking for more information. Existing research points to a lower buying intention when consumers are concerned about a company’s ethics, and when consumer confusion is higher, the intent to buy is lower, according to Mitchell and

(22)

Papavassilou (1999) and Hassan et al. (2013). It has not been established that consumer confusion is also the mechanism through which Inconsistent CSR is influencing buying behavior. Inconsistent CSR has not been linked directly to consumer confusion before, despite its usefulness as a model for explaining a decrease in buying intent. This leads to the third hypothesis:

Hypothesis 2b: Consumer confusion will mediate the relationship between inconsistent CSR

information and buying behavior.

2.5 Socially Responsible Consumer Behavior

Consumer confusion seems the likely way to model the way inconsistent CSR effects buying behavior, but consumers themselves can be considered a variable in this research as well. The last few decades have given rise to the most critical and educated consumers that marketers have ever seen before (Caminiti, 1992). Most of the consumers make their consumption decisions based on characteristics like, price, quality and convenience or based on the brand name (Boulstridge & Carrigan, 2000). This is not surprising, considering that for the average consumer price and quality are of a higher priority than being socially responsible. This is the way in which self-interest often outweighs other concerns (Etzioni, 1988; Wolf, 1987).

But when typing in “socially responsible product” in 2014, Google’s search engine finds 6 million hits compared to 2011’s finding of 1.26 million (Ha-Booksire & Normun, 2011). This large increase proves the increasing popularity of socially responsible products and the increased consumer demand for that kind of product. Although the concept of ethical purchasing has existed for a long time, the consumer awareness has increased in recent years (Shanka & Gopalan). Certifications such as the “fair-trade” label help consumers that attempt to purchase in a socially responsible way and distinguish ethical from unethical products.

The process of not buying products, which the consumer deems unethical, is sometimes called boycotting (Paek & Nelson, 2009). Consumers are continuously considering their wants and their wish to act in a socially responsible manner and their awareness about societal and environmental issues is rising (Ha-Booksire & Normun, 2011). Roberts (1995) investigated the trait ‘socially responsible’ and according to his research, 32% of the American population could be seen as socially responsible consumers. This group is more liberal, more environmentally worried and has a higher tendency to want to influence their environment with their consumer

(23)

behavior. Mohr et al. (2001) found that a small group of consumers were interested and willing to find out what companies are doing concerning CSR. This aspect shows us how important at least some consumers consider CSR to be, and how big the difference between socially involved consumers and consumers who care less could be.

According to Mohr et al. (2001) a reason for the lack of ‘socially responsible consumer buying-behavior’ is because consumers lack the knowledge about responsible products and CSR and that they experience difficulty in both purchasing and understanding the information about a company’s CSR involvement. Many studies have looked at personality traits in relation to CSR. For instance, Einwiller et al. (2006) have researched the consumer’s level of identification with a company, when coping with negative publicity about a company. Consumers who had a strong identification with a company because of its CSR efforts, would have less negative responses to the company, when dealing with negative publicity about the company. Ahluwalia, Burnkrant & Unnava (2000) added the factor commitment, which makes consumers resistant to negative information when a brand is under attack. Ahluwalia et al. (2000) state that when there is a high degree of commitment toward a brand, the consumer can filter the negative information separately, when negative publicity occurs.

Bhattachary and Sen (2003) have also looked at the impact of the identification level between the consumer and the company. They found that consumers that have a high level of identification with a company are often more loyal to them and have a higher willingness to promote the company to others, this is similar to Einwiller et al.’s (2006) findings. These consumers are resistant to negative information about “their” company.

The inverse of this research is less well documented. It stands to reason that people with a highly developed social consciousness would be more critical of companies than the average consumer. An interest in social causes can supply consumers with a lot of information from activist groups. But these activists have their own agenda, which can run counter to that of the company. It is more difficult for companies to appear to be doing good, even people with a highly developed social consciousness may be so cynical about the concept of capitalism that no company’s actions will ever satisfy their standards. Hassan et al. (2013) found a negative relationship between consumers “uncertainty” and buying behavior, but this research was qualitative and highly focused. To find the relationship, a certain type of consumer was sought out, i.e. one with a highly developed ethical sense. They were asked about their feelings about how ethical they thought their purchasing decisions were, and almost all of them were uncertain

(24)

or conflicted about them, leading to delays in almost all cases.

