Treatise - Final

1 CHAPTER ONE: INTRODUCTION TO THE RESEARCH 1.1

Introduction

The purpose of this study is to investigate the cross-cultural differences in responses to consumer dissatisfaction in the city of Port Elizabeth of the Eastern Cape, with regard to the services sector. Port Elizabeth is one of the biggest cities in the Eastern Cape, South Africa and is a fastgrowing economic power in the region. Such rapid development has given rise to an influential segment of South African consumers (Stats SA, 2005). Consumer satisfaction is widely known as leading to goodwill, development of long-term relationships and profitability, and hence all companies should endeavour to make their customers satisfied with their service offering (1997, 1993, Blodgett, Wakefield & Barnes in Snellman & Wihtkari, 2003). Conversely, dissatisfaction can result in negative word of mouth, or exit 0behaviour, both of which are extremely detrimental to the company’s bottom line (1990, Singh in Phau & Sari, 2004). Consumer Complaint Behaviour (CCB) IS an important area of study for academics in order to provide recommendations to companies for effective service recovery. 1.2

Motivation for the Study

Research conducted by Liu and McClure (2001) showed that culture does have an impact on consumer

complaint

behaviour.

This

study

investigated

the

how

Hofstede’s

individualism/collectivism dimensions of culture impacted on complaint behaviour, and found that there was a strong relationship between the two. These researchers were also concerned about the lack of studies on consumer dissatisfaction and complaint behaviour in non-Western societies. They raised questions about the transportability of such Western studies to the growing consumer market in non-Western countries, and hence saw the need for such cross-cultural studies. The study carried out by Liu and McClure showed that many previous assumptions on CCB do not apply in non-Western cultures, and that it is both of theoretical and practical importance to study the occurrence of CCB in countries other than the U.S.A and Europe.

2 The aim of the researchers is to apply a similar approach to that of Liu and McClure (20001) to the community of Port Elizabeth in South Africa. Specifically, the purpose of this followup study is to investigate cross-cultural differences in the responses to consumer dissatisfaction amongst the community of Port Elizabeth, South Africa, in the services sector. 1.3

Focus of the Study

A thorough understanding of the various responses to customer dissatisfaction and a study of consumer complaint behaviour are essential knowledge for successful service recovery. Negative information has been shown to get more attention due to its unexpected nature and is hence more believable (1984, Richins in Phau & Sari, 2004). Jacoby and Jaccard (1981, in Phau & Sari, 2004, p. 407) state that, “in sufficient numbers, consumer complaints may stimulate regulatory action against a company and destroy a product, dilute or erode brand equity and in the process, the market share”. Dissatisfaction can however prove to be extremely beneficial to the service provider if handled sensitively and effectively. Samuel, Craig and Douglas (2006) state that culture has a deep impact on all facets of human behaviour. It is a profoundly complex phenomenon due to the nature of its impact, its manifestation in human behaviour and its intrinsic dynamic nature. The world is rapidly changing in terms of political, economic, social and technological forces, and so is the cultural landscape. Indeed, Triandis (1989, in Liu & McClure, 2001) have determined that “an individual’s values, concepts of self, perceptions of others and patterns of interaction with his/her environment are influenced by the “cultural meaning systems” in which they operate”. In the same manner, as a unique type of social behaviour, patterns of response to post-purchase dissatisfaction may also be impacted by culture as well. Hence, it is necessary to take into account the cultural context in which the consumer operates in order to develop more effective forms of marketing within the global economy. This view is corroborated by Kim, Forsythe, Gu, and Moon (2002) who state that personal values seem to be the underlying determinant of consumer attitudes and consumption behaviour. A thorough understanding of this research matter will enable companies will be able to appreciate the differences with which their diverse customers respond to dissatisfaction. Therefore, they will be able to adapt their strategy towards dealing with complaints more effectively, and thus improve brand loyalty and company image. Secondly, private action, which involves negative word of mouth, is a response to dissatisfaction, which is both

3 undesirable to the customer and the company. The customer may decide, in such a situation to switch brands, without the knowledge of the company. The company will thus be unable to provide a remedy to the disgruntled customer, and may experience a loss in profits. By identifying specific groups of the population who have a tendency to follow such a route, companies may be able to initiate various strategies to help these customers voice out their complaints in a more constructive way. The researchers believe that companies can improve their complaint management strategy by tailor-making these schemes to appeal to specific target markets. This study therefore seeks to segment the large consumer market of Port Elizabeth into culture-specific groups and understand their complaint behaviours to channel this information to organisations and relevant policy-makers. 1.4

Problem Statement

Research on complaint behaviour has been quite extensive during the past years; however, the demographics of consumer complaint behaviour in Port Elizabeth have not yet been properly studied. This is rather unfortunate, considering the fact that the new democratic South Africa is a plethora of diverse cultures, which therefore represents millions of multicultural potential consumers, whose responses to consumer dissatisfaction have not been researched. Companies have much to lose if they do not learn to appreciate the differences in which their customers, coming from different ethnic backgrounds, respond to dissatisfaction. This information is of vital importance for successful service recovery. These concerns pertain specifically to a rapidly developing city like Port Elizabeth, with a large services sector and emerging investments from multi-national companies. Unfortunately there has been sparse research carried on CCB in South Africa. It is hence necessary, at this point in time, to conduct a study of this nature. It is crucial to have reliable and valid data to broaden existing knowledge about how culture impacts on CCB, and provide practical insight to channel this information to Port Elizabeth’s corporate environment. 1.5

Research Objectives

This study will aim to: a.

Investigate the responses to consumer dissatisfaction.

b.

Ascertain the likelihood of the Afrikaans-, English-, and Xhosa-speaking consumers to engage in voice, private, or third party action when dissatisfied.

4 c.

Identify other related demographical variables affecting the likelihood of dissatisfied consumers to engage in voice, private, or third party action such as age, education level, and income level.

1.6

Expected Limitations of the Study

Unfortunately, due to limitations in finance and time, it will be difficult to obtain a truly representative sample for the purposes of this study. Also, only three main language groups are being addressed in this research. This excludes the other language groups in Port Elizabeth. The questionnaire will consist of scenarios to evaluate responses from the respondents. Therefore the latter will have to make use of their imagination to be able to provide realistic answers to the questions. Many respondents may have found it difficult to adequately imagine these scenarios and thus will find the questions tough to answer. The scenarios in the questionnaire may also not always apply to a specific individual. For example a respondent who has never been in the position of buying a lounge suite will find it difficult to respond to the questions. 1.7

Conclusion

This chapter provides a broad overview of the research background, problem statement and research objectives. Chapter two provides a conceptualisation of consumer dissatisfaction and CCB. The importance of CCB with regard to companies will also be discussed. Chapter three investigates in more detail the relationship between consumer dissatisfaction and CCB. The antecedents and determinants of CCB will be dealt with in chapter four. The discussion from the previous chapter is further detailed, with an examination of diverse variables impacting on CCB. Chapter five examines the unique individual characteristics of consumers that impact on CCB, for example assertiveness, personality, emotions and so on.

5 Chapter six is an overview of CCB research that has been conducted in the medical and financial services sector. It is necessary to include such a discussion as these specific services form the theme of the measure that is being used in the study. Chapter seven investigates the meaning of culture and its occurrence in marketing research. Most importantly, research conducted on the impact of culture on CCB will be dealt with, together with an examination of the main cultural groups in South Africa. The study’s research methodology will be described in chapter eight, together with an examination of the measures used and the hypotheses derived from the literature. A description of the population and sampling technique is also included, as well as the data collection procedure. In addition, the reliability and validity of the measure will also be examined. Chapter nine will report on the main findings of the study, and lastly Chapter ten will conclude the study by examining the implications of these findings, as well as the limitations of the study and the scope for future research.

6 CHAPTER TWO: CONCEPTUALISATION OF CONSUMER DISSATISFACTION AND CONSUMER COMPLAINT BEHAVIOUR 2.1

Introduction

This chapter is the first of the seven which make up the literature review. In this chapter, the concepts of consumer dissatisfaction and CCB are described. This chapter firstly describes how consumers get dissatisfied as well as the different responses to consumer dissatisfaction. These responses are analysed according to their behavioural and non-behavioural implications. The chapter also proceeds with a definition of consumer complaint behaviour together with a discussion of the differences between consumer dissatisfaction and CCB. The importance of CCB will also be dealt with. 2.2

Consumer Dissatisfaction

Warden, Liu and Huang (2003) refer to the use of gap research as the main model for understanding how consumers of services are satisfied. When expectations are not met, the resulting gap leads to dissatisfaction, and when expectations are surpassed, the result is satisfaction (1990, Zeithaml et al. in Warden et al., 2003). A consumer who is dissatisfied may choose several kinds of responses. Table 1 is an illustration of these responses: Table 1: Types of Responses to Consumer Dissatisfaction (Crié, 2003, para.2) Towards entity Response Type

Behavioural

Public

Private

(Sellers, manufacturers, official

(Family, friends,

organisations, associations, justice) Complaint

relations)

Legal action

Word of mouth

Return of the item

Boycott/leaving

Request for repair No action, with or without modification Non-behavioural

of the attitude Forget or forgive

7 Singh (1988, in Liu & McClure, 2001) used factor analysis techniques (both exploratory and confirmatory) and discovered that the responses to dissatisfaction could generally be considered to fall into one of three categories: voice responses, private responses, and third party responses. Voice responses are aimed at objects that are external to the consumer’s social circle and are directly involved in the dissatisfying exchange (e.g. seeking redress from the seller). Private responses are those in which the objects are not external to the consumer’s social circle and are nor directly involved in the dissatisfying experience (e.g. word of mouth communication or exit). Third party responses include objects that are external to the consumer but not directly involved in the dissatisfying transaction (e.g. reporting to a consumer agency or taking legal action). Crié (2003) points out three important dimensions within this framework of responses. Firstly, the terms ‘action’ and ‘responses’ should be distinguished from each other. The former refers more to a specific action; while the latter contains a few modalities that are not exclusively behavioural, specifically change of attitude or inactivity. The second dimension is embodied by the entities towards which the responses are directed to: the public one includes sellers, manufacturers and consumer associations or legal action; the private one involves family, friends and relatives (Crié, 2003). Thirdly, the responses demonstrate diverse intensities according to the two previous dimensions. Responses may fluctuate from inactivity to legal action – either simply to express dissatisfaction or to obtain repair or compensation. The reason behind heterogeneity of these responses may be partially attributed to the cause and intensity of dissatisfaction, and the nature and importance of the service and product concerned (Crié, 2003). However it may happen that the consumers mixes or connect several kinds of responses for the same dissatisfaction. 2.3

Consumer Complaint Behaviour

Consumer complaint behaviour has been widely proposed as a phenomenon resulting from the emotions involved in consumer dissatisfaction (1983, Bearden & Teal in Andreassen, 1998; Singh, 1988). The first definition that is most closely associated with CCB was stated at the end of the 1970’s by Jacoby and Jaccard (1981, in Crié, 2003): “an action begun by an individual who entails a communication of something negative to a product or service.” This definition has an important implication. The CCB phenomenon is thought to be set off by some feelings or emotion of dissatisfaction. Hence, consumers cannot qualify for CCB without any perceptions of dissatisfaction (Singh, 1988). Day and Landon (1977, in Crié,

8 2003) suggest that it is the consequence “of a given act of consumption, following which the consumer is confronted with a situation generating a high dissatisfaction, of sufficient impact so that it is, neither likened psychologically nor quickly forgotten”. Fornell and Wernerfelt (1987, in Crié, 2003) believe that the complaint is “an attempt of the customer to change an unsatisfactory situation”. Oliver (1980, in Phau & Sari, 2004) further points out the link between CCB and the expectancy disconfirmation theory, which states that the dissatisfaction leading up to CCB is a product of the gap between the consumers’ expectations and the actual performance of the service. This finding has been corroborated by Day (1977, in Phau & Sari, 2004). The researchers believe that it is necessary for firms to avoid over-inflating their marketing promotions, such that consumers do not construct unrealistic expectations and engage in CCB when those are not met. Singh (1988) suggests that this behaviour, triggered off at an emotional or sentimental level, forms part of a more general framework of responses to dissatisfaction, consisting of two dimensions. The first dimension is behavioural and is based entirely, or in part, in actions initiated by the consumer (conveying expression of his/her dissatisfaction not only to the seller, but also to third parties, friends or relations). However, this does not necessarily require action towards the company. Singh (1988) iterates that it is essentially within this dimension that CCB should be considered. The second dimension involves the inaction by the consumer whereby he/she forgets a general episode of dissatisfaction. 2.4

Importance of Consumer Complaint Behaviour for Companies

The nature and extent of CCB in an industry appears to affect social and customer welfare (Warren & Gilbert, 1993). Indeed, firms usually consider consumer complaints to be indispensable indicators of unsatisfactory quality and performance (Phau & Sari, 2004; Dolinsky, 1994). Consumer complaints have significant repercussions on societal issues as well as strategic decisions for organisations. Understanding why customers are dissatisfied with a particular purchase forms a crucial basis for improvement in a firm’s market offering (Warren & Gilbert, 1993). McCole (2004) further emphasises upon the importance of retaining existing customers and attracting additional ones as an issue of great significance within the marketing field. Warren and Gilbert (1993) stress that consumer satisfaction is an essential factor in repeat purchase behaviour. When a dissatisfied customer seeks redress, this provides the service provider with

9 an opportunity to resolve the situation. If done successfully, this can improve the company’s image via the formally disgruntled customer, and lead to enhanced customer loyalty (2002, Crie & Ladwein in Phau & Sari, 2004). The company also benefits by learning from complaints how to prevent similar service/product failures in the future (1997, 1993, Blodgett, Wakefield & Barnes in Snellman & Wihtkari, 2003). According to Ndubisi and Ling (2006; Liu & McClure, 2001), the cost of gaining a new customer can prove to be as high as five to six times the cost of retaining an existing one. Therefore, marketing managers must evaluate the reactions of consumers after the sale in order to implement successful strategies. There should be equal emphasis on retaining current customers in addition to taking up new ones, that is, the use of defensive marketing strategies (McCole, 2004; 1987, Fornell & Wernerfelt in Dolinsky, 1994). This strategy is referred to as relationship marketing, where the primary focus in on customer retention and loyalty (1995, Bitner in While the source of customer dissatisfaction represents critical information that may prevent errors, a thorough understanding of how customers react when they are dissatisfied is also of central importance (Warren & Gilbert, 1993). Furthermore, Singh (1988) states that the study of CCB and its consequences seem to be critical in the explanation and the prediction of consumer repurchase intentions and loyalty. In addition, in terms of macro marketing, the nature and extent of CCB prevailing in an industry seems to have an influence on customer and social welfare. Crié (2003) says that, within the framework of the relationship paradigm, complaint behaviour is a potent signal that companies should take into account. If the organisation reacts appropriately, it offers the organisation a last opportunity to lock in the customer. It is also a legitimate and ethical act towards the consumer. A significant way in which companies may retain their current customers in the face of service failure is to display sincere commitment towards them (McCole, 2004). Dolinsky (1994) specifies that, though effective complaint handling is an important activity for all organisations, it is even more critical in the services industry. This is because services are highly variable compared to steady product lines, resulting in greater probability of perceived problems and customer complaints. The researcher further emphasises that customer complaint handling is vital for organisations that operate in highly volatile and competitive environments, as it can be quite difficult to attract new customers and it is more cost-effective to retain existing ones. The following conceptual model illustrates the

10 importance of effective complaint management and how it is a sound financial practice (Johnston, 2001).

Figure 1: Conceptual Model of Complaint Management (Johnston, 2001, p.62) 2.5

Conclusion

This chapter provided definitions of two of the most important concepts being considered in the treatise, that is, consumer dissatisfaction and consumer complaint behaviour. The behavioural as well as the non-behavioural responses were explained and the concept of consumer complaint behaviour was analysed within this framework of dissatisfaction. Chapter three will further investigate the relationship between consumer dissatisfaction and consumer complaint behaviour. The different responses to consumer dissatisfaction and complaint behaviours will also be explained.

