A path analysis study of the relationships among consumer satisfaction, loyalty, and market share in retail services.

Pleshko, Larry P. ; Baqer, Samar M.

INTRODUCTION

An accurate description of the basic consumer model, along with the relative influence of that model on firm performance, has interested marketing theorists for decades. The basic premises in marketing are straightforward: (a) better value for the buyer should lead to consumption and satisfaction, (b) a satisfied buyer will eventually and for various reasons become a repeat purchaser and/or loyal buyer, (c) this buyer loyalty and satisfaction should result in improved marketing performance for a variety of reasons, and (d) the improved marketing performance should lead to better overall firm performance (Leverin & Liljander 2006, Story & Hess 2006, Cooil et al 2007).

However, these general premises do not hold for every industry or under every condition. There are multitudes of examples that better mousetraps do not always result in higher sales or share, usually as the result of poor marketing, poor relative value, or a variety of macro-firm factors. Plus, satisfaction does not always lead to improved firm performance even when it does lead to improved market shares (Pleshko & Cronin 1997). Satisfaction is also not always enough to ensure customer loyalty, even though satisfaction leads to loyalty in many instances (Mitchell & Kiral 1998, Reichheld & Sasser 1990). Buyer loyalty generally does lead to improved marketing performance, but not in all investigations (Ehrenberg & Goodhardt 2002, Knox & Denison 2000). Therefore, the specific conditions for which the marketing premises hold are still under investigation.

The purpose of this study is to partially investigate the premises mentioned above. In particular, the interrelationships among consumer satisfaction, consumer loyalty, and market share are studied in four types of retailers: fast-food burgers, convenience stores, health clubs, and medical clinics. This study will initially address the literature related to the primary constructs under study. This will be followed by a discussion of the data collection and measurement, analyses, results and discussion, as well as any limitations of the study.

CUSTOMER SATISFACTION

Customer satisfaction and retention are generally considered among the most important long term objectives of firms (Cooil et al 2007). The marketing concept suggests that a satisfied buyer will likely return to purchase again, or at least, consider purchasing again (Keith 1960, Leavitt 1960). According to Reichheld and Sasser (1990) repeat customers cost less to serve than new buyers, benefiting a firm's cost structure. Additionally, maximizing customer retention rates and minimizing customer defections are primary strategic objectives for most firms, as evidenced by companies' emphasis on customer relationship management (Ching et al 2004, Verhoef 2003). Thus, previously satisfied buyers may help firms both reduce marketing costs and develop more stable levels of sales when a large number of satisfied buyers are retained to purchase again in the future.

There are several definitions of customer satisfaction in the marketing literature. It is generally accepted that satisfaction is a psychological state that a consumer experiences after consumption (Oliver 1980). Additionally, the basic conceptualizations focus on either or both of two aspects: (i) of the buyers' initial expectations in relation to the product and (ii) the buyers' perceptions of the product performance in relation to these expectations (Turan 2002, Churchill & Suprenant 1982, Oliver 1980). The general idea is that a potential buyer has specific expectations about a product before purchasing: such as, health club 'A' will provide superior customer service and maybe excellent facilities. If, after using health club "A", the buyer feels that these salient criteria are met, then he/she will be satisfied, and vice versa. The measurement of satisfaction generally focuses on the product performance aspect rather than expectations. As such, it may be suggested that a buyer's 'fulfillment response', as derived from the performance of features or benefits of the product, is the vital aspect in defining satisfaction (Turan 2002).

BUYER LOYALTY

The investigation of brand loyalty has had a long and rich tradition in the field of marketing. From early research until today, a variety of conceptualizations of the buyer loyalty concept have been developed. Purchase possibility, purchase frequency, awareness, and trust or commitment over the long-term have all been proposed as a means to measure or view brand loyalty (Story & Hess 2006, Oliver 1999, Twedt 1967, Brody & Cunningham 1968, Farley 1964). The modern conceptualization is from Dick and Basu (1994) who argue that brand loyalty should not be regarded as mere repurchase behavior, but rather as a combination of purchase behavior and attitudes. Accordingly, true brand loyalty requires repeat purchase behavior in addition to a significant psychological attachment to the brand which is consumed. Based on this two-dimensional approach, true brand loyalty is now widely defined as an individual having both a favorable attitude towards and consistent purchase pattern towards a brand over time and research has shown this to be the case (Kerin et al 2006, Pleshko & Heiens 1997; 1996).

