Vous êtes sur la page 1sur 24

8.

Introduction

This chapter commences with a brief summary of the research, and followed by discussions on the research questions based on the empirical research evidence drawn from Chapter 6 and 7. Consequently, conclusions are drawn by reflecting upon the Structural Equation Modelling which was employed in this study in order to assess the research hypotheses and to generate a plausible research model.

8.2

Research Summary

The main objective of this research endeavour was to identify and examine the key determinants that influence customer satisfaction judgements and their subsequent behavioural outcomes in a consumption system setting. In order to accomplish this objective, valid and reliable multidimensional measures had to be established and validated (Churchill 1979; Ping 2004). With this in mind, the measures were stringently assessed and validated by Item Analysis and Exploratory Factor Analysis (EFA), procedures suggested by Churchill (1979) and Confirmatory Factor Analysis (CFA), following recommendation by Anderson and Gerbing (1988). Therefore, it is reasonable to claim that the measures employed to test the hypothesised relationships among constructs as postulated in the conceptual model have successfully achieved unidimensionality, validity and reliability. Ultimately, the thesis primary goal of generating a plausible model that could be characterised as having statistical and explanatory power (McQuitty 2004) which permits interpretation of results confidently was accomplished after minor modification to the proposed model (see Section 7.6.3).

The proposed model adopts a broader conceptualisation of satisfaction by incorporating the entire consumption process, which comprises the evaluation of product, direct seller and company performance which were measured at attribute, dimension/subsystem (micro-level) and overall (macro) levels. This conceptualisation is consistent with Oliver (1993) and Rust et al. (1995), who argue that overall and attribute satisfaction must be measured because there is evidence that they are distinct but related constructs. Therefore, it was recommended that they should be assessed separately (Mittal et al. 1999)

The Expectancy-Disconfirmation Paradigm has been widely applied to account for postpurchase response in the consumer satisfaction literature (e.g. Churchill and Suprenant 334

1982; Oliver 1980, 1993; Halstead et al. 1994; Oliver and DeSarbo 1988; Westbrook 1987; Westbrook and Oliver 1991), and is the theoretical foundation for the current investigation.

The emphasis on assessing customer satisfaction goes far beyond pure academic interest (Westbrook 1997, 2000; Woodruff 1993). It appears that businesses have focused large amounts of attention and resources on the construct because of the presumption that satisfaction will precede certain behavioural outcomes such as word of mouth recommendations, repeat purchase, increased spending, and increased price insensitivity (Anderson 1996; Kassim 2001; Sderlund 1998; Zeithmal et al. 1996). Furthermore, it has been empirically revealed that customer satisfaction is an essential factor in desired outcomes such as loyalty, customer retention and financial profitability (Rusk and Zahorik 1993; Rusk et al. 1995; Strong 2000). In response to this notion, it was observed that there has been growing concern and recognition of the importance of understanding postpurchase consumer evaluation, particularly customer satisfaction, this being the focal construct in most post-purchase investigations (Babin and Griffin 1998).

Realising the importance of understanding how customers formed satisfaction judgements and the outcome of satisfaction in the direct sales environment, this research took the lead to advance the study of customer satisfaction by taking a consumption system approach.

8.3

Discussion of Research Questions

This thesis specifically seeks to address the following research questions, which were presented in Chapter 1 (see Table 1.2, p.20). The major findings will be highlighted and discussed accordingly. It is important to highlight that some of these research questions are associated with the hypotheses tested in Part II of Chapter 7. The main distinction is that in testing the hypotheses, the researcher attempts to confirm/disconfirm whether the hypothesised linkages are supported; hence, emphasis was devoted to the examination of each path. On the other hand, in addressing the research questions, the researcher is interested in investigating the interrelationships among constructs, which involves examining several pertinent constructs simultaneously, and in certain cases the purpose is to elucidate specific issues or concerns. Additionally, by responding to the research questions, the researcher is able to explicitly highlight the main contributions of this 335

study, as the comprehensive model is decomposed into smaller sub-models. In some of the figures which are used to visually represent the linkages among constructs for the research questions, the related hypotheses are exhibited in the boxes. What are the key attributes and factors that influence customers satisfaction judgements and which of the attributes have the greatest impact on each pertinent factor that constitute a direct sales consumption system?

(1)

As previously indicated, the direct sales consumption system is recognised to comprise of three main components (subsystems), which are the product, direct seller and direct selling company, and in turn, each subsystem is comprised of two factors, except the direct seller subsystem, which is a one factor construct. However, prior studies, for example Mittal et al. (1998) provide evidence that assessment of performance at attribute level is essential in the attempt to evaluate customer satisfaction. It was argued that marketers generally make strategic decisions at the attribute level (Mittal et al. 1998; Westbrook 1997). Similarly, customers are more likely to render evaluations of their postpurchase satisfaction experiences at an attribute level rather than at the product/service level (Westbrook and Oliver 1991). Indeed, it was noted that this was the main reason why most applied satisfaction studies frequently assess determinants of satisfaction (e.g. performance, expectation, importance, and disconfirmation) at the attribute level rather than at the product or dimension level. Parallel to these arguments, the current study assesses customer satisfaction at each subsystem abstraction by analysing it based on the attribute-level performance (Allen and Rao 2000; Mittal et al. 1999).