There are different personality trait variables that can affect how strong a consumer responds to a company’s level of social responsibility. One of these personal traits is the social responsibility variable of the consumers. In 1975, Webster outlined a socially responsible consumer as: “a consumer who takes into account the public consequences of his or her private consumption or who attempts to use his or her purchasing power to bring about social change” (p.188). In 2001 Mohr et al. (p.47) redefined this as SRCB, Socially Responsible Consumer Behavior, i.e. a consumer who wishes to use his or her private budget to minimize or eliminate the effect of products that have potential negative effect on the environment and maximize benefits in the long run for society. SRCB can be more prevalent in particular groups within society like women, minorities, immigrants etc. But it can also be a factor in particular protests like gay rights, animal use in products, religious associations and so forth (Roberts, 1995). SRCB is a personality trait that also reflects the self-image of a person, i.e., it is more likely to occur both in people who will benefit from a certain social action and in people who won’t but are idealistic. These groups of people are prepared to learn more about what companies are doing and how involved they are in CSR. Because these consumers feel responsible for their own contributions and think it is important for them to support causes through their behavior (Mohr et al., 2001). People that score highly in SRCB, will try to maximize benefits for society in the long run in several circumstances, through their buying behavior. With every consumption, they take into consideration what the effect will be for the rest of society, because they want to express their current social concern.

Currently, people have faster, constant access to information through the internet, making much more information about the CSR activities of companies available for the consumer. Research should redefine the socially conscious consumer and what effect inconsistent CSR information has on them. The research on the topic of consumer’s attitudes to companies where they relate to their values has been thoroughly researched, but nearly always in a different way. For this research, SRCB is an appropriate variable. It succinctly covers the concept of a consumer’s social attitude and defines it as a property that a consumer can have to a certain degree, allowing for a scale to exist.

Customers who have a high SRCB score are expected to be more likely to have a low intent to buy if they are confused than customers who are less socially involved. The reasoning being; a high SCRB score corresponds to specific values, which when exposed to more

(25)

information will lead to consequences in the buying behavior, they are more likely to act on their increased confusion. This paper assumes that there is a negative relationship between consumer confusion and intent to buy and that this negative relationship is higher for those with high SRCB scores. This negative relationship was seen in Hassan et al. (2013), but that was a small sample qualitative research, and they specifically selected a sample of people with “a highly developed ethical sense”. This paper has a somewhat larger sample and a more structured survey, and will attempt to contrast high SRCB scores with low ones, so hopefully the same relationship will be found and can be subjected to statistical significance tests. This leads us to hypothesis 3:

Hypothesis 3: SRCB will moderate the effect consumer confusion has on the intent to

buy.

2.6 Research Model

This is the proposed model that this research will attempt to prove. In the following chapter a methodology will be outlined. The various concepts and their proposed interrelationships will be operationalized. 2. Conceptual model Inconsistent CSR SRCB Buying behavior Consumer confusion H1 H2a H2b H3

(26)

3. Method

This section describes the research method, the data collection, the conduct of the survey. First, the research design will be explained and then the characteristics of the participants will be described.

3.1 Research design

The purpose of the study is to find out what kind of effect inconsistent CSR information has on customer confusion and consequently on the buying behavior of customers that are socially involved and those that are not as involved. The data will be collected through a questionnaire based-survey. Input from several respondents is necessary to complete the survey, the respondent can fill in the survey in their own time, remain anonymous and it is easy to compare groups with each other (Saunders et al., 2009), making the survey strategy reliable and valid. In this study, the perceptions of the respondents have to be researched, how the inconsistent CSR affects the customer confusion on how this will affect their buying behavior. Also, previous studies with similar research variables have also used the survey strategy (Roberts, 1995; Mohr & Webb, 2005; Yeung, 2009). The questionnaire is a standardized form (Saunders et al., 2009, p. 373). Participants who do not finish the survey will be excluded from the analysis.