11 CHAPTER THREE: EFFECT OF CONSUMER DISSATISFACTION ON LIKELIHOOD OF CCB 3.1

Introduction

Dissatisfied consumers may not always choose to complain. It is hence necessary to consider what the factors that surround such a phenomenon are. Another important dimension to be considered is that the reasons why consumers complain impact on their choice of channels. This chapter investigates these concepts and seeks to go broadly into explaining the link between consumer dissatisfaction and consumer complaint behaviour. Responses to consumer dissatisfaction and consumer complaint behaviour are discussed as well as a classification of the consumers according to their channel of choice. 3.2

Propensity of Consumers to Complain

The propensity for a consumer to complain depends on a number of factors, including the personality traits of that individual, the perceived likelihood of success, consumer’s attitude towards complaining, service importance and stability and controllability of the incident, have an impact on the choice of whether to complain to the service provider or not (1995, Blodgett et al. in Snellman & Wihtkari, 2003; Singh, 1988). In fact, certain individuals are assertive and seek redress whenever they are dissatisfied with a service, while others are hesitant to complain, independent of the severity of the dissatisfaction endured. Also, it is less probable for dissatisfied consumers to complain if they suspect that the service provider will not endeavour to offer a refund or otherwise resolve the situation. Similarly, dissatisfied customers who depend minimally on the service or product may be more disinclined to complain (1995, Blodgett et al. in Snellman & Wihtkari, 2003). These findings have been confirmed in a study carried out by CCCTF (1995, in Coyle & Williams, 1999) on consumer dissatisfaction and propensity to complain in the medical sector. Though, customers do at times complain when there is no legitimate reason to do so, it may be said that “consumer dissatisfaction is positively and significantly related to CCB” (Singh & Pandya, 1991). It has also been shown that complaining is a form of emotional release (Nyer, 2000). Mattila and Wirtz (2004) have found that aggrieved customers, whose principal goal is to vent their frustration so as to feel better, neither expect nor rely on a reply from the firm. Venting is a “fire and forget” situation, where the complainer may want to remain

12 anonymous,

to

reduce

embarrassment

and

avoid

a

potentially

nasty

(unpleasant)confrontation with the firm’s front-line staff. It has been shown that these complainers are more likely to use remote channels such as a letter or e-mail. Complaining may also be aimed at seeking redress or warning potential consumers (Nyer, 2000; Mattila & Wirtz, 2004). Mattila and Wirtz (2004) have demonstrated that this type of complainer is likely to voice his or her complaint face to face or over the telephone, probably because this type of real-time interaction enables the complainant to clarify matters, explain in detail, observe the other party’s body language and even show anger, frustration and urgency that may lead to faster resolution of the problem. 3.3

Responses to Dissatisfaction and Consumer Complaint Behaviour

CCB should be rather thought of as a process, that is, its ultimate expression does directly depend on its initiating factors but on the evaluation of the situation by the consumer, and of its evolution over time (Crié, 2003). Crié (2003) explains that the notion of ‘complaint behaviour’ includes a broader terminology including the notions of protest, communication (word-of-mouth) or counsel to third parties, even the notion of boycott. The latter is conceptually involved in a set of explicit demonstrations of the consumer’s dissatisfaction, usually directed at the organisation. The following Figure 2 is a representation of the various heterogeneous responses to dissatisfaction, a study of which is essential to understanding CCB.

13

Figure 2: Responses to Dissatisfaction and Complaint Behaviour (Crié, 2003, para. 3) The above typology shows that there are four broad ways in which dissatisfied consumers can take action. Firstly, dissatisfied consumers can take no actions following bad buying experiences. Doing nothing or not repurchasing a firm’s product or services are legitimate responses to dissatisfaction (1984, Day in Phau & Sari, 2004). Singh (1990, in Phau & Sari, 2004) classifies consumer engaging in such behaviour as “passives”. Secondly, dissatisfied consumers can take some form of private action (1990, Singh in Phau & Sari, 2004). Private actions refer to actions involving only people inside the consumer’s group in informal ways. This may include changing the brand/supplier, ceasing to use the product or service, or warning family and friends. Private actions are more likely to be driven by “getting even and punitive aims”. This group is classified as “voicers”. As stated before, negative word-of-mouth can be extremely detrimental to a company as such communication exerts on powerful influence on consumer purchasing behaviour, affecting both short-term and long-term judgments (1991, Murray in Mangold, Miller & Brockway, 1999). Consumers have been found to rely on word-of-mouth communication to lower the

14 level of perceived risk and the uncertainty that are often associated with service purchase decisions. In comparison to the purchasers of goods, it was found that service purchasers have greater confidence in personal sources of information as well as a greater pre-purchase preference for personal information sources. Richins (1983, in Mangold et al., 1999) found that the tendency to engage in negative word-of-mouth was positively related to the level of dissatisfaction and negatively related to the consumer’s perception of the retailer’s responsiveness to complaints. The reason why word-of-mouth seems to be so compelling to consumers is because of its vividness as a form of communication. It seems that dissatisfied consumers often establish the stability and controllability of the problem (1984, Folkes in Snellman & Wihtkari, 2003). The researcher defines stability as being related to the consumer’s perception of the probability of similar incidents occurring in the future, while the controllability of the problem relates to the consumer’s evaluation on how easily the problem could have been prevented. Blodgett et al. (1995, in Snellman & Wihtkari, 2003) propose that consumers, who believe that the problem has a high probability of occurring in the future, and that it could have been prevented, are likely to switch to another service provider instead of embarking on a complaint process. Thirdly, dissatisfied consumers can engage in some form of public action (1990, Singh in Phau & Sari, 2004). Public actions involve people and organisations outside the consumer’s group in more formal ways. These may be seeking redress directly from the seller or manufacturer and taking legal action against the seller and manufacturer. They may also be registering a complaint with the seller and manufacturer, a public consumer protection agency, or a private consumer organisation. Complainers may also create a new company to provide a better product or service. The main purpose of these consumer complaints is to “recover economic loss by getting an exchange or a refund and rebuild self-image”. This group is classified as “irates”. Finally, consumers may take a variety of different private and public actions (1990, Singh in Phau & Sari, 2004). Consumers may blame sellers and manufacturers for their unsatisfactory product or service. They may select to boycott sellers and manufacturers by ceasing to use their products or services and spread negative information about their products or services. This last group is classified as “activists”.

15 Consumers who engage in public actions offer are more benefits to companies than those who take private actions since public actions not only make companies aware of the problems they face, but also provides an opportunity for them to retain their customers (Phau & Sari, 2004). It is important for companies to retain their current customers as they are likely to show their loyalty by using a variety of the companies’ products or services. Their positive word of mouth will attract new customers and there is greater likelihood of them feeling sympathetic when companies face a problem such as product or service failures. Their influence may also spread to the non-complaining silent majority, by a form of osmosis or memetic diffusion. 3.4

Conclusion

This chapter considered the factors that push dissatisfied consumers to complain and also the channel of choice of dissatisfied consumers according to the reason why they set out to complain. The framework of responses to consumer dissatisfaction and consumer complaint behaviour was also explained. It was found that there are four types of complainers, classified according to their responses to dissatisfaction. Chapter four shall go into explain greater detail about the factors that impact on the types of responses that dissatisfied consumers choose to take, that is, the determinants of CCB. The other concepts that will be discussed include the modulating factors of CCB.

16 CHAPTER FOUR: DETERMINANTS AND ANTECEDENTS OF CCB 4.1

Introduction

Research has attempted to investigate why some people tend to voice their dissatisfaction while others do not. Three spheres of factors are identified as moderating CCB and each will be described in detail. The four entities involved in CBB are the suppliers, the product or service, the customer, and the episode of dissatisfaction. Usually, consumers need to be dissatisfied in order to complain but other variables are necessary to switch from dissatisfaction to complaint. Such variables may lie in attribution of the cause of dissatisfaction or in psycho-sociological consumer characteristics. These are the initiating factors of CCB, while the modulator factors refer to whether the service or product purchased is delayed or immediate. This will also influence CCB in various ways, especially in the case of a delayed experience, where various variables can intervene to modify the consumer’s response, as will be described below. 4.2

Factors of CCB

It has been found that CCB is a complex phenomenon that is impacted upon by diverse factors in the choice of a particular complaining action (1984, Dearden & Mason in Phau & Sari, 2004). Crié (2003) articulates that it is possible to distinguish within the literature related to CCB, between three spheres of factors which interact with different weights to lead eventually to complaint behaviour or other modalities of response to dissatisfaction such as defined in Figure 3 below.

17 Figure 3: Antecedents and determinants of CCB (Crié, 2003, para. 6)

The psychological sphere is made up of characteristics reflecting an individual’s propensity to engage in complaint behaviour. The economic sphere groups together elements of cost and exchange structures. Lastly, the ethical sphere incorporates transactional equity, focusing on the perception of the value of the link with the company as well as the accuracy and helpfulness of the information given, for example, regarding ways in which to resolve disputes. According to Crié (2003), the variables of the ethical sphere work as initiators and those of the economic sphere have a modulator role, while the variables of the psychological sphere can be classified as either type. Four entities or actors are described as being directly involved in CCB: the product or service, the customer, the supplier and the episode of dissatisfaction. These various entities allow the various antecedents and determinants of CCB to be re-ordered, as Table 2 shows.

Table 2: Various determinants of CCB (Crié, 2003, para. 7)

18

4.3

The Initiating Factors of CCB

Several factors may lead to CCB which allow a better understanding of consumers’ motivation for CCB and, on the other hand, a forecast of what response will be the most likely to be adopted. Usually, consumers need to be dissatisfied in order to complain but other variables are necessary to switch from dissatisfaction to complaint. Such variables may lie in attribution of the cause of dissatisfaction or in psycho-sociological characteristics of the individual consumer (Crié, 2003). 4.3.1

Dissatisfaction

According to Singh (1988), dissatisfaction is a necessary antecedent of CCB, but is often not sufficient. It is described as being the activating factor of the process. Oliver (1987, in Crié, 2003) underlines the existing relationship between the intensity of the dissatisfaction and complaint behaviour. As the dissatisfaction increases, the verbal complaint strengthens and the probability of action grows. Singh and Pandya (1991) state that the relationship intensity of response is not linear and that it has certain threshold effects. According to these authors,

19 when the level of dissatisfaction exceeds a given threshold consumers tend to use either negative word of mouth or leaving or appeal to a third party. 4.3.2

Attribution of the cause of the dissatisfaction

For dissatisfaction to lead to CCB, according to Crié (2003), the consumer has to identify clearly the party responsible for his or her dissatisfaction during a given consumption episode. In various situations, it is the consumer himself or herself, for example when he or she judges that he or she did not make the right choice. Therefore, the author argues that even if the attribution of responsibility for dissatisfaction operates as an intermediary between the confusion and the response which follow, it is suitable to classify it as an initiating factor of CCB. A success or failure can be attributed either to elements under the control of the individual, that is internal causes, or to environment or situational factors (external causes). On the other hand, Crié (2003) states that the performance can be attributed to invariant factors, which are stable, or factors evolving with time, which are unstable. The type of attribution achieved by the customer determines the response that may be chosen. Usually consumers who perceive the cause of their dissatisfaction as being stable, that is the same problem may recur, or controllable, where the individual thinks that the retailer could have prevented the problem, are more inclined to leave the product or the firm and to engage in negative word of mouth than those who think that the problem has little chance of recurring and that the supplier could not prevent it. 4.3.3

Psycho-sociological factors

Some individual characteristics can be considered as initiators of CCB, although they can also play a modulating role according to the situation, as can be seen from Figure 4. Numerous researchers associated CCB with frustration. Frustration can arise for various reasons and generally, the more substantial the frustration, the greater the risk of aggressiveness and CCB. The stress of dissatisfaction adds to daily stress and CCB is connected to a double evaluation of the situation during a cognitive process. Three components form the primary evaluation of the stress: (1) the level of modification of the individual objectives; (2) the incongruity between these objectives and the incident; (3) the level of the individual’s ego infringement (humiliation, self-esteem, ethical values). A second evaluation of possible response strategies

20 is then formulated according to: (1) the responsibility attribution for the confusion; (2) the possibility of solving the problem (capacities of the individual and probability of success); (3) the feeling that things will go better afterwards. If the resulting stress of the primary evaluation can be reduced by a solving strategy, the probability of CCB is much greater. If, on the other hand, this strategy risks increasing the initial stress, the main probability turns out to be a non-behavioural response or a behavioural response towards the market (refer to Figure 2). 4.4

Modulator Factors of CCB

Dissatisfaction can be organised around two congruent factors: the first is situational and the latter is temporal. Sometimes the product is purchased or the service is delivered and it is immediate, but it can also take place at a distance from the act of purchase. Similarly, the response to dissatisfaction can be either immediate or deferred. So the CCB can consist of a short path in the case of dissatisfaction on the spot and an immediate response, or a long course in the case of delayed dissatisfaction with regard to the purchase and a postponed response. It is especially in this last situation, that numerous variables can intervene to modify the consumer’s actual response, by moderating or aggravating it. The market structure, socio-cultural characteristics or evaluation of the various costs associated with CCB will act as modulators of the process outcome (Crie, 2003). 4.4.1

The Market Structure

Crie (2003) articulates that in a restricted competitive environment, verbal action is the only possible action for a dissatisfied customer. However, when competitors are numerous, the consumer’s leaving becomes the most likely reaction and this action is then situated at the market level (see Figure 2).Hence, the market structure appears as a powerful determinant of CCB. On the other hand, and within this framework, sensitivity of the customer segments to price or quality shapes the dominant response type. Those sensitive to price may leave the company while those sensitive to quality are more inclined to complain. Also, it seems that buyer–seller interaction frequency (or purchase rate) plays a part in the preference for verbal action. The more frequent these interactions the fewer the public actions. Barksdale et al. (1884, in Crie, 2003) found that the lower the level of purchase at the same supplier, the

21 greater the tendency to engage in CCB. New purchases lead to more complaints and it seemed that the standard suppliers receive more complaints than the new ones. 4.4.2

Socio-cultural Factors

These factors relate to education, income and ethnic identity that will be discussed in section 5.3. 4.4.3

The Costs of the Complaint

Landon (1977, in Crie, 2003) states that the profit of a complaint is a function of the result minus the cost of complaint. The result is itself estimated with regard to the importance and the nature of the damage sustained. The customer’s preference for verbal action is then related to the expected value of the complaint outcome (which is connected with the importance of the dissatisfaction) minus the associated costs. CCB requires time and effort and money, thus generating various types of costs, such as the opportunity cost of the elapsed time, deliberation costs and transportation costs. These costs are specific to the envisaged response type whether it is actual or not. The costs cannot be paid off afterwards and can at best be used as a learning stage during a later complaint, reducing later costs. Uncertainty arises from any complaint and the consumer looks for information to reduce this uncertainty to a bearable level. Furthermore, the complaint frequency has a direct influence on its organisation, perhaps becoming a ‘routine’, thereby reducing the associated costs. This theory explains more frequent CCB when consumers have superior educational level; they know their rights so the level of uncertainty associated with the complaint is reduced and hence there is a global decrease of the perceived costs of a complaint. As the search for a new product or supplier generates several costs, it is often the case that the dissatisfied customer refrains from any action (Crie, 2003). 4.5 Conclusion From the research presented in this chapter, it is clear that there are diverse variables that can influence how and if consumers will choose to complain after a dissatisfactory experience. Characteristics influencing CCB relating to the specific individual were described and the four entities involved in CBB were identified. These are the initiating factors of CCB while

22 the modulator variables of CCB explain that dissatisfaction can be organised around situational and temporal factors. These variables will have a large part to play, with a delayed experience or purchase. The next chapter will specifically focus on and examine in much more detail all those unique individual characteristics of consumers moderating their complaint behavior.