In this section, the answer to each research question is addressed and discussed explicitly when appropriate. The findings of the research questions are based on the validated measurement model which has been assessed via CFA (see Section 7.4 and 7.5) and the assessment of the research hypotheses (see Section 7.6). It is essential to highlight that, in order to determine the magnitude of impact on the outcome variable, such as satisfaction or behavioural intentions propensity, the diagram is presented with the value of the regression coefficient, which is shown between each significant path. Clearly, this mechanism could provide valuable knowledge to both researchers and managers, with road maps for maximising an outcome variable such as satisfaction at subsystem or global abstractions (Allen and Rao 2000, p. 70). 336

(i)

Product Satisfaction

The product component is comprised of two factors, Product Quality and Product Offerings and Information. Table 8.1(a) exhibits that the Product Quality factor is composed of four attributes, whereas Product Offerings and Information constitutes three indicators. The importance (i.e. magnitude of impact) of the attributes is determined by their corresponding standardised regression coefficient values, as suggested by Hayes (1998) and Mittal et al. (1998). A higher value indicates that the specific attribute is an important predictor to the factor and to customer satisfaction or any other given outcome variable (Allen and Rao 2000). As shown in Table 8.1 (a), in the Product Quality construct, the highest standardised loading is effectiveness of product (0.81), hence it is considered the key driver for Product Quality in the product subsystem.
Table 8.1(a) The product aspect of the direct sales consumption system
Constructs Product Quality 1. Effectiveness of product 2. Product performed as claimed 3. Product guarantee 4. Product is multifunctional Product offerings and information 1. Product information adequacy 2. Product catalogue / brochure 3. Innovative and unique product
Standardised Regression (Loading)

0.81 0.80 0.68 0.64

0.73 0.64 0.57

In this regard, in order to establish a sound brand, the effectiveness of the product is unquestionably important to consumers, especially when the direct seller or the company has made claims about it. Generally, the importance of maintaining an above-standard product quality can never be over-emphasised. Product information adequacy is the most influential predictor (0.73) for the Product Offerings and Information factor in the same subsystem. In essence, the direct sales industrys resources will be best served if it focuses on improving product effectiveness and provides sufficient information pertaining to the products marketed. Improvement in these product criteria might be reflected in increases in customer satisfaction scores. Furthermore, with the rising level of standard of living and purchasing power, consumers today demand quality and value products, and in

337

return, they are willing to pay a premium. Bartlett supports this notion (1994, p. 1) by asserting that: The future of retailing lies in delivering genuine value to customerssurvival of retail enterprises will depend on selling the right products at the right prices in the right way. Poorly made products are outValue is in; value in retailing is everything.

(ii)

Direct Seller Satisfaction

The direct seller satisfaction component is a one-factor construct, which is comprised of eight attributes. Table 8.1(b) exhibits the direct seller construct with eight attributes.

Table 8.1 (b) The direct seller aspect of the direct sales consumption system
Constructs Direct Seller performance 1. Maintaining a professional appearance 2. Have customer interest at heart 3. Continuity of contact 4. Availability of direct seller 5. Being consistently courteous 6. Giving personal advice and attention 7. Knowledge of products and services 8. Effectiveness of sales demonstration
Standardised Regression (Loading)

0.78 0.78 0.72 0.72 0.68 0.67 0.65 0.63

The importance of the attributes is determined by their standardised regression values (Mittal et al. 1999). As depicted in Table 8.1 (b), both maintaining a professional appearance and have customer interest at heart exhibit the highest standardised loadings (0.78), hence they are considered as the most important predictors of Direct Seller Satisfaction. In addition, continuity of contact and availability of direct seller are regarded as the second most important predictors of Direct Seller Satisfaction. Fascinatingly, these two closely related items were revealed to be the most important findings by Raymond and Tanner (1994), which demonstrate their crucial role for the direct seller in generating repeat sales. In essence, if companies emphasise these factors (e.g. maintaining a professional appearance and have customer interest at heart) to direct sellers, it might possibly translate into higher customer satisfaction toward the direct seller.

338

(iii)

Direct Selling Company Satisfaction

The direct selling company component consists of two factors, Corporate Image and Corporate Customer Service Table 8.1(c) shows that both Corporate Image and Corporate Customer Service factors are comprised of three indicators. As indicated in Table 8.1 (c), corporate information/publicity (0.79) is the key driver for Corporate Image, while handle complaints promptly (0.83) is the most influential driver for the Corporate Customer Service measure.