Aside from direct questions about participant’s behavior, meant to gauge their SRCB score, the survey will attempt to measure behavioral intent and confusion by presenting participants with scenarios. The survey has a quasi-experimental design because of these scenarios. These Scenarios will test how respondents react to inconsistent CSR information, i.e. positive and negative CSR information, and compare this to consistent CSR messages, i.e. wholly positive or negative CSR messages. The scenarios were written about a fictional company, to prevent any outside factors to intrude. Two of the scenarios will have consistent CSR (one positive, one negative) and two of the scenarios will have inconsistent CSR, (one starting with a negative piece of information and then a positive, and one starting with a positive piece of information and then presenting a negative piece), to control for order effects. Each respondent reads only 1 of the 4 scenarios. After reading the scenario, the participants are asked to describe their confusion and their intent to buy.

The scenarios are similar to the ones used by Wagner et al. (2009) and the questions were based on Walsh et al. (2007), Chen &Chang (2012) and Hassan et al. (2013). The actual

(27)

text of the scenario is loosely based on a real incident, clothing brand H&M’s publicity crisis after a factory that made their clothing collapsed. The company decided to invest more in working conditions in the country in the wake of the scandal (Forbes, 2013). Using the actual company and incident was considered, but this could have influenced results. Wagner et al. (2009) used a fictional company specifically to avoid this, and this research does the same.

The design of the survey allows for 8 conditions to be tested, because it is a 4 CSR messages x 2 SRCB design. Namely, (2 inconsistent (positive -negative and negative-positive) x 2 consistent (positive and negative)) x 2 SRCB (high, low).

3.2 Measures

Independent measures: Half of the scenarios in the survey will present respondents with a

fictional company’s contradictory CSR information. The other half will be scenarios with consistent information, used as control. The company is fictional to control other factors, specifically respondents other possible reasons to buy or not to buy from a real world brand. This is how “inconsistent CSR” is modeled in this research, similar to Wagner et al. (2009). The survey present the respondents with general information about the company first, it is stated that the company produces affordable clothes and that the respondent has no strong opinions about the company before the scenarios are presented. The scenarios describe that the respondents see newspaper articles about the company and articles from the company’s newsletter.

In scenario A, the respondent is told that the newsletter states that the company will invest 1 percent of its profits in Bangladesh, because that is where its clothes are made. Subsequently, a newspaper article states that the company is under investigation for the “inhuman” conditions in its factories in Bangladesh.

In scenario B, the respondent is told that a newspaper article states that the company is under investigation for the “inhuman” conditions in its factories in Bangladesh. Subsequently, the newsletter states that the company will invest 1 percent of its profits in Bangladesh, because that is where its clothes are made.

In scenario C, the respondent is told that the newsletter states that the company will invest 1 percent of its profits in Bangladesh, because that is where its clothes are made. Subsequently, a newspaper article states the same and that the company has been given a development award by the government.

(28)

In scenario D, the respondent is told that the newsletter states that the company has moved its production to Bangladesh to deliver cheaper clothing for its customers. Subsequently, a newspaper article states that the company is under investigation for the “inhuman” conditions in its factories in Bangladesh.

Dependent Measures: To measure lack of consistency confusion a 7 point Likert scale is used,

with three questions adapted from Walsh et al. (2007). The questions were:

Ik weet genoeg over Van der Linden: adapted from: When buying a product I feel sufficiently

informed

Ik voel me goed geïnformeerd over Van der Linden: adapted from: When buying a product I feel

sufficiently informed

De informatie over Van der Linden is duidelijk: adapted from: The information I get is so vague

I often do not know what a product can actually do.

The same scale is used for intent to buy, with three questions adapted from Schweizer et al. (2006). Those questions were:

Het is zeer waarschijnlijk dat ik nog iets van Van der Linden koop: adapted from: It is likely that

I will buy a product from this manufacturer again

Als ik nu moet beslissen, dan zal ik waarschijnlijk iets van Van der Linden kopen: adapted from:

If I have to decide now, I will buy a product from this manufacturer

De kans dat ik iets van Van der Linden zal kopen is hoog: adapted from: The likelihood that I

will buy a product from this manufacturer is high.