23 CHAPTER FIVE: CHARACTERISTICS OF COMPLAINERS AND NONCOMPLAINERS 5.1

Introduction

As was detailed in the previous chapters, failure to complain is indicative that dissatisfaction is necessary, but often not a sufficient cause for complaining. Researchers have tried to explain this phenomenon in terms of variations in market structures, product and service characteristics, costs and benefits of complaining as well as characteristics of consumers. This chapter will examine such unique characteristics of consumers related to CCB in greater detail, referring to customer attitudes to complaining, assertiveness, personality and emotions and demographic and socio-economic variables as influencing factors. 5.2

Personality Characteristics

Harris and Mowen (2001, in Bodey and Grace, 2006) emphasise that it is increasingly being recognised that personality accounts for much more variance in behaviour and intentions than has been noted in past literature. Bodey and Grace (2006) state that complainers can be differentiated from non-complainers based on three personality characteristics. These are: perceived control, Machiavellianism, self-efficacy and self-monitoring. Perceived control is defined as “the expectation of having the power to participate in making decisions in order to obtain desirable consequences and a sense of personal competence in a given situation” (1990, Rodin in Bodey & Grace, 2006). Skinner (1996, in Bodey & Grace, 2006) drew a distinction between individuals who believe they have a high degree of control as opposed to those individuals who believe they have a low level of control over their environment. Such persons are often referred to as individuals with an internal locus of control and those with an external locus of control (Robbins & Judge, 2005). Individuals with an internal locus of control are generally considered to exercise greater effort to achieve goals by taking action; are highly motivated and display attributes of competence in situations. Therefore, Bodey and Grace (2006) suggest that individuals who maintain a high level of perceived control would be more likely to complain than those who do not. In the event of service failure, for example, it would be expected that such persons would engage in complaint behaviour as a means by which to take control of the situation, therefore

24 accomplishing their goals. On the other hand, individuals with an external locus of control are often described as being passive and fearful and, thus, would not be expected to engage in complaint behaviour. Research by Bodey and Grace (2006) has supported this finding. According to Robbins and Judge (2005), those who are considered to be Machiavellians tend to engage in more manipulative behaviour by using indirect, irrational tactics such as exploitation and persuasion to gain an advantage. As Machiavellianism has been found to influence other forms of consumer behaviour such as purchasing behaviour, sociability, bargaining, ethical behaviour and interpersonal interactions, it may also assist in explaining consumer complaint behaviour. Bodey and Grace (2006) suggest that Machiavellians may view complaining as a mechanism to be utilised in order to get what they want. However, no research up to date has been able to establish a correlation. On this basis, it may be that situational factors, such as the severity of the situation or the degree to which the benefits of complaining outweigh the costs, impact on the relationship between Machiavellianism and complaint behaviour. Similarly, researchers have not been able to link self-efficacy to complaint behaviour. Selfefficacy refers to the belief in “one’s capabilities to mobilise the motivation, cognitive resources and courses of action needed to meet given situational demands” (1989, Wood & Bandura in Bodey & Grace, 2006). Luthans and Peterson (2002, in Bodey & Grace, 2006) report that the higher a person’s self-efficacy, the more persistent he or she is when faced with setbacks or failure. Therefore one might expected that individuals with low self-efficacy would tend to avoid complaining behaviour rather than engage in it; however, once again, it may be that situational factors impact the probability of individuals with high self-efficacy actually engaging in complaint behaviour and therefore a correlation could not be established. According to Robbins and Judge (2005), self-monitoring refers to an individual’s ability to adjust his or her behaviour according to external situational factors. Such individuals are said to be highly sensitive to external cues and can behave differently in different situations. On the other hand, Browne and Kaldenberg, (1997, in Bodey & Grace, 2006) state that those individuals who are considered low self-monitors are relatively insensitive to social cues and tend to present themselves in the same light in every situation As low self-monitors mostly rely on dispositional information they tend to behave in a manner that is conducive to their inner beliefs and feelings of the self. Most research to date examining self-monitoring has

25 done so in relation to products such as clothing, alcoholic soft drinks and job performance. In the consumer behaviour context, it has been found that self-monitoring is related to advertising, decision-making, purchasing and consumption, materialism, and interpersonal influence. According to Richins (1981, in Bodey & Grace, 2006), regarding complaint behaviour, individuals who are said to be socially anxious are expected to complain less often. In fact, a study by Kowalski and Cantrell (1995, in Grace and Body, 2006) found moderate negative correlations between complaining and social anxiety. Based on these findings, selfmonitoring may well be an important variable in understanding the differences between complainers and non-complainers. Research by Bodey and Grace (2006) also found that, high self-monitors tend to be noncomplainers whereas low self-monitors are more likely to be complainers. This supports the finding of Keng et al. (1995, in Bodey & Grace, 2006) who also found that high self-monitors are generally relatively conservative; have negative attitudes towards complaining and view complaining less as a social responsibility. Thus such individuals may shy away from complaining in their desire to maintain a socially desirable image of themselves. This is related to research by Lerman (2006) who refer to the politeness theory in explaining complaint behaviour. According to this theory, Brown and Levinson (1987, in Lerman, 2006) distinguish between two kinds of face: positive and negative. Goffman (1967, in Lerman, 2006) defines face as “the positive social value a person effectively claims for himself.” Positive face concerns the need for approval and it recognises that everyone has a self-image while negative face concerns the desire to be unhindered in one’s actions. The author argues that politeness as an interaction style must be studied with respect to consumer complaining behaviour because complaining by its nature is a face-threatening act. The degree to which a consumer does not want to insult another, relates to the degree that he or she will not engage in complaining behaviour. In his research, Lerman (2006) found that there is indeed a negative relationship between politeness and complaining behaviour. 5.3

Non-personality Variables

Lastly, Bodey and Grace (2006) also discuss an individual’s attitude in relation to complaint behaviour. Many researchers have linked attitude towards complaining to individual

26 confidence, perceived risk and attitude towards the situation. This means that an individual’s reluctance to complain is often associated with a lack of confidence, risk perceptions concerning public complaining and being put in the uncomfortable position of making a complaint. One such study by Blodgett et al. (1995, in Bodey & Grace, 2006) revealed that consumers who perceived a high likelihood of success in taking action, such as complaining, maintained a more positive attitude towards complaining and therefore were less likely to engage in negative word-of-mouth communication. Furthermore, various researchers, such as Keng et al., (1995, in Bodey & Grace, 2006), revealed that attitude towards complaining was linked to an individual’s intentions or behaviour to complain. This means that generally consumers with a more positive attitude towards complaining have a greater propensity to complain and hence, engage in complaint actions more frequently. It would thus be expected that complainers and non-complainers would differ on the basis of their attitude toward complaining and this is indeed supported by research conducted by Bodey and Grace (2006). Their finding supports that of others such as Bearden and Crockett (1981) and Keng et al. (1995) who found that those consumers who have a more positive attitude towards complaining are more inclined to complain. Similarly, the results also suggest that a negative attitude toward complaining may well impede noncomplainers from taking complaint action (Bodey & Grace, 2006). Im, Kim, Kim, Kim and Shin (2003) refer to the influence of prior complaint experience impacting upon an individual’s attitude toward complaint behaviour. Prior complaint experience can be conceptualised as a consumer’s tendency to appeal in the case of dissatisfaction. According to Singh and Wilkes (2006, in Kim et al., 2003), the extent, which may be frequent or infrequent, of past complaining experiences can reinforce a customer’s attitudinal and behavioural dispositions in future situations. It is proposed that as consumers learn about the mechanisms, options and positive outcomes of previous complaint experiences, they develop more positive attitudes with regard to complaining. In a study by Kim et al. (2003) it was found that a greater level of complaint behaviour does indeed lead to a more positive attitude toward complaining. However, the study indicated that the level of prior complaint experience does not influence the perceived value of the complaint or the perceived likelihood of a successful complaint.

27 5.4

Demographical Characteristics of Consumers

Morganosky and Buckley (1986, in Tronvoll, 2003) examined the correlation between demographic variables and complaint behaviour. Analysis revealed that higher income and better educated consumers were significantly more likely than lower income and less educated consumers to partake in complaint action. Other demographic variables such as age, employment status and family type did not reveal significant differences in this study. Consistently, Crosier et al. (1999, in Tronvoll, 2003) and Hogarth and English (2002, in Tronvoll, 2003) found that characteristics that complainants have in common to be above average education and income. Bearden et al. (1980) and Grønhaug and Zaltman (1980, in Tronvoll, 2003) have put forward theoretical reasons for complainers tending to belong to higher socio-economic groups. They have suggested that people who belong to these groups have greater resources in the form of self-confidence and information in dealing with marketplace problems. These consumers also tend to perceive less of risk of humiliation in complaining, they have more purchasing power and tend to buy better quality products. Tronvoll (2003) articulate that the notion that complainers typically belong to upper socioeconomic groups in society was established in the1970s and 1980s. However, the author’s present study did not confirm this pattern. Indeed, it was found that complainers came predominantly from lower socio-economic groupings. A change in complaint behaviour over time is a possible explanation for these findings according to Tronvoll (2003). The author also refers cultural issues as another possible explanation for the present findings. Various researchers, such as Furrer et al (2000, in Tronvoll, 2003) and Hui and Au (2001, in Tronvoll, 2003) have noted that cultural differences, such as individualism and collectivism, affect perceptions of service quality and complaint behaviour . Furthermore, managerial or professional status was also found to increase the likelihood of complaining by Crosier et al. (1999, in Tronvoll, 2003) while Hogarth and English (2002, in Tronvoll, 2003) found it to be more likely that complainants are female, and more probable to engage in voice complaint behaviours. Interestingly, Hogarth and English (2002, in Tronvoll, 2003) also concluded that race had a part to play. Their findings revealed that when it came to written complaints regarding the banking services in the United States, complainers were slightly more likely to be non-white than Caucasians.

28 5.5

Conclusion

Personality characteristics such as perceived control, Machiavellianism, self-efficacy and self-monitoring may have a part to play in CCB, although it is often difficult determining the influence of these variables, as situational factors may also impact the probability of individuals engaging in complaint behaviour. Hence a correlation cannot always be established and research studies may often contradict each other. Furthermore, non-personality variables are discussed, such as an individual’s attitude in relation to complaint behavior, individual confidence, perceived risk and attitude towards the situation. The correlation between demographic variables such a higher income and better educated consumers were established by some studies while others found no relationship. Other demographic variables such as age, employment status and family type did not reveal significant differences. What follows is a description of CCB in the service sector, specifically focusing on the medical and financial sectors.

29 CHAPTER SIX: CONSUMER DISSATISFACTION IN SERVICES SECTOR 6.1

Introduction

The following chapter considers research that has been carried out on consumer dissatisfaction and CCB in the medical and financial sector. These sectors are specifically being discussed as they form the theme of the scenarios that make up the questionnaire. Section 6.2 considers the reasons why patients often experience dissatisfaction in the medical sector, together with a discussion of the factors that may impact on the propensity of these dissatisfied consumers to complain in such situations. Then in section 6.3, consumer dissatisfaction in the financial sector is examined, and possible reasons for consumers’ apparent lack of engaging in complaint behaviour discussed. 6.2

Consumer Dissatisfaction in the Medical Sector

Research carried out on complaint behaviour in the medical field showed that many patients who experienced problems with health care failed to complain (1995, CCCTF in Coyle & Williams, 1999). This study identified a number of factors that contributed to this phenomenon, namely perceptions of powerlessness, fear of retribution, emotional costs of complaining and feelings of gratitude. The primary reason for not complaining was quoted as being people’s perceptions of their lack of power (that is their inability to change the practitioner’s behaviour or the unresponsiveness of organisations). Some people may perceive that they will suffer from some form of reprisal after complaining, for example be unable to find another health practitioner after complaining about a current one. In fact, these feelings of retribution seem to be strongest in certain vulnerable groups, such as the elderly, the disabled, and the mentally ill, who may be at a disadvantage due to the nature of their critical needs (1992, Mulcahy in Coyle & Williams, 1999). Gratitude has also been identified as a factor in affecting patients’ propensity to complain in the face of dissatisfaction, for example Altschul (1983, in Coyle & Williams, 1999) discovered that patients felt grateful for the care provided by nurses and hence made allowances for the nurses’ shortcomings by deflecting responsibility for poor service delivery to lack of resources or to the system in general. There also seems to be a certain stigma attached to the act of complaining which reduces the propensity to do so in some households (1994, Mulcahy & Tritter in Coyle & Williams, 1999). This reluctance to complain may be

30 observed in the way that certain patients are concerned about the image they send forth when complaining; they feel the need to justify extensively the fact that they are complaining and establish themselves as worthy complainants. Hence, Coyle and Williams (1999) state that studies of complaint behaviour and qualitative studies of patients’ viewpoints have shown that there exist some cultural biases towards criticising health care practitioners. Along with the reluctance to complain in such instances, there also exists the strong need to justify one’s complaints with a worthy cause, such as endeavouring to improve the quality of service delivery for third parties. 6.3

Consumer Dissatisfaction in the Financial Sector

The complexity, variability and uncertainty of the performance of many financial services providers make consumer evaluation difficult, and have accounted for many purchases failing to meet consumers’ expectations. However, since many financial services are purchased infrequently; many consumers are inexperienced and unfamiliar with complaint procedures and behaviour (Davies, 1996). According to Zeithaml (1981 in Lopez, Hart and Rampersad, 2007) due to the fact that most services are intangible, definitive quality measurement systems are challenging because their delivery is performance based (Zeithaml, 1981). Parasuraman et al. (1985, in Lopez et.al 2007) published a well-known study on customer satisfaction in the service sector in which they identified ten determinants of service quality that contribute to customer satisfaction. The authors found that ten criteria were used by customers when evaluating service quality, and these were reliability, responsiveness, competence, access, courtesy, communication, credibility, security, understanding, and tangibles dimensions. From subsequent studies, it was found that customers’ perceptions of service quality and resulting levels of satisfaction are linked to individual tolerance levels. Parasuraman et al. (1985, in Lopez et al., 2007) concluded that if a service is provided within an acceptable quality range, a customer will likely be satisfied with both the service and the service provider. According to Lopez et al. (2007), in an analysis of the United States and Latin American private banking sector, financial institutions could only achieve high levels of customer satisfaction by providing consistently high levels of reliability. In South Africa, Bick et al. (2004, in Lopez et al., 2007) also discovered that customers perceived reliability, which was

31 referred to as operational excellence, to be the most important contributor to consumers’ satisfaction. Hirschman (1970, in White & Yanamandram, 2004) contends that some customers react to dissatisfaction passively, preferring to remain with a service provider in the belief that the probability of an improvement outweighs the cost of searching for another firm. Therefore loyalty is one of the reactions a customer may have to a service failure. The findings from a study conducted by Levesque and McDougall (1993, in White & Yanamandram, 2004) found that even in those cases where a problem is not solved, about half of the respondents will remain with the financial firm. Accordingly, Day (1984, in White & Yanamandram, 2004) articulates that a majority of customers do not undertake any action following a negative service experience. Colgate and Lang (2001) refer to various reasons for lack of behaviour in the financial sector, specifically focusing on the banking segment. The authors refer to the “relationship investment barrier”, which is associated with loyalty, confidence in the provider, receiving “the best deal” as well as being known by the bank or the financial organisation’s staff. The second barrier factor identified, negativity, captured issues such as being locked in to an organisation and the financial costs and or uncertainty associated with changing providers. The apathy factor related to participants’ perceptions that changing involved too much time and effort and that all banks or financial institutions were similar in any case. The fourth barrier factor described was service recovery, and this factor was present when a complaint had been satisfactorily resolved. Hence, in essence, service recovery can compensate customers for a dissatisfactory experience. 6.4

Conclusion

This chapter discussed research carried out on consumer dissatisfaction and CCB in the medical and financial sector. It was found that there a number of factors that impact on patients in the medial sector and customers in the financial sector’s unwillingness to complain even when they are dissatisfied. Some of these factors in the medical sector include are perceptions of powerlessness, fear of retribution and feelings of gratitude. Lack of power was most frequently cited as the reason for patients not complaining in such instances. In the financial sector, the “relationship

32 investment barrier”, time and financial costs, apathy due to the perception that all financial institution are the same and service recovery influenced customers propensity to complain. Chapter seven conceptualises culture within the framework of consumer behaviour.