Table 8.1 (c)

The company aspect of the direct sales consumption system


Standardised Regression (Loading)

Constructs Corporate Image 1. Corporate information / publicity 2. Sales campaigns / promotion 3. Companys reputation Corporate Customer Service 1. Handle complaints promptly 2. Concern about customers 3. Product return / adjustment policy

0.79 0.77 0.62

0.83 0.81 0.67

As described earlier in Chapter 2, even though customers seldom make direct contact with the direct selling companies, under certain circumstances they have to deal directly with the company, such as during sales campaigns and promotions, to request product information/ service, product returns and complaints about products and direct sellers. Therefore, just like other conventional companies, customer care that provides an efficient mechanism of responding to customers might help to generate stronger customer relationships and subsequently garner greater profitability (Rust et al. 1996). This notion was supported by empirical evidence in Strongs (2000) study; it was found that customer care strategies have a significant impact on market performance (market share, customer satisfaction, customer retention, sales volume and return on investment). In conclusion, it could be assumed that companies that allocate more resources and effort toward improving the key indicators (e.g. corporate information/publicity and handle complaints promptly), will experience a greater likelihood of customer satisfaction toward the direct selling company.

(2)

What are the determinants and consequences of product, direct seller and direct selling company satisfaction?

339

There is a rich body of literature concerning the determinants of customer satisfaction judgments (e.g. Mittal et al. 1998; Oliver 1980; Spreng et al. 1996). Such studies focus upon the impacts of various response determinants (e.g. performance, expectation, disconfirmation, equity, etc.) on satisfaction judgments, and report that these impacts on satisfaction are varied and are often dependent upon the product/service class under investigation (see Appendix 3.1 for a comprehensive summary of major research findings on response determinants). Markedly, these studies contribute significantly to the understanding of the salience of the various response determinants on overall customer satisfaction judgments. In the current study, both tangible products and service were integrated in one coherent framework (see Figure 4.2, p. 134). Hence, in order to respond to this research question, it is possible to decompose the comprehensive framework into three subsystems, which are product satisfaction, direct seller satisfaction and company satisfaction.

(i)

Product Satisfaction

Figure 8.1 (a) illustrates that Product Performance is the only significant determinant of Product Satisfaction and the consequence of Product Satisfaction is customer Overall Satisfaction with the direct sales channel.
Figure 8.1 (a) Determinant and Consequence of Product Satisfaction

Purchase Decision Involvement (H17)


Disconfirmation

Product Performance (H8)


0.74****

Product Satisfaction (H14)


0.85 ****

0.17 ***

Overall Satisfaction

Product Disconfirmation (H 5)

Note: **** Significant at p< 0.001 (t> 3.29) *** Significant at p< 0.01 (t> 2.57)

Non-significant path

340

The diagram above clearly illustrates that product performance affects only product satisfaction directly. The effects of product disconfirmation and purchase decision involvement are disregarded as these links are not significantly related to product satisfaction. In summary, product performance is the most influential determinant of customer satisfaction with the product and subsequently it has a direct significant impact on overall satisfaction with the direct sales channel.

(ii)

Direct Seller Satisfaction

Figure 8.1 (b) demonstrates that Direct Seller Performance and Direct Seller Disconfirmation are the key determinants of Direct Seller Satisfaction and in turn the consequences of Direct Seller Satisfaction are Relational Commitment and Overall Satisfaction.
Figure 8.1 (b) Determinant and Consequence of Direct Seller Satisfaction

Perceived Equity (H23)

Overall Satisfaction (H13) Direct Seller Satisfaction


0.91**** - 0.11* 0.40 **** 0.38 **

Direct Seller Performance (H10)


0.19 *** Direct Seller

Relational Commitment (H25)

Disconfirmation (H4)
Note: **** Significant at p< 0.001 (t> 3.29) *** Significant at p< 0.01 (t> 2.57) ** Significant at p< 0.05 (t> 1.96)

Significant at p< 0.10 (t> 1.65) Non significant path

The diagram above clearly portrays that direct seller performance directly affects only direct seller satisfaction. In summary, the direct seller performance is the key determinant of customer satisfaction with the direct seller and in turn, the consequences of direct seller satisfaction are relational commitment and overall satisfaction.