Moderating Measures: To measure SRCB, a series of questions on political engagement,

attitudes and knowledge of social and environmental issues is added to the Questionnaire. These were adapted from Schweizer et al. (2006), Roberts (1995), Mohr & Webb (2005) and Antill (1984). The answers to these no questions are tallied, and the highest scorers will be in the ‘high’ SRCB group and low scorers in the ‘low’ SRCB group.

Controls and demographics: Control questions about the questionnaire were also included in the

survey. Three questions about “Trust” were added to round out the questionnaire, but these are mostly there to check for attitudes that can not be captured in purchase intent. These questions were adapted from Doney & Cannon, (1997).

Ik vertrouw er op dat Van der Linden verantwoord goede producten maakt.: adapted from: I

(29)

Ik kan er van op aan dat Van der Linden ethisch handelt. : adapted from: I can rely on this

company’s ethical behaviour.

Ik heb vertrouwen in Van der Linden als merk: adapted from: I trust this brand.

Additionally, two questions about how realistic and believable the scenarios were, were added.

3.3 Pre-test

A short quantitative pre-test was performed with 20 respondents through an online questionnaire. Four scenarios would be used, two with consistent CSR information (negative CSR information or positive CSR information) and two with inconsistent CSR information (positive- negative CSR information and negative-positive CSR information). The respondents were told that the topic of the questionnaire was CSR information. The most important aspect was to find out if the four scenarios were suitable for the experiment. Therefore, the scenarios were rated on how trustworthy the respondent found the scenarios. Secondly, some practical matters about the questionnaire could be determined, i.e. the length it would take to fill out, any mistakes in the wording, etc. Other than these additions, the questionnaires were identical to the ones used in the actual survey.

For the pre-test, 10 respondents received the scenario texts about inconsistent CSR, of which 5 received the positive negative version and 5 respondents the negative-positive version, and 10 respondents the other two scenarios of consistent CSR, 5 the positive and 5 the negative version. Also included in the questionnaire were control questions about credibility and realism. At the end of the questionnaire the respondents were asked if they had trouble with any of the questions, if aspects were unclear and if the survey took too long to fill in. The respondents deemed the survey clear and easy to fill in. The only issue in the pre-test was the lack of high SRCB scores in the sample. The sample consisted of students at the university, mostly. The exact results can be found in Appendix C, an excel spreadsheet with all the results.

Tabel 1: Cronbach’s Alpha for Credibility

Number of items N Mean Std. Deviation Cronbach's Alpha Credibility 3 20 5.1091 2.40883 0.963

(30)

Table 2: One sample T-test

t df

Sig.

(2-tailed) Mean Difference

Credibility 4.985 19 0 1.10914

Table 2: Level of education

Level of education Respondents

Elementary school 0

MBO (low vocational education) 2 HAVO (high vocational education) 3 VWO (high vocational education) 3 HBO (higher professinal education) 4

WO (university) 8

Table 2: Political Affiliation

Political Affiliation

Political Party Respondents

VVD 7 PvdA 1 PVV 2 CDA 2 SP 1 D66 5 Groenlinks 2 Christenunie 0

Partij voor de Dieren 0

Did not vote 0

Total 20

Table 3: Income

Income distribution in Euros High of income Respondents

Lager dan € 10.000 6 €10.000 - €20.000 1 €20.000 - €30.000 3 €30.000 - €40.000 2 €40.000 - €50.000 2 Hoger dan €50.000 6 Total 20

(31)

3.4 Sample

The responses were collected in the Netherlands, including both female and male respondents, and the questionnaire was written in Dutch. Sample size has to be around 200 respondents. The amount of respondents has been chosen because within the limited time schedule, this is the most reasonable amount of respondents that can be collected that will still yield reliable results. All respondent will receive the closed-end questionnaire with one inconsistent or consistent CSR information scenario. To test all 8 conditions in sufficient numbers, 50 people will have to fill out each of the 4 scenarios, and at least half will have to have a high SRCB score. At least 25 respondents for each condition, but because the respondents can not be pre-selected for SRCB scores, the numbers have been carefully monitored during the data collection. To ensure enough respondents with a higher SRCB, various protests, found through the schedules of activists have been attended. The subjects that are needed for this research can be found through the website:

http://socialisme.nu/blog/agenda/. There are several nationwide protests against climate chance, animal cruelty, the food industry and working conditions. Also through the Facebook page of Greenpeace a randomly selected group of respondents that demographically represents the population will be approached. It was expected that attendants of these protests will have a higher SRCB score than the general public. Additionally, a low cost vegan restaurant with a particular focus on left-wing politics was visited and the survey was distributed there. Only 9% percent (22 surveys) of the total data collected was done in these locations. The purpose was solely to gather enough respondents with a high SRCB score, these were difficult to find in the general population. Keeping the percentage collected at protests and the restaurant was

considered vital, to reduce any possible bias from the sample location. Demographic information