33 CHAPTER SEVEN: CULTURE AND CONSUMER COMPLAINT BEHAVIOUR 7.1

Introduction

This chapter is a conceptualisation of culture and investigates the occurrence of culture in marketing research. Culture is an important focal point in this treatise as language, as a proxy of culture, is being investigated as an independent variables impacting on consumer dissatisfaction responses. First, several pertinent definitions of culture will be provided. The next focus will be on investigating culture within the framework of consumer behaviour. The notion of culture will be broken into measurable constructs and their impact on consumer behaviour will be explained. Thirdly, findings from previous research conducted on the impact of culture on CCB will be analysed. The main language groups of South Africa will be classified according to Hofstede’s cultural dimensions. 7.2

Conceptualising Culture

Culture may be defined as “that complex whole which includes knowledge, belief, art, morals, law, custom and any other capabilities and habits acquired by man as a member of society” (1881, Tylor in Samuel Craig & Douglas, 2006). McCracken (1986, in Samuel Craig & Douglas, 2006) further described culture in a more holistic way as being “the lens through which the individual views phenomena”. Culture may also be defined as, “the collective programming of the human mind that distinguishes the members of one human group from those of another. Culture, in this sense, is a system of collective held values” (1981, Hofstede in Communal & Senior, 1999). The nature and complexity of culture is such that there are numerous definitions allocated to it. Consumer researchers generally follow these views of culture. Samuel Craig and Douglas (2006) state that culture consists of three components, namely language and communication systems, material culture and artefacts, and values and belief systems. 7.3

Impact of Culture on Consumer Behaviour

Figure 4 below is a framework for understanding the ways in which culture has an impact on consumer behaviour and vice versa (Luna & Gupta, 2001). A person’s behaviour is a result of that person’s cultural value system for a specific context. Individuals’ cultural value systems are developed over time as they are socialised into a particular community. The formation of

34 an individual’s cultural value system is influenced by societal culture, together with regional subculture and familial values. Hence, the cultural value system includes cultural elements that individuals have in common with groups(s) to which they belong, together with idiosyncratic values unique to the individual. As seen in the model, culture affects consumer behaviour, which itself may reinforce the manifestations of culture (1998, Peter & Olsen in Luna & Gupta, 2001). The consumption patterns of an individual may be viewed and imitated or rejected by others. It can then become the group’s norm of behaviour and be identified as part of the culture of a given population. Marketing communications serve as a vehicle to transfer meanings of values from the culturally constituted world to consumer goods, so marketing communications are represented in the model as a moderator of the effect of culture on consumer behaviour (1988, McCracken in Luna & Gupta, 2001). At the same time, marketing communications may also impact on a culture’s manifestations through advertising. It is important to note that marketing communications also impact on consumer behaviour independent of culture. This model is particularly relevant in the field on consumer complaint behaviour. The latter is an integral part of the discipline of consumer behaviour; hence this model illustrates some of the ways in which culture impacts on complaint behaviour.

Figure 4: A model of the interaction of culture with consumer behaviour (Luna & Gupta, 2001)

35 According to Samuel Craig and Douglas (2006) and Andreasen (1990, in Luna & Gupta, 2001), a consequence of consumer research involving a cultural dimension is the phenomenon of cultural interpenetration, which refers to the merging of aspects of one culture with another. This can lead to cultural contamination, which implies that the central values of one culture are no longer discrete and cannot be readily measured. This phenomenon is further discussed by Luna and Gupta (2001), who describe it as being “acculturation”. This is caused by the increased migration of people around the world, creating complexity in most markets. As these people are gradually exposed to their host culture, they begin to acculturate and may adopt its norms of behaviour. Increased migration is one of the forces that are causing “radical changes” in consumer behaviour. Such a phenomenon may be very much present in South Africa, a country that is characterised by a highly ethnically diverse population. In such contexts, marketing research becomes difficult due to the complexity of the cultural factors involved. Research in marketing on culture follows three major streams (1995, Sojka & Tansuhaj in Samuel Craig & Douglas, 2006), namely: • Abstract or intangible elements of culture such as values and belief systems, • Material aspects of culture, such as artefacts, symbols and rites, • The communication links which bind and perpetuate a cultural system. Each of these provides unique and distinct perspectives and insights into a facet of culture. In the past, these three elements have often been discussed independently or attention has been focused on a single element. They are, however, closely linked together as shown in Figure 5 below.

36 Figure 5: Components of Culture (Samuel Craig & Douglas, 2006, p. 324) Communication is a means of transmitting the intangible aspects of culture, such as values and beliefs from one person to another, or from one generation to another (Samuel Craig & Douglas, 2006). This communication process is inherently dynamic and at the same time continually evolving. It is a key element of culture as it is a means of relaying and interpreting messages relating to the world around an individual. Communication takes place in a physical and social context such as time, location and the social relationship of the participants, as well as in relation to other competing messages (9173, Hall in Samuel Craig & Douglas, 2006). All these impact on how a communication is received. Members of a culture share a common key for interpreting their social surroundings, which creates rules for governing the interaction. Members of dissimilar cultures may not know how to interpret these signs, resulting in miscommunication. Language is an essential component of communication as it provides a mechanism for encoding and decoding messages (Samuel & Craig, 2006). Therefore, a shared language is a vital factor unifying members of a common culture. Language provides an organising schema for the interpretation and understanding of the world. Language also seems to play an essential role in the formation of thought patterns and behavioural responses as well as in the transmission of cultural norms and behaviour patterns from one generation to another. Language also assigns meanings to objects and symbols for the individual. Samuel and Craig (2006) further sate that language has many facets that relate to the meaning of consumer products. Linguistic structure has a significant impact on the formation of cognitive processes, such as perception and hence judgement and choice, the encoding and recall of information as well as in brand recall and recognition (1998, Schmitt & Zang in Samuel & Craig, 2006). 7.3

Socio-cultural Factors in Consumer Complaint Behaviour

Several socio-cultural factors have been recognized as being modulators of the complaint process. (Crié, 2003). According to Day and Landon (1977, in Crié, 2003), the three groups of characteristics to be considered for a cross-cultural approach to the phenomenon are: a) The standards of life and the effectiveness of the marketing system; b) The degree of regulation and control of economic activities and marketing practices;

37 c) The availability of information to help consumers make their choice or know where to complain. The fact that ethnic origin plays a significant part in CCB, especially through values and way of life, is clearly emphasized by Cornwell, Bligh and Babakus (1991, in Crié, 2003). This is corroborated by Webster (1991, in Crié, 2003) who finds an influence of ethnic factors when the effect of social variables is controlled. Moreover, people inhabiting rural areas are more likely to negative word of mouth (1993, Kolodinsky in Crié, 2003). Farhangmehr and Silva (1995, in Crié, 2003) state that educational level is a decisive variable; the higher it is, the more consumers tend to complain in a dissatisfaction situation. Gronhaug and Zaltman (1981, in Crié, 2003) have demonstrated that economic indicators such as income only have a weak explanatory power and that it is the same for demographic factors. Singh’s (1988) research shows that “complainers” tend to have superior incomes, a higher educational level, are still working and are younger. 7.3.1

The Impact of Individualism, Collectivism and the Importance of In-groups and Outgroups on CCB

One of the most widely studied and validated constructs used to compare and describe cultures is that of individualism and collectivism, as shown in a study carried out by Liu and McClure (2001). Traditionally, individualism-collectivism have described as being on a continuum. People in individualistic cultures value independence and self-sufficiency and think in terms of “I”. In individualistic countries, people are concerned with themselves and their families rather than others. The individual is important, and each person’s rights are highly valued (Francesco & Gold, 2005; 1980, Hofstede in Sun, Horn & Merritt, 2004). Conversely, people on collectivist cultures have a propensity to behave according to the social norms that are designed to maintain social harmony amongst the members of the ingroup and think in terms of “we”. Collectivistic cultures value the overall good of the group because the expectation is that people should subordinate their individual interests and needs for the benefit of the group (Francesco & Gold, 2005). There are clear expectations about how people in the group should behave because being part of a collective is so important. Collectivists have a tendency not to express their emotions in an outward fashion, particularly negative emotions that are often held back in the public. If negative emotions are to be expressed, they are likely to be done so in discussions within the context of intimate social settings. “In contrast, an individualist’s self-esteem and attitudes depend more on his/her

38 success in being unique, in self-expression (including expressing negative emotions), and in validating internal “defining” attributes.” Individualistic and collectivist cultures also classify in-groups and out-groups in very dissimilar ways (Sun, Horn & Merritt, 2004). For example, in individualistic cultures, the ingroup is characterized by “people who are like me in social class, race, beliefs, attitudes and values.” while the in-group in the collectivist culture is defined as “family and friends and other people concerned with my welfare”. This has important implications for the differences in which an individualist and a collectivist would interact with many businesses (especially foreign firms). A collectivist would consider associations with the business as being an outgroup interaction, and consequently, would avoid any unnecessary involvement, a phenomenon that is likely to be higher than for an individualist. In addition, in a collectivist culture, “avoiding doing business with a firm because of my bad experience with that firm” may be of no less importance than “convincing my friends and relatives not to do business with that firm”, since it is the duty of the individual not to let his/her in-group have the same bad experience. Hence, third party actions and voice actions are schemes that necessitate interactions with the out-groups, and, thus, are more likely to be avoided. However, when dissatisfied, consumers from a collectivistic culture are more likely to engage in private behaviour than those from an individualistic culture. The study also showed that dissatisfied individualistic consumers who voiced dissatisfaction are more likely to stay than those who did not. However, because of cultural differences, dissatisfied consumers from collectivistic cultures who voiced dissatisfaction are less likely to stay than those who did not. The results revealed that the differences on third-party behaviour as well as non-action response are not significant (Sun, Horn & Merritt, 2004). This implies that when dissatisfied, consumers in either a collectivistic or an individualistic culture are less likely to engage in a non-action response but are more likely to engage in action responses. Among the three types of action responses (that is voice, private, and third-party), they are more likely to take either voice or private action, but less likely to take third-party action. McFarlin and Coster (1999) conducted a study to describe the three subcultures present in South Africa, that is the Afrikaners (the Afrikaans-speaking community), the Anglos (the English-speaking community) and the Africans (including the Xhosa-speaking community), using Hofstede’s cultural dimensions. This is shown in Table 2 below.

39

Table 2: Predicted differences among major South African cultural groups using Hofstede’s cultural dimensions (McFarlin & Coster, 1999, p. 77). SOUTH AFRICAN CULTURAL GROUPS Cultural dimensions Afrikaner Anglo African Uncertainty High Low Very High Avoidance Power Distance

High

Moderate

High

Individualism/

Individualist

Individualist

Collectivist

Collectivism Masculinity/

Masculine

Masculine

Feminine

Femininity Long-term

Long-term

Long-term

Short-term

orientation

vs.

Short-term orientation Several predictions can be made from the findings of McFarlin and Coster (1999) and Liu and McClure (2001): a) The Africans, being part of a collectivistic community, are more likely to engage in private behaviour when dissatisfied, rather than engage in third party actions and voice behaviours. Individualistic consumers from the Afrikaans and Anglo communities are however more likely to engage in third part actions and voice behaviours. b) The Afrikaans and the Anglos (both being individualistic) who voice their dissatisfaction are more likely to stay than those than those who did not. However, Africans (being collectivistic) who voice their dissatisfaction are less likely to stay than those who did not. c) When dissatisfied, consumers from the Afrikaner, Anglo and African community are less likely to engage in a non-action response but are more likely to engage in action responses. All three of these communities are more likely to take either voice or private action, but less likely to take third-party action.

7.4

Conclusion

40 This chapter provided a comprehensive definition of culture that is readily quantifiable within the framework of consumer behaviour. The proxy of language was discussed as an important proxy of culture. Findings from similar studies showed that individualism and collectivism are the two cultural dimensions that readily explain responses to dissatisfaction. Findings from a study classifying the main language groups in South Africa according to Hofstede’s cultural dimensions were included. This information is critical in the formation of hypotheses.

CHAPTER EIGHT: RESEARCH METHODOLOGY

41

8.1

Introduction

This chapter describes the measuring techniques used and the objectives of the main study. The study involves both a qualitative and a quantitative approach. The measuring instruments used were in-depth interviews and a survey. Information from the in-depth interviews was used in the design of the questionnaire. The purpose of this mixed methodology is it provides a more holistic understanding of the subject being studied. The in-depth interviews provided dense descriptions of the participants’ thinking processes in their own terms, and the survey allowed for the recording of responses from a larger diversity of sample as it is cheap (inexpensive) and easily analysed. Drawing from the literature and the model for consumer dissatisfaction, hypotheses are formulated. The population and sampling technique are described as is the research instrument. The data collection technique is outlined as well as the method employed in data analysis. The reliability and the validity of the measure will also be assessed. 8.2

Objectives of the Main Study

The aim of the study was to investigate cross-cultural differences in the responses to consumer dissatisfaction, in the services sector, in Port Elizabeth. The secondary objectives of the study are as follows: a. What are the responses to consumer dissatisfaction, with regard to services? b. What is the likelihood of the different cultural groups to engage in voice, private and third party action? 8.3

In-depth Interviews

In-depth interviews were carried out to determine the type of services that people in Port Elizabeth feel most strongly about for the research proposes. Such services are regarded as high-involvement services.

42 The objectives of the in-depth interviews were to: a. Identify the types of services that consumers regarded as being most critical b. Gather information for the design of scenarios to be used in survey. The respondents were provided with a list of services and they were required to choose the two types of services that they had the highest expectations of, with regard to service quality. The list of services categorised services according to the medical sector, insurance services, government services, transport services, grooming services, restaurant and hotel services, customer services in the retail sector, and telecommunications services. Respondents had the opportunity to suggest other services that they thought were important but that had not been included in the list. Interviewees were also required to describe a real-life experience of poor service quality with those services they had previously selected. The biographical details of the people interviewed are shown in Table 3 below (n=8). Table 3: Biographical details of Respondents in In-depth Interview Demographic Characteristics GENDER Male Female Total RACE White Black Indian Coloured Total INCOME (Per Month) R 1,001 – R 2000 R 2,001 – R 4,000 R 4,001 – R 6,000 R 6000+ Total AGE 15 – 25 26 – 36 37 – 47 48 – 58 Over 50 Total

Values in numbers 3 5 8 3 3 1 1 8 2 3 2 1 8 3 4 1 0 0 8

43 The two most frequently selected services were those relating to the medical sector and the insurance services. Therefore, scenarios using the respondents’ descriptions of real-life experiences within these two particular services sector were constructed to be used in the questionnaire. 8.4

The Measure

The aim of the survey was to determine the types of actions that consumers are most likely to undertake, following an experience of service failure. It comprised of scenarios that describe a specific event of dissatisfaction that required the respondent to project his/her intended behaviours in that particular situation. The scenario approach was used by Rousseau (1987) in an investigation of complaint behaviour. After reading the scenarios, the respondent had to choose, from a list of CCB intentions, all the possible routes of action that he/she would be likely to undertake. The biographical details considered were home language, race, gender, age, income level, and education level. The survey comprised of 10 items, adapted from research conducted by Singh (1988). Table 4 shows the items measured in the survey.

44

Table 4: Items Used to Measure CCB Intentions (Singh, 1988, p.105) How likely is it that you would: VOICE CCB1

Forget about the whole incident and do nothing?

CCB3

Definitely complain to the store manager/service provider?

CCB4

Go back or call the manufacturer/service provider immediately and ask them to take care of your problem?

PRIVATE CCB2

Decide not to use that product/service again?

CCB6

Speak to your friends and relatives about your bad experience?

CCB7 THIRD PARTY CCB8

Convince your friends and relatives not to use that product/service? Complain to a consumer agency and ask them to make the manufacturer/service provider take care of your problem?

CCB9

Write a letter to the local newspaper about your bad experience?

CCB10

Report to the consumer agency so that they can warn other consumers?

CCB5

Take some legal action against the manufacturer/service provider?

8.5 Hypotheses The following hypotheses were constructed from the literature review. H1: When dissatisfied, Afrikaans-speaking consumers are more likely to engage in voice behaviours (seeking redress from the seller) than Xhosa-speaking consumers. H2: When dissatisfied, English-speaking consumers are more likely to engage in voice behaviours than Xhosa-speaking consumers. H3: When dissatisfied, Xhosa-speaking consumers are more likely to engage in private behaviour (word-of-mouth communication or exit) than Afrikaans-speaking consumers. H4: When dissatisfied, Xhosa-speaking consumers are more likely to engage in private behaviour than English-speaking consumers.