341

(iii)

Direct Selling Company Satisfaction

Figure 8.1 (c) demonstrates that Company Performance is the only determinant of Direct Selling Company Satisfaction and consequently it does not have a significant effect on Overall Satisfaction.
Figure 8.1 (c) Determinant and Consequence of Direct Selling Company Satisfaction
Company Performance (H9)
0.13** 0.84****

Company Satisfaction (H15)

Overall Satisfaction

Company Disconfirmation (H6)

Note: **** Significant at p< 0.001 (t> 3.29) ** Significant at p< 0.05 (t> 1.96)

Non-significant path

The findings of this research question suggest that the process by which consumers arrived at satisfaction judgements on each subsystem level were heavily influenced by perception of performance. These findings concur with Oliver and DeSarbo (1988); Spreng and Olshavsky (1991); Swan (1988) and Tse and Wilson (1988), who suggest that when a product performs well, the consumer is likely to be satisfied regardless of disconfirmation evaluation. Interestingly, disconfirmation did not have a significant effect on satisfaction except for the direct seller disconfirmation-direct seller satisfaction link, although disconfirmation has a significant effect but in a negative direction. As for the negative disconfirmation, this could be explained by the fact that the customers have higher expectation, and when the performance of the direct seller falls short (expectation not met), this will cause them to feel less satisfied (Oliver 1980). An alternative explanation concerning why performance is the key predictor of customer satisfaction was suggested by Swan (1988) and Yi (1993). They contend that performance is strongly correlated with satisfaction with the product (e.g. food) and on the contrary, disconfirmation is a significant predictor for service satisfaction. It was argued that products may be less ambiguous (i.e. easy to evaluate), while the service dimension (e.g. sales persons trustworthiness) is relatively ambiguous, and thus 342

difficult to judge. Perhaps, based on this contention, the results of the current investigation revealed that only direct seller disconfirmation was significantly related to customer satisfaction. Additionally, several scholars suggested (e.g. Churchill and Suprenant 1982; Oliver 1997; Swan 1981) that the effect of relationships between variables in satisfaction models is idiosyncratic to the characteristics of the product (e.g. durable or high involvement items) being investigated. It is reasonable to believe that the result of this study might be determined by the product category investigated. In fact, prior research has demonstrated that performance evaluation is the only determinant to satisfaction for high involvement product (Hong and Rucker 1995; Oliver and Bearden 1983). It could be logical to assume that the healthcare and beauty care products, which were the focal products in this study, are considered high involvement products. In light of these assertions, this clearly explains why performance is a significant determinant of satisfaction judgements at subsystem level. In conclusion, performance is significantly related to disconfirmation for each subsystem, which in line with Cardotte et al. (1987) and Oliver (1980). Noticeably, performance is the key determinant of customer satisfaction at the subsystem level (product satisfaction, direct seller satisfaction and direct selling company satisfaction).

(3)

How do product satisfaction, direct seller satisfaction and direct selling company satisfaction judgements aggregate to form a global (overall) satisfaction and which of these aspects is the most predictive of the customers overall satisfaction with the direct sales marketing system?

As portrayed visually in Figure 8.2, both product and direct seller satisfaction have been established to be significant determinants of customers overall consumption satisfaction, but direct selling company satisfaction was found to be a non-significant predictor of the overall satisfaction judgement.

343

Figure 8.2

Determinants of Overall Satisfaction

Product Satisfaction (H14)


0.17 ***

Direct Seller Satisfaction (H13) Direct Selling Company Satisfaction (H15)


Note: *** Significant at p< 0.01 (t> 2.57)

0.38***

Overall Satisfaction

Non significant path

One notable finding is that the direct seller satisfaction has the strongest influence (0.38) on overall satisfaction, compared to product satisfaction, which is supposed to be the core of the exchange transaction. This implies that the direct seller is essential in influencing customer satisfaction in the direct sales channel, and further demonstrates the significance of personal contact in the direct sales industry. It is conceivable to compare these results with Crosby and Stephenss (1987) empirical evidence in the insurance industry. They tested two alternative models: the Rational Evaluation Model (REM) and the Relational Generalization Model (RGM). The REM hypothesised that satisfaction with the core product drives satisfaction with peripheral services; in contrast, the RGM proposed that peripheral services (i.e. the insurance agent) drive satisfaction with the core product (i.e. the insurance policy). In their investigation, insurance policy was found to drive satisfaction with the contact person (insurance agent), thus supporting the REM model. Conversely, the findings of the current study support Crosby and Stephenss (1987) RGM model.

It is could be speculated that, through continuous sales and services, the direct seller can develop a long-term relationship with their clients, particularly in this case, where it was revealed that majority of respondents were experienced customer on the direct sales channel. (see Section 6.4). Perhaps, over time the exchange encounters have moved to a social rather than economic basis (Beatty et al. 1996; Frenzen and Davis 1990). Hence, the respondents satisfaction with the direct seller drives their satisfaction with the product. Clearly, this result indicates that respondents evaluation of their overall satisfaction with the direct sales channel has been moderately determined by the direct 344

seller performance evaluation, whereas the product judgement shows a very weak effect. Conversely, this phenomenon might not hold in the situation when the purchase encounter was with a first time customer, who might be less familiar with the direct seller and product (Frenzen and Davis 1990).