The eventual sample size was 204 respondents. This meant that the reliable amount of 200 responses was completed. The average age of the respondents was 25 years, considerably less than the average age of 39. The different categories for the ages were: 18-25, 25-35, 45-55, 55-65, 65 and up. 45,3% of the respondents were men and 54,7% were women, slightly more women than the national averages of 49,5 and 50,5. (CBS, http://www.cbs.nl/nl-NL/menu/cijfers/kerncijfers/default.htm)

(32)

Table 1: Level of education

Education

Level of education Respondents

Elementary school 1

MBO (low vocational education) 9 HAVO (high vocational education) 8 VWO (high vocational education) 14 HBO (higher professinal education) 55

WO (university) 117

Table 2: Political Affiliation

Political Affiliation

Political Party Respondents

VVD 83 PvdA 9 PVV 6 CDA 5 SP 4 D66 66 Groenlinks 6 Christenunie 2

Partij voor de Dieren 2

Did not vote 21

Total 204

Table 3: Income

Income distribution in Euros High of income Respondents

Lager dan € 10.000 75 €10.000 - €20.000 21 €20.000 - €30.000 28 €30.000 - €40.000 27 €40.000 - €50.000 22 Hoger dan €50.000 31 Total 204

(33)

Table 4: Age Distribution

Tabel 5: Number of respondent across 8 conditions SRCBd Total Laag Hoog Cond P-N 32 15 47 N-P 32 18 50 N 27 29 56 P 21 30 51 Total 112 92 204

(34)

4. Results

In the first section of the results the sample will be discussed, after which reliability of the items will analyzed. Subsequently, the hypotheses will be tested according to the results.

4.1 Reliability Analysis

To make sure of the reliability of the items, the reliability was measured using Cronbach’s Alpha. This will ensure that the measures are consistent, that the scales are a coherent group.

Table 1: Cronbach’s Alpha coefficients for multiple-item scales

Mean Std. Deviation Cronbach's Alpha Conf 3.6601 1.39589 0.700 BB 3.9183 1.65757 0.917 SRCB 4.0103 1.27874 0.903

Tabel 2: Inconsistent vs. Consistent CSR

Cons Items Min. Max. N Mean Std. Deviation Cronbach's Alpha

Conf Incons 3 1.000 7.000 97 3.5636 1.35004 0.7 Cons 107 3.7477 1.43689 BB Incons 3 1.000 7.000 97 3.7045 1.55819 0.917 Cons 107 4.1121 1.72717 SRCB Incons 9 1.000 7.000 97 3.7801 1.23509 0.903 Cons 107 4.2191 1.28742

Regarding Consumer Confusion (Conf), the third item was recoded: “Het recentelijke nieuws over het bedrijf is tegenstrijdig met het beeld dat ik eerder had van het bedrijf.” The Cronbach Alpha of Consumer Confusion is 0.70, which is exactly the minimum norm of reliability. This means that the scale of Consumer Confusion used was acceptable. The scale of Buying Behavior (BB) is 0.917, which means that this item scale is also very reliable. The Cronbach’s Alpha for the moderator SRCB (SRCB) also has a high score of 0,883. The Cronbach’s Alpha’s of all the scales are above 0,7 therefore the reliability of the scales can be seen as acceptable, ensuring internal consistency.

The ‘one sample t-test’ tests if the means of the 4 scales is significantly different compared to the neutral result of 4. The result shows that this is the case for Confusion. So it can be

Confused consumers and inconsistent CSR : the effect of inconsistent CSR information on consumer confusion and buying behavior of customers that are socially involved