45 H5: When dissatisfied, female consumers are more likely to engage in voice as a means of complaint behaviour than males consumers. 8.6

Sampling Technique and Description of the Sample

It is important to draw a distinction between a population and a sample. According to Malhotra (2006, p. 335), a population is “the aggregate of all the elements that share some common set of characteristics and that comprise the universe for the purpose of the marketing research problem”. On the other hand, a sample is “a subgroup of the population selected for participation in the study”. Due to time and financial constraints, it was not possible to measure the characteristics of the entire population; hence through sampling, inferences about the population parameters were made. The population in this study consists of the inhabitants residing in the city of Port Elizabeth, comprising of approximately 1, 005, 776 people (Stats SA, 2005). 8.6.1

Sampling Technique

There are two major sampling techniques. Firstly, there is non-probability sampling, which refers to “sampling techniques that do not use chance selection procedures, rather they rely on the personal judgement of the researcher”. Secondly, there is probability sampling which “relies on sampling methods that are selected by chance. Each element of the population has a fixed probabilistic chance of being selected for the sample” (Malhotra, 2006, p. 340). The type of sampling that was used in this study is non-probability convenience sampling. The proxy of culture to be studied is home language, namely Afrikaans, English and Xhosa. According to the census report 2001, these are the main language groups in Port Elizabeth (Stats SA, 2005). For the purposes of the study, the researchers decided that the sample size should be equally divided between the three language groups being investigated. The researchers approached respondents from a number of locations around Port Elizabeth (Greenacres Shopping Mall, Walmer Park Shopping Mall, Summerstrand Village Shopping Centre and the Nelson Mandela Metropolitan University South Campus), in order to complete questionnaires. Before embarking on data collection, the researchers familiarized themselves with fieldwork techniques and ethical considerations as indicated by Malhotra (2006).

46

8.6.2

Description of the Sample

A sample of 180 was drawn of which 150 were usable. It was the largest possible sample given the resource constraints. 8.6.1.1 Respondent home languages Table 5 below indicates that the number of participants per language group was almost equal, that is 50 participants for Xhosa, 49 respondents representing English speaking individuals and 51 participants that were Afrikaans speaking. Table 5: Home Language Distribution of Sample (n=150) Language Afrikaans English Xhosa

Count 51 49 50

Cumulative Count 51 100 150

Percent 34 % 33 % 33%

8.6.1.2 Respondent ages As can be seen from Table 6, the majority of participants in this study belonged to the youngest age-group, that is, between the ages of 18 and 35. Such a phenomenon may be due to the fact that a greater percentage of the sampling was carried out in NMMU, where students form the majority of the university body. Table 6: Age distribution of Sample (n=150) Age 18-30 31-51 52-72

Count 93 40 15

Cumulative Count 93 133 150

Percent 63 % 27 % 10 %

8.6.1.3 Respondent races Table 7 below shows the racial distribution of the sample. The White and Black racial groups formed the majority of the respondents. These two racial groups are the main racial groups in Port Elizabeth. The Coloured and Indian racial groups, who form the minority of the Port Elizabeth population, only added up to 9% of the sample. Table 7: Racial Distribution of Sample (n=150) Race group White Black

Count 82 54

Cumulative Count 82 136

Percent 56 % 36 %

47 Indian Coloured

12 2

148 150

8% 1%

8.6.1.4 Respondent Education Levels As can be observed from Table 8, a large percentage of the respondents possessed a degree or a diploma. This distribution may be due to the fact that a large percentage of the sampling was carried out on the premises of the NMMU, where there is a strong presence of students and academics. where individuals are predominantly students or academics. Table 8: Education Levels Distribution of Respondents (n=150) Education Levels
Count 7 43 62 38

Cumulative Count 7 50 112 150

Percent 5% 29% 41% 25%

Degree/Diploma 8.6.1.5 Respondent income levels Table 9 show that there is greater percentage of lower income earners in the sample. There are two possible reasons for this phenomenon. Firstly, a large part of the sampling was undertaken at the NMMU and therefore many students were involved in this research project. Students usually belong to lower-income groups. Secondly, the Eastern Cape is the poorest province of South Africa, and hence possesses a majority of lower income earners.

Table 9: Income Levels Distribution of Respondents (n=142) Income Levels R1,000 – 3,000 R 3,001-5,001 R 5,001-10,000 R 10,001-15,000 R 15,001+

Count 59 14 39 15 15

Cumulative Count 59 73 112 127 142

Percent 42% 10% 27% 11% 11%

48 8.6.1.6 Respondent gender The gender distribution reflects a normal trend where there is a greater percentage of females than males. This is shown in Table 10. Table 10: Gender Distribution of Sample (n=150) Gender Males Females

8.7

Count 65 85

Cumulative Count 65 150

Percent 43% 57%

Data analysis

The data was prepared through coding and was transcribed into an Excel Spreadsheet on a computer. Those questionnaires with missing or invalid answers were disregarded, leaving 150 usable questionnaires. The data was analysed using Stastistica, a software program specifically designed for data analysis. 8.7.1

Descriptive statistics

Descriptive statistics are those that present data in a way that is informative (Singh, 2001). Descriptive statistics summarise large amounts of data into useful information through the synopses of general trends and findings, which can be easily identified by the use of tables and graphics. It involves the calculation of means, modes, standard deviation and frequencies in order to make sense of the findings.

8.7.1.2 Mean According to Malhotra (2006), the mean is the average value of a series of numbers and is used to measure central tendency. It is obtained by summing all the elements in a set and dividing by the number of elements. 8.7.1.3 Mode The mode refers to a value in a sample distribution that occurs the most frequently, and it is useful when there are values that are extreme outliers, which would alter the mean unnaturally (Malhotra, 2006).

49

8.7.1.4 Median Malhotra (2006) articulates that median of a sample is the value that falls directly in the middle of all the values in the sample distribution arranged in ascending or descending order. Half the values are lower than this point and half are higher. 8.7.1.5 Standard Deviation Variance refers to the mean squared deviation of all the variables from the mean. Standard deviation is the square root of the variance and explains the average amount that the sample scores vary from the mean (Malhotra, 2006). 8.7.1.6 Frequency distribution A frequency distribution counts the responses associated with different values of the same variable, such as how may participants spoke Afrikaans, for example. It is expressed as a number and as a percentage of total responses (Malhotra, 2006). 8.7.2

Inferential statistics

Inferential statistics allow the researchers to learn more about the characteristics drawn from the population (Malhotra, 2006). These statistics are used to explain how variables are related to each other, rather than just suggesting that there exists a relationship. From a well-drawn sample and credible inferential statistics one is able to generalise findings to the population. The measures of inferential statistics used for data analysis in this study are Cohen’s D and analysis of variance (ANOVA).

8.7.2.1 Cohen’s D Cohen’s D is the difference between the means of two groups divided by the standard deviation of either group. It is able to describe the size of the effect of a variable. It is usually accepted that a 0.2 value indicates a small effect, 0.5 a medium effect and anything higher than 0.8 denotes a large effect (Venter, 2006). 8.7.2.2 Cronbach’s alpha Cronbach’s alpha is measure of internal consistency reliability that is the average of all possible split-half coefficients resulting from various splittings of the scale items. This

50 coefficient varies from 0 to 1 and a value of 0.6 or less usually indicates unsatisfactory internal consistency reliability (Malhotra, 2006). 8.7.2.3 Analysis of Variance (ANOVA) Malhotra (2006) states that ANOVA is a technique to examine the differences in the means of two or more populations. In the simplest form, one must have a metric dependant variable and one or more independent variables that are all categorical. 8.8

Reliability and validity

In order to authenticate any research study, it is necessary to investigate the reliability and validity of the measuring instrument used. In this study, Cronbach’s alpha was used to analyse the reliability and validity of the measure. The following Table 11 shows the reliability of the measure.

Table 11: Item Analysis Results for the 3 Dimensions

Dimensions/ Items Voice CCB 03 CCB 04 %Var Cronbach's α

Scenario 1

Scenario 2

.829 .829 69% .54

.839 .839 70% .58

.731

.767

Private CCB 02

51 CCB 06 CCB 07 %Var Cronbach's α

.796 .729 57% .62

.819 .615 55% .57

Third Party CCB 05 CCB 08 CCB 09 CCB 10 %Var Cronbach's α

.646 .788 .777 .504 47% .61

.431 .799 .830 .714 51% .64

The factor loadings indicate all the items were acceptable, except CCB5 for Scenario 2. However, this factor loading of 0.431 comes very close to the minimum factor loading of 0.45, and is acceptable in light of the small sample used. Cronbach’s alpha is also acceptable for all the three dimensions of voice, private and third party for both of the scenarios, as it ranges from 0.50 to 0.70, which is the acceptable range for exploratory research (Malhotra, 2006). However, Cronbach’s alpha can still be considered low for the voice complaint dimension (both scenarios) and private complaint dimension (scenario 2). CCB1 was not considered during the analysis as the researchers felt it did not follow a logical fit with the other items in the survey. The researchers believe that this may be the cause for the low Cronbach’s score in the voice complaint dimensions. The researchers deem the reliability scores of this measure can be improved by increasing the number of items under each dimension. Malhotra (2006, p. 269) states that “validity is the extent to which differences in observed scale score reflect true differences among objects on the characteristic being measured, rather than systematic or random errors”. For this study, content validity was achieved through expert analysis of the content, drawing from extensive research. These items have been extensively researched by Singh (1988) and have been successfully used for a number of studies conducted on CCB. 8.9

Conclusion

52 This chapter described the methodology utilised in the study. The main objectives of the study were outlined to direct the research process. The objectives of the two measuring instruments were also included. In-depth interviews were conducted in order to select the types of high-involvement services to be targeted for the study. Information was also obtained to construct scenarios, which were used in the survey. The themes selected for the scenarios were the medical and financial sectors. The survey captured the respondents’ responses to dissatisfaction and their biographical details. The survey items were drawn from previous studies conducted by Singh (1988), and were thoroughly researched. Hypotheses were drawn from the extensive literature review and constructed based on a satisfactory model. Data collection was undertaken by the researchers who approached respondents in various strategic locations within Port Elizabeth in order to capture a sample that is as representative as possible. The data was sorted and analysed through the use of statistical software packages. Both descriptive and inferential statistics were utilised in the data analysis process. The measure was deemed to be valid as it was an instrument that has been widely used in previous studies. Cronbach’s alpha confirmed the reliability of the measure as it was within acceptable range for exploratory research. The next chapter will provide an account of the main findings of the study as well as a discussion of these results. Hypothesis testing will be undertaken by use of descriptive and inferential statistics. The results are discussed drawing from the various theories and concepts outlined in the study. CHAPTER NINE: MAIN FINDINGS 9.1

Introduction

This chapter reports on the findings from the analysis of the impact of culture on responses to dissatisfaction. The proxy used for culture is language. The impact of other demographic details, namely gender, race, education, income and age, on the responses to dissatisfaction will also be investigated. These responses have been divided into three constructs, namely voice, private and third party complaint behaviours.

53

The results of the responses to the measure will be discussed under each demographic detail, starting first with language groups. The findings from both scenarios will be discussed together and, the similarities and differences between the two will be highlighted in the analysis. The results from the hypothesis testing will also be dealt with. Conclusions will be drawn from the relationship between these demographical details and the three complaint behaviour dimensions. These discussions include the practical implications of these findings for South African service providers. 9.2

Results of the study

The results of this study will be disclosed using descriptive and inferential statistics. 9.3

Testing of hypothesised relationships: Impact of Language on Voice, Private and

Third Party Complaint Behaviours Inferential statistics were used to test the relationship between independent variable language and the dependent variables voice, private and third party complaint behaviours. Statistical and practical significance was also computed in order to determine whether the findings were reliable and useful. The hypotheses test whether language, as a proxy of culture, has an impact on the likelihood that the dissatisfied consumer will engage in any one of these complaint behaviours, which are voice, private or third party behaviours.

9.3.1

H1: When dissatisfied, Afrikaans-speaking customers are more likely to engage in

voice complaint behaviours (seeking redress from the seller) than Xhosa-speaking customers. H1: µ1 > µ2 F≠0 P = 0.05

54 H0: When dissatisfied, Afrikaans-speaking customers are less likely, or equally likely, to engage in voice complaint behaviours than Xhosa-speaking customers. H0: µ1 ≤ µ2 Where:

µ1 = Propensity of dissatisfied Afrikaans-speaking customers to engage in voice behaviours µ2 = Propensity of dissatisfied Xhosa-speaking customers to engage in voice behaviours Table 12: Mean Values for Language and Complaint Behaviours for Scenario 1 (n= 150) Overall Mean

Afrikaans (n=51)

English (n=49)

Xhosa (n=50)

F

df Effect

p

Voice

0.60

0.34

0.81

0.67

22.66

2

.000

Private

0.51

0.22

0.69

0.63

36.98

2

.000

Third Party

0.48

0.29

0.63

0.52

17.72

2

.000

Table 13: Mean Values for Language and Complaint Behaviours for Scenario 2 (n= 150) Overall Mean

Afrikaans (n=51)

English (n=49)

Xhosa (n=50)

F

df Effect

p

Voice

0.61

0.35

0.89

0.59

31.69

2

.000

Private

0.56

0.33

0.73

0.61

20.77

2

.000

Third Party

0.41

0.21

0.53

0.49

17.49

2

.000

From Table 12 and 13, it can be seen that dissatisfied Afrikaans-speaking customers are much less likely to engage in voice complaint behaviours than Xhosa-speaking customers for both scenarios, when F is higher than zero, with a p value of 0.000. This indicates that there is a statistically significant difference between dissatisfied Afrikaans- and Xhosa-speaking consumers’ likelihood to engage in voice behaviours. Table 14: Practical Significance of the Impact of Language on Voice, Private and Third Party Complaint Behaviours in Scenario 1 (n=150)

55 Overall Mean Voice

0.60

Privat e

0.51

Third Party

0.48

Afrikaan s (n=51) 0.34

English (n=49)

Xhosa (n=50)

Scheffé p

Cohen's d

Interpretation

0.81

0.67

1:2 (.000) 1:3 (.000) 1:2 (.000) 1:3 (.000) 1:2 (.000) 1:3 (.000)

0.93

Large

0.65

Moderate

0.93

Large

0.82

Large

0.68

Moderate

0.45

Moderate

0.22

0.69

0.29

0.63

0.63

0.52

Table 15: Practical Significance of the Impact of Language on Voice, Private and Third Party Complaint Behaviours in Scenario 2 (n=150) Overall Mean Voice

Privat e Third Party

0.61

0.56

0.41

Afrikaan s (n=51) 0.35

English (n=49)

Xhosa (n=50)

Scheffé p

Cohen's d

Interpretatio n

0.89

0.59

1:2 (.000) 1:3 (.002) 2:3 (.000) 1:2 (.000) 1:3

1.07

Large

0.47

Moderate

0.60

Moderate

0.79

Large

0.56

Large

1:2 (.000) 1:3 (.000)

0.64

Moderate

0.55

Moderate

0.33

0.21

0.73

0.53

0.61

0.49

Table 14 and 15 shows that this relationship has a moderate practical significance. Thus, µ1 > µ2, and H1 is rejected. H0 is not rejected. This implies that Afrikaans-speaking consumers are less likely to engage in voice complaint behaviours when dissatisfied than Xhosa-speaking consumers in both scenarios.

56 9.3.2

H2: When dissatisfied, English-speaking customers are more likely to engage in voice

complaint behaviours (seeking redress from the seller) than Xhosa-speaking customers. H2: µ1 > µ2 F≠0 P = 0.05 H0: When dissatisfied, English-speaking customers are less likely, or equally likely, to engage in voice complaint behaviours than Xhosa-speaking customers. H0: µ1 ≤ µ2 Where:

µ1 = Propensity of dissatisfied English-speaking customers to engage in voice complaint behaviours

µ2 = Propensity of dissatisfied Xhosa-speaking customers to engage in voice complaint behaviours There is no statistically significant difference between the likelihood of dissatisfied Englishand Xhosa-speaking consumers engaging in voice complaint behaviours. The results are similar for both scenarios as can be seen from tables 12 and 13.