It is interesting to reflect upon the in-depth interviews which were conducted during the exploratory phase of this study. A participant substantiated the above contention anecdotally. She admitted that, she is loyal to a particular beauty care product distributed by a direct selling company. However, if the specific beauty consultant who had provided her with regular services for several years was not available anymore, she would most probably not patronise the beauty salon again, even though there might be another beauty consultant available to offer the same service with a similar brand . She proudly mentioned that the particular beauty consultant came to her wedding reception. She observed that she had been given much personal attention, a tremendous amount of information pertaining to this specific product, as well as a special pricing plan.

Implicit in her comment, it could be concluded that one approach to developing commitment between buyer and seller is to focus on relationship-building activities. Obviously, this quality of relationship is very valuable for generating future sales even when the client is dissatisfied with the product and level of service. Undoubtedly, this strong relationship acts as a switching barrier, which represents an important strategy for customer retention (Jones et al. 2000; Patterson and Smith 2003). Indeed, it has been established that satisfaction with the seller could be a very productive mechanism in diminishing the customers discontent (Priluck 2003). Interestingly, these findings are consistent with Merrilees and Millers (1999) empirical investigation in a comparative study between two countries (Australia and China) concerning direct selling effectiveness; they found that in China (i.e. Eastern culture) direct sellers evaluate that relationship marketing factors elicit a greater magnitude than product factors in determining their success, which is contrary to the evaluation of their Australian (i.e. Western culture) counterparts. Similarly, Malaysians generally embraces Eastern culture and typically belong to a collectivist society, described as one in which from birth, they are integrated into strong, cohesive in-groups and they are 345

characterised as a relationship rich society

(cf. Patterson and Smith 2003). Perhaps

for this reason, the respondents of this survey judged the direct seller to be the most important aspect in the direct sales system, as compared to the product or the direct selling company.

However, it could be argued that when relationship marketing is the dominant strategy in an industry, buyers may feel satisfied for the wrong reasons; they become enamoured with the service rendered to them and failed to judge the product rationally (Lehtinen 1985). In other words, buyers generalise positive feelings with the salesperson who is the provider of the product to the core product. However, the author agrees with Crosby and Stephenss (1987) views that it is rather nave to assume that customers do not validate their relationship occasionally and demand evidence of product benefits to justify long-term commitments to the direct seller or the company. To conclude, the results indicate that customers satisfaction with the direct seller is the most predictive determinant of their overall satisfaction with the direct sales channel.

(4)

What are the determinants and consequences of overall satisfaction and how are these determinants interrelated to each other to provide an explanation of customers overall satisfaction with the direct sales channel?

Figure 8.3 clearly illustrates that perceived value is the most influential determinant of customers overall satisfaction with the direct sales channel and in turn h as a significant impact on their behavioural intentions.
Figure 8.3 Determinants and Consequence Overall Satisfaction
Product 0

Satisfaction (H14)
0.17 ***

Direct Seller Satisfaction (H13) Direct Selling Company Satisfaction (H15) Perceived Value (H20)
Note: *** Significant at p< 0.01 (t> 2.57) **** Significant at p< 0.001 (t> 3.29)

0.38 ***

Overall Satisfaction (H27)

0.78 ****

Behavioural Intentions

0. 94 ****

Non- significant

346

The results exhibited in Table 8.2 shows that perceived value affects only overall satisfaction directly, and that indirectly it emerged as having the strongest effect on behavioural intentions via overall satisfaction, which acts as a mediating construct. Direct seller satisfaction emerged as having a stronger indirect effect on behavioural intentions compared to product satisfaction. In conclusion, these findings clearly illustrate that perceived value is the most influential determinant of overall satisfaction and in turn, it has a very strong significant impact on behavioural intentions.

(5)

What are the determinants of behavioural intentions and which of these determinants has the most influential impact on the intentional behaviour propensity?

Figure 8.4 visually depicts that, in tandem, both relational commitment and overall satisfaction are significant determinants of behavioural intentions, however overall satisfaction is the most influential predictor of behavioural intentions.
Figure 8.4 Determinant of Behavioural Intention
Relational Commitment (H26)
0. 27 ****

Behavioural Intentions
0.78 *****

Overall Satisfaction (H27)

Note: **** Significant at p< 0.001 (t> 3.29) Non-significant path

Customer satisfaction has become a well-studied construct in marketing, given its importance and established relationship with behavioural intentions, customer retention and firm profitability (e.g. Anderson and Sullivan 1993; Anderson et al. 1994; Strong 2000). Generally, customer satisfaction seems to be a relatively stable and powerful predictor of customers subsequent behaviour (e.g. Cronin et al. 2000; Garbarino and Johnson 1999; Kassim 2001). In the current investigation, overall customer satisfaction is demonstrated as having the strongest impact on behavioural intentions, which are comprised of indicators such as word-of-mouth communication, repurchase intention, 347

price tolerance, customers intention to complain and intention to switch. Therefore, these results confirm the previous research that suggests that overall satisfaction is the key predictor of customers behaviour intentions. On the other hand, relational commitment elicits a moderately significant effect on behavioural intention. In essence, this finding offers empirical evidence that relationship commitment motivates the buyer to engage in certain behaviours, as suggested by previous scholars, for example Bloemer and Schrder (2002), Hennig-Thurau and Klee (1997), Johnson et al. (2001) and Macintosh and Lockshin (1997). In conclusion, overall satisfaction has the most impact on the intentional behaviour propensity, compared to relational commitment.