Thus:

µ1 ≤ µ2, and the null hypothesis is not rejected. H0 is not rejected. This means that English-speaking consumers are as likely as Xhosaspeaking consumers of engaging in voice complaint behaviours when dissatisfied. Tables 12 and 13 also show that there is a statistically significant difference in the likelihood of voice complaint behaviours between the English- and Afrikaans-speaking dissatisfied

57 consumers, with F being higher than zero and a p value of 0.000. This means than the Afrikaans-speaking consumers are less likely to engage in voice behaviours than the Englishspeaking consumers when dissatisfied, for both scenarios. Tables 14 and 15 further indicate that this relationship has a large practical significance. 9.3.3

H3: When dissatisfied, Xhosa-speaking consumers are more likely to engage in

private complaint behaviours (word-of-mouth communication or exit) than Afrikaansspeaking consumers. H3: µ1 > µ2 F≠0 P = 0.05 H0: When dissatisfied, Xhosa-speaking consumers are less likely, or equally likely, to engage in private complaint behaviours than Afrikaans-speaking consumers. H0: µ1 ≤ µ2 Where:

µ1 = Propensity of dissatisfied Xhosa-speaking consumers to engage in private complaint behaviours

µ2 = Propensity of dissatisfied Afrikaans-speaking consumers to engage in private complaint behaviours Tables 12 and 13 indicate that there is statistically significant difference in the likelihood of engaging in private complaint behaviours between the Xhosa- and the Afrikaans-speaking dissatisfied consumers, with F being higher then zero and a p value of 0.000. This means that the Xhosa-speaking consumers are more likely to engage in private complaint behaviour when dissatisfied than the Afrikaans-speaking dissatisfied consumers, in both scenarios. Tables 14 and 15 further show that this relationship has a large practical significance. Thus,

58

µ1 > µ2, and H3 is not rejected. H3 is not rejected as the difference in the likelihood to engage in private complaint behaviour between the Xhosa- and the Afrikaans-speaking dissatisfied consumers has both practical and statistical significance. 9.3.4

H4: When dissatisfied, Xhosa-speaking consumers are more likely to engage in

private complaint behaviours than English-speaking consumers. H4: µ1 > µ2 F≠0 P = 0.05 H0: When dissatisfied, Xhosa-speaking consumers are less likely, or equally likely, to engage in private complaint behaviours than English-speaking consumers. H0: µ1 ≤ µ2 Where:

µ1 = Propensity of dissatisfied Xhosa-speaking consumers to engage in private complaint behaviours

µ2 = Propensity of dissatisfied English-speaking consumers to engage in private complaint behaviours Tables 12 and 13 show that there is no statistically significant difference in the likelihood of engaging in private complaint behaviours between the Xhosa- and the English-speaking dissatisfied consumers, with the p value being greater than 0.05. This means that both language groups are equally likely of engaging in private complaint behaviours for both scenarios. Thus,

µ1 ≤ µ2, and the null hypothesis is not rejected.

59

H0 is not rejected. This implies that both the Xhosa- and the English-speaking consumers are just as likely to engage in private complaint behaviours when dissatisfied. 9.4

Testing of relationships: Impact of Race on Voice, Private and Third Party

Complaint Behaviours Inferential statistics were used to test the relationship between the independent variable race and the dependent variables voice, private and third party complaint behaviours. Statistical and practical significance was also computed in order to determine whether the findings were reliable and useful. Table 16: ANOVA Grouping for Race in Scenario 1 (n=148) Overall White Black Indian Mean (n=82) (n=54) (n=12)

F

Voice

0.60

0.51

0.68

0.83

5.21

Private

0.51

0.38

0.64

0.81

15.6 2

Third Party

0.48

0.37

0.54

0.85

df Effec t 2 2

17.0 4

2

p

Scheffé p

Cohen' sd

Interpretation

. 007 . 000

1:3 (.032) 1:2 (.000) 1:3 (.000) 1:2 (.005) 1:3 (.000) 2:3 (.003)

0.64

Moderate

0.53

Moderate

0.85

Large

0.34

Moderate

0.97

Large

0.63

Moderate

. 000

Table 17: ANOVA Grouping for Race in Scenario 2 (n=148) Overall Mean

White (n=82)

Black (n=54)

Indian (n=12)

F

df Effect

p

Scheffé p

Cohen 's d

Interpreta tion

Voice

0.61

0.55

0.61

0.92

4.68

2

.011

0.74

Moderate

Private

0.56

0.48

0.62

0.75

4.89

2

.009

0.54

Moderate

Third Party

0.41

0.31

0.50

0.60

8.69

2

.000

1:3 (.011) 1:3 (.044) 1:2 (.004) 1:3 (.010)

0.37

Moderate

0.59

Moderate

60

9.4.1

Voice Complaint Behaviours

This section reports on relationships found between race and the likelihood of engaging in voice complaint behaviours, as indicated in Table 16 and Table 17. In scenario one, there is statistically significant difference between the likelihood that White and Indian consumers will engage in voice complaint behaviours when dissatisfied, when F is greater than zero with a p value of 0.007. In Scenario 2, the results are similar, with a p value of 0.011. This means that in both scenarios, dissatisfied Indian consumers are more likely to engage in voice complaint behaviours than dissatisfied White consumers. For both scenarios, this relationship has moderate practical significance. 9.4.2

Private Complaint Behaviours

This section reports on relationships found between race and the likelihood of engaging in private complaint behaviours, as shown in Table 16. In Scenario one, there is a statistically significant difference between the likelihood that White and Black consumers will engage in private complaint behaviours when dissatisfied, when F is greater than zero with a p value of 0.000. This implies that Black consumers are more likely to engage in private complaint behaviours than White consumers when dissatisfied. This relationship has moderate practical significance. In Scenario 2, there was no relationship found. In both scenarios, there is a statistically significant difference between the likelihood that White and Indian consumers will engage in private complaint behaviours when dissatisfied, when F is greater than zero with a p value of 0.000 in the first scenario and a p value of 0.009 in the second scenario. This implies that Indian consumers are more likely to engage in private complaint behaviours than White consumers when dissatisfied. This relationship has large practical significance in scenario one and a moderate practical significance in scenario two. 9.4.3

Third Party Complaint Behaviours

This section deals with the relationships found between race and the likelihood of engaging in private complaint behaviours, as indicated in Tables 16 and 17.

61

In both scenarios, Black consumers are more likely to engage in third party complaint behaviours than White consumers when dissatisfied. This relationship has both statistical significance when F is greater than zero with a p value of 0.000, and moderate practical significance. Table 16 and Table 17 also indicate that Indian consumers are more likely to engage in third party complaint behaviours than White consumers when dissatisfied. In scenario one, this relationship has both statistical significance when F is greater than zero with a p value of 0.000, and large practical significance. In scenario two, the p value of 0.010 points to a smaller statistical significance and only moderate practical significance. In scenario one, the third relationship that has statistical significance when F is greater than zero with a p value of 0.000 is the difference in third party complaint behaviours between Indian and Black consumers. This implies that in scenario one, Indian consumers are more likely to engage in third party complaint behaviours than Black consumers when dissatisfied. This relationship has moderate practical significance. 9.5

Testing of hypothesised relationships: Impact of Gender on Voice, Private and

Third Party Complaint Behaviours Inferential statistics were used to test the relationship between independent variable gender and the dependent variables voice, private and third party complaint behaviours. Statistical and practical significance was also computed in order to determine whether the findings were reliable and useful. The hypothesis tests whether gender has an impact on the likelihood that the dissatisfied consumer will engage in any one of these complaint behaviours, which are voice, private or third party behaviours. H5: When dissatisfied, female consumers are more likely to engage in voice as a means of complaint behaviour than males consumers. H5: µ1 > µ2

62 F≠0 P = 0.05 H0: When dissatisfied, female consumers are less likely, or equally likely, to engage in voice complaint behaviour than male consumers. H0: µ1 ≤ µ2 Where:

µ1 = Propensity of dissatisfied female consumers to engage in voice complaint behaviours µ2 = Propensity of dissatisfied male consumers to engage in voice complaint behaviours

Table 18: t-test Grouping for Gender in Scenario1 (n=150) Overall Mean

Males (n=65)

Females (n=85)

tvalue

df

p

Cohen's d

Voice

0.60

0.53

0.66

1.95

148

.053

n.a.

Private

0.51

0.47

0.54

1.17

148

.245

n.a.

Third Party

0.48

0.52

0.44

-1.61

148

.109

n.a.

Table 19: t-test Grouping for Gender in Scenario 2 (n=150) Overall Mean

Males (n=65)

Females (n=85)

tvalue

df

p

Cohen's d

Voice

0.61

0.52

0.68

2.49

148

.014

0.32

Private

0.56

0.54

0.56

0.37

148

.715

n.a.

Third Party

0.41

0.47

0.36

-1.96

148

.052

n.a.

63 As can be seen from Table 18, in Scenario one, there is no statistically significant difference between the likelihood that male and female consumers will engage in voice complaint behaviours when dissatisfied, with a p value of 0.053. Therefore in the first scenario, male and female consumers are equally likely of engaging in voice complaint behaviours when dissatisfied. However, the p value closely approaches a value less than 0.05. The researchers believe that this relationship may have been established with a larger sample. Thus,

µ1 > µ2, and H0 is not rejected. H0 is not rejected as the difference in the likelihood that male and female dissatisfied consumers will engage in voice complaint behaviours is not statistically significant. From Table 19, however, it is clear that certain situations, as in Scenario 2, there is a statistically significant difference between the likelihood that dissatisfied male and female consumers will engage in voice complaint behaviours, with a p value of 0.014. Hence in the second scenario, female consumers are more likely to engage in voice complaint behaviours when dissatisfied than male consumers. This relationship has moderate practical significance. Thus,

µ1 ≤ µ and H5 is not rejected. H5 is not rejected as the difference in the likelihood that male and female dissatisfied consumers will engage in voice complaint behaviours is statistically and practically significant. 9.6

Testing of relationships: Impact of Education Level on Voice, Private and Third

Party Complaint Behaviours Inferential statistics were used to test the relationship between the independent variable education level and the dependent variables voice, private and third party complaint behaviours. Statistical and practical significance was also computed in order to determine whether the findings were reliable and useful.

64 Table 20: ANOVA Grouping for Education Level in Scenario 1 (n=150) Ove rall Mea n


Matric (n=43)

Degree/ Diploma (n=62)

Post Grad Deg/Dip (n=38)

F

df Effect

p

Scheffé p

Cohen' sd

Voice

0.60

0.43

0.60

0.61

0.62

0.46

3

.708

n.a.

n.a.

Private

0.51

0.62

0.48

0.52

0.51

0.33

3

.807

n.a.

n.a.

Third Party

0.48

0.61

0.45

0.48

0.48

0.50

3

.684

n.a.

n.a.

Table 21: ANOVA Grouping for Education Level in Scenario 2 (n=150) Ove rall Mea n


Matric (n=43)

Degree/ Diploma (n=62)

Post Grad Deg/Dip (n=38)

F

df Effect

p

Scheffé p

Cohen' sd

Voice

0.61

0.57

0.56

0.63

0.63

0.34

3

.797

n.a.

n.a.

Private

0.56

0.67

0.53

0.56

0.55

0.33

3

.802

n.a.

n.a.

Third Party

0.41

0.54

0.38

0.40

0.43

0.59

3

.620

n.a.

n.a.

Table 20 and Table 21 show that there is no statistically significant relationship between education level and the likelihood of engaging in either one of voice, private or third party complaint behaviours. 9.7

Testing of relationships: Impact of Age on Voice, Private and Third Party

Complaint Behaviours Inferential statistics were used to test the relationship between the independent variable age and the dependent variables voice, private and third party complaint behaviours. Statistical and practical significance was also computed in order to determine whether the findings were reliable and useful. Table 22: ANOVA Grouping for Age in Scenario 1 (n=148) Overall Mean

18 - 30 (n=93)

31 - 51 (n=40)

52 - 72 (n=15)

F

p

Scheffé

Cohen's d

0.09

df Effec t 2

Voice

0.60

0.61

0.59

0.63

.916

n.a.

n.a.

Private

0.51

0.50

0.56

0.44

0.63

2

.533

n.a.

n.a.

Third Party

0.48

0.44

0.56

0.48

1.96

2

.145

n.a.

n.a.

65 Table 23: ANOVA Grouping for Age in Scenario 2 (n=148) Overall Mean

18 - 30 (n=93)

31 - 51 (n=40)

52 - 72 (n=15)

F

p

Scheffé

Cohen's d

0.10

df Effec t 2

Voice

0.61

0.60

0.63

0.63

.902

n.a.

n.a.

Private

0.56

0.53

0.63

0.55

1.06

2

.348

n.a.

n.a.

Third Party

0.41

0.38

0.46

0.48

1.41

2

.247

n.a.

n.a.

Table 22 and 23 show that there is no statistically significant relationship between age and the likelihood of engaging in either one of voice, private or third party complaint behaviours in both scenarios. 9.8

Testing of relationships: Impact of Income Level on Voice, Private and Third

Party Complaint Behaviours Inferential statistics were used to test the relationship between the independent variable income level and the dependent variables voice, private and third party complaint behaviours. Statistical and practical significance was also computed in order to determine whether the findings were reliable and useful. Table 24: ANOVA Grouping for Income Level in Scenario 1 (n=142) Overa ll Mean

1,0003,000 (n=59 )

3,0015,000 (n=14 )

10,00115,000 (n=15)

15,00 1+ (n=15 )

F

df Effe ct

p

Scheff é

Cohe n's d

0.46

5,00110,00 0 (n=39 ) 0.62

Voice

0.60

0.62

0.73

0.57

0.85

4

.495

n.a.

n.a.

Private

0.51

0.51

0.33

0.55

0.62

0.51

1.32

4

.266

n.a.

n.a.

Third Party

0.48

0.43

0.38

0.46

0.58

0.65

2.28

4

.064

n.a.

n.a.

Table 25: ANOVA Grouping for Income Level in Scenario 2 (n=142)

Voice

Overa ll Mean

1,0003,000 (n=59 )

3,0015,000 (n=14 )

0.61

0.61

0.50

5,00110,00 0 (n=39 ) 0.63

10,00115,000 (n=15)

15,00 1+ (n=15 )

F

df Effe ct

p

Scheff é

Cohe n's d

0.80

0.50

1.42

4

.230

n.a.

n.a.

66 Private

0.56

0.55

0.43

0.56

0.76

0.53

1.67

4

.161

n.a.

n.a.

Third Party

0.41

0.37

0.38

0.40

0.52

0.53

1.23

4

.300

n.a.

n.a.

Tables 24 and 25 shows that there is no statistically significant relationship between income level and the likelihood of engaging in either one of voice, private or third party complaint behaviours in either scenarios. 9.9

Comparing Findings of Scenarios One and Two

The purpose of this section is to draw comparisons between findings from scenario one and two. Table 19: Descriptive Statistics for Scenario 1 and 2 (n=150) SCENARIO ONE Statistics

Voice

Private

SCENARIO TWO

Third Party

Voice

Private

Third Party

n

150

150

150

150

150

150

Mean

0.60

0.51

0.48

0.61

0.56

0.41

SD

0.40

0.36

0.32

0.40

0.36

0.32

Min

0.00

0.00

0.00

0.00

0.00

0.00

Quartile 1

0.50

0.33

0.25

0.50

0.33

0.25

Median

0.50

0.67

0.50

0.50

0.67

0.25

Quartile 3

1.00

0.67

0.75

1.00

1.00

0.75

Maximum

1.00

1.00

1.00

1.00

1.00

1.00

Table 19 above is a comparison of the descriptive statistics for scenario one and two. This information is illustrated in Figure 6 below.

67

Figure 6: Comparing the Mean Values of Complaint Behaviour Dimensions for Scenarios 1 &2 Figure 6 shows the mean values comparison of voice, private, and third party complaint behaviours for scenario one and two. The values for these complaint dimensions follow a similar pattern across both scenarios, with voice complaints being the most popular, followed by private then third party complaint behaviours. Voice behaviours as a form of complaint were found to be equally popular across both scenarios. Private complaint behaviours were more present for scenario two than for scenario one, whilst the reverse was true for third party complaints.

68 9.10

Discussion of results

The following section will discuss and interpret the results of the study, with a particular focus on the practical implications of the findings. 9.10.1 Language This section discusses the findings of the relationship between language and complaint behaviours. The main findings can be summarised as such: a. Dissatisfied Afrikaans-speaking customers are less likely to engage in voice complaint behaviours than Xhosa-speaking customers, and this relationship has moderate practical significance. b. English-speaking consumers are as likely as Xhosa-speaking consumers to engage in voice complaint behaviours when dissatisfied, and this relationship has large practical significance. c. Xhosa-speaking consumers are more likely to engage in private complaint behaviours when dissatisfied than the Afrikaans-speaking dissatisfied consumers, and this relationship has moderate practical significance. d. Both the Xhosa- and the English-speaking consumers are just as likely to engage in private complaint behaviours when dissatisfied. The first finding from the study, in both scenarios, did not support the researchers’ hypothesis. Afrikaans-speaking consumers were expected to engage in voice behaviours more than their Xhosa counterparts when dissatisfied. This is because the Afrikaans-speaking community has been previously shown to be part of an individualistic culture, which implies that it is more likely to engage in voice and third party behaviours (McFarlin & Coster, 1999; Liu & McClure, 2001). Collectivistic cultures, such as the Xhosa-speaking consumers, on the other hand, have been previously shown to be more likely to engage in private behaviour as this does not necessitate them to go outside their in-group. Such an unexpected finding has several implications. South Africa is an ethically diverse country where there is great complexity in accurately defining market segments in terms of culture. It may be that many cultures in South Africa have undergone a process of acculturation, whereby there is an active interchange of cultural practices amongst different cultures with rapid integration (Luna & Gupta, 2001; Samuel Craig & Douglas, 2006). This would imply that the Afrikaans-

69 speaking culture has elements of both individualistic and collectivistic cultures. The study that these hypotheses were based on was conducted by McFarlin and Coster in 1999, which was eight years ago from now. It may be that, in this time period, that there has been a rapid acculturation process that has contaminated the primary cultures of South Africa such that they cannot be strictly classified as being collectivistic or individualistic. This finding can also be explained in terms of non-culturally related variables. Previous theoretical reasons suggested that complainers would likely to be from higher-income groups as those consumers have better resources in terms of information and self-confidence. However, more recent studies have shown that complainers come from predominantly lower income groups (Tronvoll, 2003). The Afrikaans-speaking consumers emerged as a predominantly wealthy segment of the market after the apartheid era. The fact that this language group complains minimally compared to other language groups can be hence explained in terms of their higher socio-economic status. This finding has significant practical implications for service providers. Afrikaans-speaking consumers are the least likely to complain compared to other cultures. It is difficult to find concrete reasons behind such an unexpected phenomenon as this study does not provide information on its causes. It is extremely detrimental for companies not to be aware of the dissatisfaction of the Afrikaans-speaking consumers as they represent a large and prosperous section of the South African market with considerable buying power. Afrikaans-speaking consumers would tend not to bring their dissatisfaction to the attention of service providers such that the companies are not afforded with the opportunity to rectify the problem. This may result in the service providers losing their Afrikaans-speaking consumers to competitors, which implies a loss of profitability and poor corporate image. This finding provides crucial information that service providers must take into account for effective complaint management when dealing with this section of the South African population. The second finding for scenario one was also contrary to the initial hypothesis. It was found that English-speaking dissatisfied consumers are as likely as their Xhosa counterparts to engage in voice complaint behaviours. What was initially hypothesised was that Englishspeaking consumers would engage more in voice behaviours than the Xhosa-speaking consumers when dissatisfied. This is because the English-speaking consumers belong to an individualistic culture whilst Xhosa-speaking consumers belong to a collectivistic culture

70 (McFarlin & Coster, 1999; Liu & McClure, 2001). Such a finding may also be attributed to a process of acculturation that has taken place in South Africa, such that Xhosa-speaking consumers may have integrated elements from their individualistic counterparts (Luna & Gupta, 2001; Samuel Craig & Douglas, 2006). However, the second scenario showed that the English-speaking consumers did engage in voice behaviours more than their Xhosa counterparts. A possible explanation for the difference in results between the two scenarios is that dissatisfied consumers in the medical sector are frequently reluctant to voice out their complaints because of diverse reasons, for example feelings of powerlessness and feelings of gratitude (Coyle & Williams, 1999). Xhosa-speaking consumers are part of previously disadvantaged and vulnerable groups in South Africa, and may be more prone to these feelings of powerlessness and gratitude such that they are less likely to voice their complaints than the English-speaking consumers. The third finding is consistent with prior research. Xhosa-speaking consumers were more likely to engage in private complaint behaviours when dissatisfied than Afrikaans-speaking consumers. Being from a collectivistic culture, Xhosa-speaking consumers prefer to associate with members of their in-group and share their bad experiences with them in order to prevent them from having a similar experience (McFarlin & Coster, 1999; Liu & McClure, 2001). However, this finding also has to be viewed in context with other results for Xhosa-speaking consumers. The latter are also more likely, than their Afrikaans-speaking counterparts, to engage in both voice and third party complaint behaviours. In fact, the fourth finding shows that, for scenario one, English- and Xhosa-speaking consumers are just as likely to engage in private behaviours when dissatisfied. This means that such a characteristic cannot solely be attributed to stereotypical cultural dimensions of individualism and collectivism as the English- and Xhosa-speaking consumers, though belonging to highly different cultures; both have similar attitudes towards private complaint behaviours. Private behaviours are usually considered to be detrimental for a company’s complaint management practices as these complaints are not brought to the attention of the service provider in question. Instead, dissatisfied consumers engage in negative word-of-mouth or exit behaviours that do not allow the service provider to fix the problem, and hence retain the consumer and enhance its corporate image. It would detrimental for companies if the Englishand Xhosa-speaking consumers only scored high in private behaviours as it would indicate that these cultures do not bring their dissatisfaction to the attention of the service provider.

71 However, these two cultures also engage in voice and third party behaviours, particularly in voice behaviours, such that companies can still have the opportunity to resolve the situation. One possible explanation behind this finding is that these consumers do not have positive expectations of service recovery from the service provider and hence engage in private and third party behaviours as well. The literature states that, when dissatisfaction exceeds a certain threshold, consumers tend to use either negative word-of-mouth or leaving or appeal to a third party (Singh & Pandya, 1991). Therefore, this finding has mixed implications for companies. On the one hand, companies benefit by having knowledge of service failure incidents from English- and Xhosa-speaking consumers such that they can engage in service recovery and retain these consumers. On the other hand, these two language groups also engage extensively in private complaints which involve highly detrimental behaviours such as negative word-of-mouth and switching service providers. This will severely affect on the service provider’s corporate image resulting in a loss of market share. This is a problem area for service providers which they have to attend to. 9.10.2 Race Race, which is closely related to first language and hence produced similar results, was also found to have a part to play with regard to voice, private and third party complaint behaviours. The main findings for this analysis can be summarised as: a. In scenario one and two, dissatisfied Indian consumers are more likely to engage in voice complaint behaviours than dissatisfied White consumers. This relationship has moderate practical significance. b. Black consumers are more likely to engage in private complaint behaviours than White consumers when dissatisfied in scenario one, and this relationship has moderate practical significance. c. Indian consumers are more likely to engage in private complaint behaviours than White consumers when dissatisfied in both scenarios. In the first scenario this relationship has large practical significance and moderate practical significance in the second scenario. d. In both scenarios, black consumers are more likely to engage in third party complaint behaviours than White consumers when dissatisfied and this relationship has moderate practical significance.

72 e. In scenario one and two, Indian consumers are more likely to engage in third party complaint behaviours than White consumers when dissatisfied. In the first scenario this relationship has large practical significance and moderate practical significance in the second scenario. f. In scenario one, Indian consumers are more likely to engage in third party complaint behaviours than Black consumers when dissatisfied. This relationship has moderate practical significance. The results showed that for both scenarios, dissatisfied Indian consumers are more likely to engage in voice complaint behaviours than dissatisfied White consumers. It was previously expected that White consumers would engage more in voice behaviors than Indian consumers when dissatisfied because Whites belong to a more individualistic culture whilst Indians belong to a collectivistic culture. Yet again, this finding may point to a process of acculturation in South Africa. Furthermore, it can be said that Indians are from a lower socioeconomic group compared to Whites, as they were part of the previously disadvantaged groups in South Africa. Hence the results of this study support the findings of Tronvoll (2003), who found that complainers come predominantly from lower socio-economic groupings. With regard to private complaint behaviours in scenario one, it was found that Black and Indian consumers to engage in private complaint behaviours than White consumers when dissatisfied. This finding is consistent with research stated by Francesco and Gold (2005) that Blacks and Indians, who are from a collectivist culture, were more likely to engage in private complaint behaviors than Whites, who are from an individualistic culture. As Black and Indian consumers are from a collectivistic culture, they prefer to associate with members of their in-group and share their unpleasant experiences with them as to prevent them from having a similar experience, rather than to voice a complaint which may be seen as going against social norms and not maintaining social harmony. The results change slightly in scenario two, where Black consumers are equally likely as White consumers to engage in private complaint behaviour. However, Indians are still more likely than the other two races to engage in private complaint behaviours. It was also found that Indian consumers were more likely to engage in third party complaint behaviours than White consumers when dissatisfied in both scenarios. In scenario one, Indian

73 consumers were more likely to engage in third party complaint behaviours than Black consumers when dissatisfied. The researchers suggest that Black consumers may not be fully aware of their rights as consumers due to their historically disadvantaged past such that they are not thoroughly acquainted with consumer associations and other such regulatory bodies. Hence, Black consumers are less likely to engage in third party behaviours when dissatisfied as they do not have adequate knowledge of the consumer associations they must contact. Due to the low literacy rates of the Black population, these consumers may also be reluctant to write letters to the newspapers due to their lack of skill. The above finding has important practical implications for companies. Private complaint behaviours have serious detrimental effects on an organisation because such complaints are not bought to a company’s attention. Hence the service provider cannot take any action for service recovery. Indians are highly likely of engaging in private complaint behaviour when dissatisfied, followed by Blacks. This could be potentially harmful for companies as they do not have the opportunity to take care of the problem. However, Indians are the most likely of engaging in both voice and third party behaviours when dissatisfied, compared to the other races. Hence, companies will have the opportunity of service recovery for Indians. However, the high incidence of private complaints will have a harmful effect on the company’s reputation and it may lose future consumers. Therefore, this finding has mixed implications for companies and needs to be further investigated. It is of utmost importance that companies pay attention to dissatisfied White consumers as they are the least likely, compared to the other races, to engage in any form of complaint behaviour. The researchers believe that these consumers, who are usually of higher socioeconomic status, may be more inclined to switch service providers when dissatisfied rather than to voice out their complaint and seek redress. These consumers may be better educated with better resources, and have more choices in the services market. Companies have much to lose if they do not pay attention to this segment of the market. White consumers are sizeable proportion of the South African consumer base with high purchasing power. Companies need to find strategies to tap into the dissatisfaction of these consumers to have an opportunity for service recovery and hence retain them Black consumers are also a potentially problematic segment of the consumer base for companies. Though they voice out their complaints more than Whites, it may be that they are

74 not fully aware of their options to express their dissatisfaction, especially appealing to third parties such as consumer associations. This may put them at risk for exploitation by companies. However, companies will find it detrimental that they are not fully aware of this segment’s dissatisfaction. Hence firms may be unaware of Black customers’ dissatisfaction, and since the Black consumer market in South Africa is currently the most rapidly expanding segment, companies’ ignorance in this respect will result in losing valuable market share and profit to competitors.

9.10.3 Gender Research conducted by Hogarth and English (2002, in Tronvoll, 2003) indicated that female consumers more often engage in voice behaviours as a means of complaining than male consumers. The findings of the present study, however, show that in examining the impact of CCB on gender, it is essential to consider the situational context. In scenario one, relating to the medical service sector, it was found that dissatisfied female customers do not engage in voice as a means of complaint behaviour more often than males consumers. In this case, the researchers’ hypothesis was not supported. However, in scenario two, involving the financial services, it was found that female consumers do indeed use voice as a means of complaining more often than male consumers, that is, the research hypothesis was supported. Possible reasons for the fact that male and female customers tend to complain less in the medical service could be due to feelings of powerlessness; a perception that they are unable to impact the practitioners’ actions or the organisations’ behaviour, or that they will be unable to find another health practitioner after complaining about the previous one. Feelings of gratitude as a hindrance to complaining may have a part to play in the specific context described in scenario one. Reasons why females tend to engage in voice as a complaint behaviour more often than males is that research shows that they tend to be better communicators and more often express their feelings than males (Robbins & Judge, 2005). Why there only seems to be a gender difference in CCB in scenario two, relating to the financial sector, and not in the medical services, is unclear. A possible reason for this dissonance may be the small sample size which has affected the consistency of the results.

75 Companies will find such information as being highly significant for their complaint management strategies. This finding allows companies to target the female population in diverse approaches aimed at service recovery. 9.10.4 Education Level, Income Level and Age The results showed that there was no significant relationship between education level, income or age on the likelihood of engaging in either one of voice, private or third party complaint behaviours between the various groups. However, these results have to be treated with caution as the sample size was too small for significant differences to appear. Consistent with the result by Morganosky and Buckley (1986, in Tronvoll, 2003), age in this research report was not revealed to be a significant demographic variable influencing consumer complaint behavior. The reported non-significant impact of education level and income level, however, is contrast with the rest of the findings by Morganosky and Buckley (1986, in Tronvoll, 2003) and research by Hogarth and English (2002, in Tronvoll, 2003). These authors found that higher income and better educated consumers were considerably more likely than lower income and less educated consumers to partake in complaint action. Theoretical reasons put forward by these researchers for complainers belonging to higher socio-economic groups are that they have greater resources in the form of self-confidence and information in dealing with marketplace dissatisfaction. Also, these consumers may perceive less of risk of humiliation in complaining; they have more purchasing power and can purchase better quality products. Rather, the findings of this research report is consistent with the results of Tronvoll’s (2003) more recent study that established that complainers came predominantly from lower socioeconomic groupings. A change in complaint behaviour over time could be a possible explanation for this result. Nevertheless, since Hogarth and English’s (2002) research was conducted fairly recently, Tronvoll (2003) also refers cultural elements as another possible explanation for the findings which may also explicate the results of this research report. As described in the literature overview, various researchers, such as Furrer et al (2000, in Tronvoll, 2003) and Hui and Au (2001, in Tronvoll, 2003), have noted that cultural differences, such as individualism and collectivism, may affect perceptions of service quality and complaint behaviour.

76 9.10.5 Comparison of Complaint Behaviour Dimensions across Scenarios One and Two The main finding from this analysis was that voice complaint behaviours are the most popular form of complaining for both scenarios. The literature supports this finding. Dissatisfied consumers usually prefer to let the service provider know about the problem, that is, voice out their concerns, and give the company the opportunity to fix it, rather than involving a third party to fix the problem (Singh & Pandya, 1991). Also, voice behaviours, as the primary form of complaining, are indicative of a restrictive and competitive market structure where consumers are faced with a lack of service providers. Hence, they are unwilling to stop using the service provider they are dissatisfied with as the search for another one may be tedious. This type of structure is rampant in South Africa where there is a monopoly amongst services, especially in the banking sector. Thus, consumers often refrain from switching service providers due to the costs and time involved in searching for a new one. Such a finding has beneficial practical implications for companies as it means that dissatisfied consumers are more prepared to voice out their concerns than engage in detrimental negative word-of-mouth or involving consumer associations which would reflect badly on the service provider’s reputation. Companies will hence be allowed adequate opportunities to engage in service recovery through the knowledge of service failure incidents brought to their attention by dissatisfied consumers. 9.11

Conclusion

This chapter discussed the results of the statistical analysis performed to investigate the relationship between language and complaint behaviour, as well as other relevant consumer demographics. It was found that language, race and gender are consumer demographics that have an impact on CCB. In an analysis of the relationship between language and complaint behaviours, it was found that English- and Xhosa-speaking consumers complain more extensively than their Afrikaans counterparts when dissatisfied. The researchers suggest that culture is not the only variable affecting this relationship and the socio-economic status of these groups may also have a role to play in these findings. In an analysis of the relationship between race and complaint behaviours, it was found that White consumers complain the least compared to Indian and Black consumers. As discussed before, this finding may be due to historical advantage that was accorded to White consumers

77 such that they may be more willing to switch service providers than complain and seek redress. When the impact of gender on complaint behaviours was investigated, it was found that gender only had a role to play in the second scenario, that is, in the financial services sector. Women were found to voice their complaints more often than men. The relationship between income, education and age was not conclusive. The researchers attribute this to the small sample size. The small sample size may have a role to play.