(6)

What is the role of Purchase Decision Involvement (PDI) in the proposed customer satisfaction model?

PDI has been postulated to be related positively with perceived value and product satisfaction in the present studys conceptual model. The results illustrated in Table 7.9 suggest that PDI is not a significant determinant of product satisfaction (H17); however, the significant effect of PDI on perceived value (H16) is established. This result indicates that, no matter how involved the customers were with the product, they displayed similar judgements on their product evaluation. However, it is logical to propose that PDI could be put forth as a moderating variable in the customer satisfaction framework, but this aspect is not within the scope of the current investigation. In conclusion, the findings imply that PDI is not a significant predictor of customers satisfaction with the product but it has a significant effect on perceived value.

(7)

What is the role of Perceived Equity in the proposed customer satisfaction model?

Figure 8.5 depicts that in this study, perceived value is the only antecedent of perceived equity, and subsequently relational commitment is its consequence.

348

Figure 8.5

The Role of Perceived Equity


Perceived Value (H21)

Relational Commitment (H24)


0.56 *** 0.37 ****

Perceived Equity

Direct Seller Performance (H22)

Direct Seller Satisfaction (H23)

Note: **** Significant at p< 0.001 (t> 3.29) *** Significant at p< 0.01 (t> 2.57)

Non-significant path

The result of the structural model estimation reveals that direct seller performance was not a significant predictor of perceived equity, which contradicts Armstrong and Tans (2000) and Patterson et al. (1997) empirical findings. On the other hand, perceived value is demonstrated to be significantly related to perceived equity. As acknowledged in Chapter 4, even though this hypothesised path seems very promising, this study was the first to present empirical evidence in support of its significance within the buyerseller relationship, particularly in the direct sales environment.

In the prediction of the effect of perceived equity on direct seller satisfaction, perceived equity was found not to be significantly related to direct seller satisfaction. This finding clearly contradicts Oliver and Swan (1989a, 1989b) and Johnson et al.s (2001) findings that perceived equity (fairness) dominates respondents satisfaction judgements, but lends support to Armstrong and Tans (2000) empirical work. This unexpected result could perhaps be explained by the Satisfaction First Model (see Section 3.6.3 and Figure 3.9) which posits that equity bridges the gap between satisfaction and loyalty, and is assumed to occur as a result of the consumers cumulative satisfaction experience. It is reasonable to assume this contention might hold in this research sample because it was reported that the majority of the respondents (see Section 6.4) are experienced shoppers within the direct sales channel, and most certainly, the current purchase transaction was not their first purchase of the specific product under study. Perhaps, in this case, perceived equity is not a predictor of the customer satisfaction process, but acts in the opposite direction; that is, customer satisfaction is the predictor

349

of perceived equity as postulated by Olsen and Johnson (2003) in the Satisfaction First Model. This notion will be expanded in the suggestions for future research.

In addition, it is worthwhile to highlight that the current investigation has found strong empirical support for the hypothesised path of perceived equity and relational commitment. In conclusion, this study affirms Johnson et al.s (2001) study that illustrates the significant effect of perceived equity on relational commitment towards the direct seller.

(8)

What is the role of Perceived Value in the proposed customer satisfaction model?

It was argued that if customer satisfaction is viewed as an outcome variable, then focusing the investigation on antecedents and determinants of customer satisfaction is necessary to affect the desired outcome. Taking the position that satisfaction is largely based on a value analysis prior to and during purchase and consumption, the author has undertaken a major review and evaluation of both the academic and practitioneroriented literature related to the value construct. It is believed that understanding the value assessment process can lead to a better understanding of the process that begets CS/D. The current investigation presents the synthesis of the value judgement process that occurs with a consumption system approach. Figure 8.6 shows that direct seller performance and product performance are antecedents of perceived value, and perceived equity and overall satisfaction are its consequences.
Figure 8.6 The Role of Perceived Value
Overall Satisfaction (H20)
0.56 0.53****

Direct Seller Performance (H18) Perceived Value


0.94****

Product Performance (H19)

0.40**** 0.56 ***

Perceived Equity (H21)