CHAPTER TEN: CONCLUSIONS AND RECOMMENDATIONS The purpose of this chapter is to provide an outline of the research conducted, its implications for stakeholders, as well as recommendations for follow up-studies. 10.1

Introduction

78 The main purpose of this study was to explore the cross-cultural differences in response to consumer dissatisfaction in the city of Port Elizabeth in the Eastern Cape, with regard to the services sector. The central aim of the research was to determine how cultural elements of the three main language groups in Port Elizabeth; Afrikaans, English and Xhosa, affect their behaviour toward dissatisfaction in the services sector, focusing specifically on the medical and financial segments. Three possible responses of consumer complaint behaviour were identified; voice, private, and third party actions and the likelihood of various ethnic identities engaging in such behaviours researched. Other variables, which may also affect consumers’ propensity to engage in voice, private, and third party actions, such as age, education level, income level, gender and race were also investigated. Based on theoretical research, it was hypothesised that: 1. Afrikaans-speaking customers are more likely to engage in voice behaviours (seeking redress from the seller) than Xhosa-speaking customers when dissatisfied. 2. When dissatisfied, English-speaking customers are more likely to engage in voice behaviours (seeking redress from the seller) than Xhosa-speaking customers. 3. When dissatisfied, Xhosa-speaking customers are more likely to engage in private complaint behaviours (word-of-mouth communication or exit) than Afrikaansspeaking consumers. 4. Xhosa-speaking consumers are more likely to engage in private complaint behaviours than English-speaking consumers when dissatisfied. 5. When dissatisfied, female consumers are more likely to engage in voice as a means of complaint behaviour than males consumers. The study comprised of in-depth interviews and a main study. In-depth interviews were carried out to determine the type of services various ethnic groups in Port Elizabeth feel most strongly about. Respondents were provided with a list of services and were required to choose two types of services that they had the highest expectations of, in terms of service quality. The two most frequently selected services were found to be medical and insurance services. The main study then aimed to explore cross-cultural differences in terms of CCB with regard to these two identified service segments.

79 A sample of 150 was drawn using non-probability convenience sampling. The proxy of culture that was studied was home language, namely Afrikaans, English and Xhosa. These are the three main language groups in Port Elizabeth. The researchers decided that the sample size should be more or less equally divided between the three language groups being investigated. Before embarking on data collection, the researchers familiarized themselves with fieldwork techniques and considerations as provided by Malhotra (2006). The data was analysed using descriptive and inferential statistics. ANOVA, Cohen’s D and Cronbach’s alpha, were utilised to analyse the data inferentially in terms of the findings’ practical and statistical significance. The measure was found to be reliable, with Cronbach’s alpha being satisfactory for almost all the items. Validity… 10.2

Summary of Findings

The following section summarises the main findings of the study. 10.2.1 Impact of first language on voice, private and third complaint behaviour can be summarised as: a. Dissatisfied Afrikaans-speaking customers are less likely to partake in voice complaint behaviours than Xhosa-speaking customers, and this relationship has moderate practical significance. b. English-speaking consumers are as likely as Xhosa-speaking consumers to engage in voice complaint behaviours when dissatisfied, and this relationship has large practical significance. c. Xhosa-speaking consumers are more likely to engage in private complaint behaviours when dissatisfied than the Afrikaans-speaking dissatisfied consumers, and this relationship has moderate practical significance. d. Both the Xhosa- and the English-speaking consumers are just as likely to engage in private complaint behaviours when dissatisfied. 10.2.2 The impact of race on voice, private and third complaint behavior were found to be: a. In scenario one and two, dissatisfied Indian consumers are more likely to engage in voice complaint behaviours than dissatisfied White consumers. This difference has moderate practical significance.

80 b. Black consumers are more likely to engage in private complaint behaviours than White consumers when dissatisfied in scenario one. c. Indian consumers are more likely to engage in private complaint behaviours than White consumers when dissatisfied in both scenarios. In the first scenario this relationship has large practical significance and moderate practical significance in the second scenario. d. In both scenarios, black consumers are more likely to engage in third party complaint behaviours than White consumers when dissatisfied and this relationship has moderate practical significance. e. In scenario one and two, Indian consumers are more likely to engage in third party complaint behaviours than White consumers when dissatisfied. In the first scenario this relationship has large practical significance and moderate practical significance in the second scenario. 10.2.2 The impact of gender, income level, education and age on voice, private and third complaint behaviour were found to be: a. There was no relationship between voice, private and third complaint behaviour and income level, education and age. b. In scenario one, relating to the medical service sector, results showed that dissatisfied females do not engage in voice as a means of complaint behaviour more often than males, while in scenario two, relating to financial services, the findings were the converse. 10.3

Implications

This chapter will provide an outline of the research that was conducted, offer recommendations for stakeholders and make suggestions for future research. The main results from the research will be summarised. Implications of these findings are discussed and practical proposals are made to overcome the identified obstacles. Limitations of the research report will be noted as well as a scope for follow-up studies suggested. 10.3.1 Language

81 The English- and the Xhosa-speaking consumers were found to engage extensively in private complaint behaviour. As stated before, this may be due to poor perceptions of service recovery on behalf of the service provider where consumers have little confidence in the ability of the company to take care of the problem. Companies need to take a three-pronged approach in dealing with this problem. The aim of the complaint management strategy should be to tailor-make any marketing approach according to the cultural practices of the language group being targeted. Therefore, companies wishing to reduce negative word-of-mouth in the Xhosa community should use Xhosa role models in any advertising or informational approach together with the use of the native language. The same should be done for English-speaking consumers. Secondly, one of the most effective ways in which to manage the negative word-of-mouth of these two language groups is to provide more informational marketing strategies aimed at dispelling previous harmful beliefs about the services (Mangold et al., 1999). Marketers should aim at providing consumers with independent information about their services from a number of sources that a particular language group finds reliable. Therefore, marketers should research the sources of information that the Xhosa-speaking consumers would find to be reliable, for example, and channel information about the services through that medium. Examples of channels are magazines, religious brochures and the Internet. It is essential that the information is presented in a captivating and vivid way, where the emotional content is high, such that this information over-rides other negative word-of-mouth beliefs. Thirdly, companies need to have more effective service recovery strategies in place. All service providers, no matter how quality-driven, will come across situations where there are failures in service quality vis-à-vis their customers (Lewis & Spyrakopoulos, 2001; 1997, Boshoff in Johnston, 2001). The actions that a service provider takes in response to service failures are known as service recovery. The latter can be further defined as “the specific actions taken to ensure that the customer receives a reasonable level of service after problems have occurred to disrupt normal service” (1995, Armistead, Clarke & Stanley in Lewis & Spyrakopoulos, 2001). The reason why English- and Xhosa-speaking consumers are heavily engaging in private complaint behaviours, despite also following other complaint routes, is probably due to the fact that these consumers are not satisfied with service recovery strategies in place by the company. Therefore, service providers need to conduct a comprehensive audit

82 on their service recovery strategies and improve such policies. Such strategies will have a beneficial impact on customer satisfaction and retention (Johnston, 2001). Complaint management will also result in better operational functioning by highlighting key areas that need to be improved. Companies need to be sensitive to the cultural differences in complaint behaviour, as previously discussed. This paper highlighted the groups of people who are more prone to private behaviour, and hence negative word-of-mouth. Companies should structure their service recovery policies in a way that is appropriate for these specific groups of people, for example by communicating to them in their own language. They should target English- and Xhosa-speaking consumers, and prioritise service quality to those sections of the population, such that the latter receive optimal service delivery. Also, through marketing strategies such as advertising and special benefits targeted to those sensitive communities, companies can help restore positive relations with them. An example of such community activities includes going into the Xhosa townships and offering social benefits whilst talking positively of the services that a company offers. Companies need to further research the best ways to approach these target communities and appeal to them positively. 10.3.2 Race The recommendations for companies about the White consumers are the same as for the Afrikaans-speaking consumers. This is because they both display the same type complaint pattern, that is, minimal complaining compared to other communities. Recommendations for the Indian community are the same as for the English- and Xhosaspeaking consumers as the patterns of complaint are also similar. The results also suggested that Black consumers may not be fully aware of their rights as consumers, and hence may not have a holistic understanding of their options when faced with service failure. This may make them vulnerable to exploitation from companies. Consumer associations and other South African non-profit organisations should approach the Black consumer market in order to educate them about their options when dissatisfied. These organisations need to boost the Black consumer’s confidence levels so that he/she can appeal to various third parties for help when a relevant company is unwilling to fix a problem.

83 10.3.3 Gender The findings of the study revealed that women engaged in voice complaints more often than men do in the scenario referring to financial services. 10.4

Limitations of the study

The sample size of this study was small and thus limited the accuracy of the findings. The researchers were unable to augment the sample size due to resource constraints. However, this limitation was addressed by having a minimum number of participants per language group, which was the main concern of the study, in order to improve the accuracy of this particular analysis. Thought the measure was found to be reliable for exploratory purposes, the researchers believe that the reliability of the questionnaire was compromised by eliminating an item from the voice complaint dimension. By doing so, Cronbach’s alpha for that particular section of the measure was lowered, affecting the consistency of the measure. 10.5

Scope for future research

The measure should be further tested and improved in terms of its reliability, by adding more items under each of the complaint behaviour dimensions, that is, voice, private and third party behaviours. This study is a description of the various relationships described between consumer demographics and complaint behaviours. Further qualitative research should be conducted in order to investigate the actual causes behind the findings of this study. One area for such research could, for example, determine the reason why Afrikaans-speaking consumers are so unwilling to complain, compared to other language groups. The researchers noted that, despite having done their utmost best to construct a simple and concise questionnaire, Black respondents were often confused about the meaning of the items in the measure. The researchers suggest that, for future similar research, questionnaires should be available in both English and a native language to ensure that Black respondents understand and answer the questions correctly. 10.6

Conclusion

84 The main results of the research were summarised and the implications of these findings were discussed so that practical proposals could be made. Such practical suggestions can be of imperative value, as not only would it allow companies to overcome some obstacles identified in managing consumer complaint behaviour, but it can also reveal existing and potential problems in consumer complaint management that a corporation may be unaware of. It was shown that it is vital that firms are sensitive to cultural differences, in particular the gender, race and first language variables of their consumers, as it affects CCB, and can ultimately impact corporate profit and prosperity. Limitations of the research report, however, were identified, most notably the small sample size used due to lack of resources. This implies improvements for follow-up studies, and in junction with improving the reliability of the research instrument, it is recommended that qualitative research is conducted to scientifically determine reasons for the findings rather than solely relying on theoretical speculation.

85

REFERENCES Andreassen, T. W. (1998). Antecedents to Satisfaction with Service Recovery. European Journal of Marketing, 34 (1/2), 156-175. Bendall-Lyon, D. & Powers, T. L. (2002). The Impact of Gender Differences on Change in Satisfaction Over Time. Journal of Consumer Behaviour, 19 (1), 12-23. Bodey, K. & Grace, D. (2006). Segmenting service complainers and non-complainers on the basis of consumer characteristics. Journal of Services Marketing, 20 (3), 178-187. Colgate, M & Lang, B. (2001). Switching barriers in consumer markets: an investigation of the financial services sector. Journal of Consumer Marketing , 18 (4), 332-347. Communal, C. & Senior, B. (1999). National Culture and Management: Messages conveyed by British, French and German Advertisements for Managerial Appointments. Leadership & Organisation Development Journal, 20 (1), 26-36. Coyle, J. & Williams, B. (1999). Seeing the wood for the trees: defining the forgotten concept of patient dissatisfaction in the light of patient satisfaction research. International Journal of Health Care Quality Assurance incorporating Leadership in Health Services, 12 (4), iix. Crié, D. (2003). Consumers’ Complaint Behaviour - Taxonomy, Typology and Determinants: Towards a Unified Ontology. Database Marketing and Customer Strategy Management, 11 (1), 60-79. Davies, M. (1996).

Image problems with financial services: some considerations for

improvement. Journal of Management Decision, 34 (2), 64-71.

86 Dolinsky, A. (1994). A Consumer Complaint Framework with Resulting Strategies: An Application to Higher Education. Journal of Services Marketing, 16 (4), 345-354. Francesco, A. M. & Gold, B. A. (2005). International Organisational Behaviour. New Jersey: Pearson Prentice Hall. Johnston, R. (2001). Linking Complaint Management to Profit. International Journal of Service Industry Management, 12 (1), 60-69. Im, S., Kim, S., Kim, C. & Shin, C. (2003). The effect of attitude and perception on consumer complaint intensions. Journal of Consumer Marketing, 20 (4), 352-371. Kim, J. O., Forsythe, S., Gu, Q. & Moon, S. J. (2002). Cross-cultural Consumer Values Needs and Purchase Behaviour. Journal of Consumer Marketing, 19 (6), 481-502. Lerman, D. (2006). Consumer politeness and complaining behaviour. Journal of Services Marketing 20 (2), 92-100. Lewis, B. R. & Spyrakopoulos, S. (2001). Service Failures and Recovery in Retail Banking: The Customers’ Perspective. International Journal of Banking, 19 (1), 37-47. Liu, R. R. & McClure, P. (2001). Recognising Cross-cultural Differences in Consumer Complaint Behaviour and Intentions: An Empirical Examination. Journal of Consumer Marketing, 18 (1), 54-74. Lopez, J. Hart, L.K. & Rampersad, A. (2007). Ethnicity and customer satisfaction in the financial services sector. Journal of Managing Service Quality, 17 (3), 259-274 Luna, D. & Gupta, S. F. (2001). An Integrative Framework for Cross-cultural Consumer Behaviour. International Marketing Review, 18 (1), 45-69. Malhotra, N. K. (2006). Marketing Research: An Applied Orientation (5th ed.). New Jersey: Person Prentice Hall.

87 Mangold, W. G., Miller, F. & Brockway, G. R. (1999). Word-of-mouth Communication in the Service Marketplace. The Journal of Services Marketing, 13 (1), 73-89. Mattila, A. S. & Wirtz, J. (2004). Consumer Complaining to Firms: The Determinants of Channel Choice. Journal of Services Marketing, 18 (2), 147-155. McCole, P. (2004). Dealing With Complaints in Services. International Journal of Contemporary Hospitality Management, 16 (4), 345-354. McFarlin, D. B. & Coster, E. A. (1999). South African Management Development in the Twenty-first Century. Journal of Management Development, 18 91), 63-78. Nyer, P. U. (2000). An Investigation into Whether Complaining Can Cause Increased Consumer Satisfaction. Journal of Consumer Marketing, 17 (1), 9-19. Phau, I. & Sari, R. P. (2004). Engaging in Complaint Behaviour. Marketing Intelligence & Planning, 22 (4), 407-426. Robbins, S.P & Judge, T.A. (2005). Organizational Behaviour. New Jersey: Pearson Prenctice Hall. Samuel Craig, C. & Douglas, S. P. (2006). Beyond national culture: Implications of cultural dynamics for consumer research. International Marketing Review, 23 (3), 322-342. Singh, J. & Pandya, S. (1991). Exploring the Effects of Consumers’ Dissatisfaction Level on Complaint Behaviours. European Journal of Marketing, 25 (9), 7-21. Singh, J. (1988). Consumer Complaints Intentions and Behaviour: Definitional and Taxonomical Issues. Journal of Marketing, 52, 93-107. Snellman, K. & Wihtkari, T. (2003). Customer Complaining Behaviour in Technology-based Service Encounters. International Journal of Service Industry Management, 14 (2), 217231.

88 Statistics SA. (2005). Census 2001. Available: http://www.statssa.gov.za/PublicationsHTML/Report-03-02-052001/html/Report-03-02052001.html. (accessed 2 September 2007) Sun, T., Horn, M. & Merritt, D. (2004). Values and Lifestyles of Individualists and Collectivists: A Study on Chinese, Japanese, British and US Consumers. Journal of Consumer Marketing, 21 (5), 318-331. Tronvoll, B. (2006). Complainer characteristics when exit is closed. International Journal of Service Industry Management, 18 (1), 25-51. Warden, C. A., Liu, T. & Huang, C. (2003). Service Failures away from Home: Benefits in Inter-cultural Service Encounters. International Journal of Service Industry Management, 14 (4), 436-457. Warren, W. E. & Gilbert, F. W. (1993). Complaint Intentions and Behaviour for Products versus Services. The Journal of Marketing Management, Spring/ Summer, 12-22. White, L & Yanamandram, V. (2004). Why customers stay: reasons and consequences of inertia in financial services. Journal of Management Service Quality, 14 (2/3), 183-194. Venter, D. (2006). WSD400 Class Notes. Unpublished class notes. Nelson Mandela Metropolitan University.