Note: **** Significant at p< 0.001 (t> 3.29) *** Significant at p< 0.01 (t> 2.57)

350

The results demonstrate that both direct seller and performance are significant predictors of perceived value. This implies that perception of direct seller and product performance could enhance customers judgement of the perceived value they have received from the direct sales channel as compared to the conventional retail system. It is important to emphasise that the current studys conceptualisation of perceived value includes indicators that take into account product and direct seller aspects, which is considered appropriate in the direct sales context. Importantly, perceived value has been found to have a strong significant effect on overall satisfaction. This finding is consistent with several scholars who confidently suggested that perceived value played a critically important role in customers satisfaction evaluation (e.g. Athanassopoulos 2000; Kristensen et al. 1999; Patterson and Spreng 1997; Webb and Jagun 1997). In

conclusion, the current study provides evidence that perceived value is positively related to perceived equity and it is the most important determinant (as shown by its high regression coefficient value) of customer overall satisfaction channel. with the direct sales

(9)

What is the role of Relational Commitment in the proposed customer satisfaction framework?

Figure 8.7 illustrates that direct seller satisfaction and product perceived equity are determinants of relational commitment and in turn influenced customer behavioural intentions.

Figure 8.7

The Role of Relational Commitment


Direct Seller Satisfaction (H25)
0.56 0.40 ****

Perceived (H24)

Equity

0.37 ****

Relational Commitment (H 26)

0.27****

Behavioural Intentions

Note **** Significant at p< 0.001 (t> 3.29)

351

This study found strong empirical support for the effects of direct seller satisfaction and perceived equity on relational commitment. These findings imply that salespersons play a crucial role in the formation of long-term buyer-seller relationships (Crosby and Stephens 1987; Gabarino and Johnson 1999; Johnson et al. 2001; Priluck 2003). As expected, relational commitment significantly affects behavioural intentions, which lends support to studies by Bloemer and Schrder (2002), De Wulf and Schrder (2003) Johnson et al. (2001) and Macintosh and Lockshin (1997) that provide empirical evidence that committed customers have a greater propensity to develop and maintain relationships with their sellers and, consequently, influence their behaviours. In conclusion, this study indicates that relational commitment is a significant predictor of customers behavioural intentions.

In summary, the major findings derived from the research questions are as follows: Product effectiveness is the most important driver of the Product Quality factor of the product subsystem, whereas product information adequacy is the most important driver of the Product Offering and Information factor of the product subsystem. Maintaining a professional appearance and have customer interest at heart are the key drivers of customers satisfaction toward the direct seller. Corporate information/publicity is the most important driver of the Corporate Image factor of the direct selling company subsystem, whilst handle complains promptly is the most important driver of Corporate Customer Service factor of the company subsystem. Performance is the most influential determinant of customer satisfaction at the subsystem level, whereas perceived value is the most important determinant of customers overall satisfaction with the direct sales channel. In addition, performance is significantly associated with disconfirmation, however disconfirmation-satisfaction link is not evidenced by this study. Customers satisfaction with the direct seller is the most predictive determinant of their overall satisfaction with the direct sales channel as compared to product satisfaction. Direct selling company satisfaction is not a significant predictor of customers overall satisfaction. Purchase decision involvement is not a significant predictor of customers satisfaction with the product but its impact on perceived value is established. Overall satisfaction has the most influential impact on behavioural intentions and relational commitment is also revealed to be a significant predictor of behavioural intentions, but of a smaller magnitude compared to overall satisfaction. 352

Interestingly, the study reveals two new linksnamely perceived valuerelational commitment and perceived valueperceived equity.

8.4 8.4.1.

Structural Equation Modelling (SEM) Measurement Model Conclusions

Twenty measurement scales were identified and applied for the current customer satisfaction investigation. Of these, the core constructs are product performance /disconfirmation, direct seller performance/disconfirmation and direct selling companies performance/disconfirmation, which were developed specifically for this study by strictly following Churchills (1979) Procedure for Developing Measure (see Section 5.6). The measures were purified via Item Analysis and EFA (see Section 5.6.4) and subsequently refined and verified for unidimensionality, validity and reliability by CFA, as recommended by Gerbing and Anderson (1988) (see Part 1 of Chapter 7).

Validity is concerned with how well the concept is defined by the measures (Hair et al. 1998). On the other hand, reliability is defined as the degree to which the observed variable measures the true value and is error free (Hair et al. 1998). In brief, validity emphasises what should be measured, whilst reliability focuses on how it is measured (Hair et al. 1998). Reflecting back upon the methods utilised to validate the measures utilised in the current investigation, they could be described as having been scrutinised rigorously via CFA. Due to the comprehensiveness of the hypothesised model, it was then decomposed into four measurement models in order to achieve a case / parameter ratio of at least 5:1 (Baumgartner and Homburg 1996; Bentler and Chou 1987; Jap and Ganesan 2000). Subsequently, all the sub-models were assessed for convergent validity, unidimensioanlity, reliability and discriminant validity (see Section 7.4). The summarised results of each of the measurement models were presented explicitly in Tables 7.2 (b), 7.3 (b), 7.4 (b) and 7.5 (b). In light of the results derived from the assessment of the measurement model, it could be succinctly concluded that all the measures successfully achieved adequate measurement quality in terms of unidimensionality, reliability, and convergent and discriminant validity.

The main criticism of this study is perhaps from the perspective of external validity, as the respondents were not selected randomly, and hence may not be an accurate representation 353

of the population of direct sales customers. Generally, in most studies, the external validity (i.e. generalisability), is considered quite difficult to assess (Lynch 1999). However, it is worth noting that several of the descriptive analysis findings with regard to the respondents consumption pattern and demographic profiles, exhibited in Chapter 6, demonstrate that these variables are relatively consistent with the existing published literature pertinent to this aspect (e.g. Barnowe and McNabb 1992; Endut 1999; Peterson et al. 1989; Raymond and Tanner 1994). Hence, this clearly offers some indications that the results and interpretations may be generalisable, specifically with regard to the direct sales consumption context in Malaysia.

Furthermore, this thought is in line with Lynch (1999, p. 368), who professes that external validity can never be increased or established because it is a function of law of behaviour that is whether the manipulated variables combine additively or interactively with a host of background factors. He further asserts that external validity could be achieved by having pertinent knowledge of how the focal constructs in one theory interact with moderator variables. In this regard, he recognised that most implementable sampling plans in consumer research engaged in selecting samples that might affect the dependent variables or they might act as proxies for the true causal factors that influence the dependent construct. Therefore, Lynch (1999) points out that in this condition, the statistical sampling theory could not make any rigorous claim pertaining to the generalisability of ones results. In fact, Ferber (1977) suggest s that there is no place for probability samples in basic or applied consumer research, quota samples are most appropriate because the researcher has better insights of the background variables, which are suitable for quota characteristics. Accordingly, Lynch (1999) argues that this specific knowledge acquired by the researcher is essential in assessing the generalisability of ones research findings.

8.4.2

Structural Model Conclusions

SEM was employed to exemplify the simultaneous effect of all the constructs incorporated in the conceptual model and demonstrate how they are interrelated to explain and predict the focal construct (i.e. overall satisfaction) and subsequently affect behavioural intentions (see Figure 4.2, p. 134). The hypothesised relationships among the

354

constructs, as posited in the conceptual model in Figure 4.2, were assessed using the AMOS 4 programme.

In addition, these results were also utilised to address the research questions formulated for this study (see Table 1.2, p.20). The overall fit of the hypothesised model seemed to marginally fit the observed data, hence the model was re-specified as recommended by Byrne (2001) in order to establish a more parsimonious and best-fit model. The initial evaluation of the proposed model showed twenty out of twenty-seven hypothesised linkages were statistically significant. Two alternative models were formulated (alternative # 1 and # 2); model # 1 posits an additional path to the proposed model (perceived value - behavioural intentions) which was not supported by the studys finding. This implies that perceived value does not have a significant direct effect on customers future behavioural consequences. On the other hand, model #2, which postulates an additional path to the proposed model (perceived value relational commitment), provides evidence that this new link was highly significant and it exhibited a better overall fit than model # 1. Finally, model #2 was re-estimated after deleting the non-significant paths (Bagozzi 1992).

8.5

Concluding Remarks

The discussion of the research questions unequivocally illuminates several key findings of this study to the marketing theory and direct sales industry. The results demonstrate that performance is the most important determinant of customer satisfaction at the subsystem level, whereas perceived value is the most influential predictor of global (overall) satisfaction with the direct sales channel. Interestingly, it was revealed that customer satisfaction with the direct seller is more influential than the product in contributing to customers overall satisfaction with the direct sales channel. The results of this study also imply that customers overall satisfaction is the strongest, but not the only driver of their future behavioural intentions. Relational commitment also exhibits a significant positive influence on behavioural intentions. It was found that perceived value does not have direct significant impact on behavioural intentions; hence, the impact of perceived value on customers behavioural outcomes is completely mediated through overall satisfaction (Patterson and Spreng 1997). Most importantly, in the current investigation, two new paths (perceived value relational commitment and perceived 355

value perceived equity) have been identified and sufficient evidence of their plausibility is provided.

The following chapter (Chapter 9) will illuminate the pertinent results reported in this chapter with a view to present the studys contributions to the existing body of knowledge. The implications of the key empirical findings on the theoretical and practitioners perspective will be discussed and the studys limitations will be presented. Finally, potentially fruitful future research opportunities and directions will be suggested.

356

357

Vous aimerez peut-être aussi