It is accepted that maintaining and increasing loyalty is a primary responsibility for any marketing manager. Customer retention programs may lead to positive increases in buyer loyalty, but with no guarantees (Story & Hess 2006). Plus, today's loyal buyers might not be so loyal in the future since loyalty is transient (Oliver 1980). Regardless, across industries, it is a firm's ability to manage both the penetration levels and the repeat purchases of its buying market which dictates the extent of customer base retention now and in the future (McDowell & Dick 2001, Lehmann & Winner 1997).

RELATIONSHIP BETWEEN CUSTOMER SATISFACTION AND BUYER LOYALTY

A good indicator of buyers' commitment towards specific retail brands should be the loyalty of customers (Cooil et al 2007, Rosenberg & Czepial 1983). Since the formalization of the marketing concept, the idea that satisfied buyers will (hopefully) return for future purchases and eventually become loyal buyers has been the basis for marketing thought. Loyalty and satisfaction can make customers more forgiving of disappointing situations associated with a certain brand name or with a store (Shankar et al 2003). In fact, this relationship between satisfaction and loyalty has been shown to be the case throughout much of the literature. Many studies have focused on the positive relationship between customer satisfaction and loyalty, with researchers considering satisfaction as one of the primary antecedent of loyalty, especially in retail businesses (Dong 2003, Bloemer & Ruyter 1998, Oliver 1997, Bitner 1990).

However, despite the expectations of both common sense and empirical research, there is a growing school of thought which suggests that satisfaction is not a reliable predictor of buyer loyalty (Story & Hess 2006, Reichheld 2003). In this perspective, it is possible to be a satisfied buyer but not a loyal buyer. The primary point of emphasis is this: loyal customers are always satisfied but satisfied customers are not always loyal. If we consider that satisfaction is an attitude, then this follows the attitude-behavior definition of loyalty presented previously discussed (Dick and Basu 1994). Thus, the important questions relate to: (i) the direction of the relationship, and (ii) the conditions under which this relationship exists. The current study accepts general marketing thought, in that a positive relationship should be evident. Hence, the following hypothesis:

H1: There is a positive relationship between customer satisfaction and buyer loyalty in the Retail Business.

MARKET SHARE AND ITS RELATIONSHIPS WITH CUSTOMER SATISFACTION AND BUYER LOYALTY

It should be logical that both satisfaction and loyalty are positively related to market share. Increases in satisfaction hopefully lead to repeat purchases and positive word-of-mouth between buyers. Increases in loyalty should also lead to positive word-of-mouth, increasing purchase volume, and also to lower marketing retention costs (Zeithaml 2000, Tellis 1988, Rosenberg & Czepial 1983). Those firms which continually and properly reinvest these potential higher profits should develop an advantage leading to higher market shares (Day & Wensley 1988). The advantages garnered from loyalty are especially relevant in mature markets as increases in share become more expensive. Improvements in the loyalty base are one viable means of increasing and maintaining share (Ghounaris & Stathakopoulos 2004).

Previous empirical research and conceptualizations link loyalty to either market share or the profitability of the firm (Leverin & Liljander 2006, Ehrenberg & Goodhardt 2002, Fader & Schmittlein 1993, Colombo & Morrison 1989, Raj 1985, Robinson 1979). These studies generally have found a positive relationship whereby brands with large market shares usually have the most brand loyal buyers (and vice versa). One explanation is that, over time as buyers switch between brands, a percentage of those who switch remain buyers of the new brand. And since the larger firms gain more of the switchers then these larger firms gain share over time (and vice versa). Another explanation is that buyers might use brand popularity (high market share) as a surrogate for quality, resulting in higher purchases from new buyers (Caminal & Vives 1996). Hence, the following hypothesis:

H2: Buyer loyalty is positively related to market share in the Retail Business.

All retail and/or service businesses rely heavily on their customers being happy and satisfied in order to provide repeat purchases and continued patronage. Firms which do not continually satisfy their buyers run the risk of customer-base reductions, resulting in smaller sales and market shares over time (Tellervision 2006). The link between customer satisfaction and performance can be established through many factors, including cost reduction derived from repeat purchases or positive word of mouth leading to increased penetration (Cooil 2007, Zeithaml 2000, Reichheld 1990). The following hypothesis summarizes the expected satisfaction-share relationship. See Figure 1 for the general model to be tested.

H3: Customer satisfaction is positively related to market share in the Retail Business. .

[FIGURE 1 OMITTED]

DATA COLLECTION

The data for the current study is gathered from a buyer group in a large university town in the southeastern USA. The sampling frame is comprised of undergraduate business students, a group of consumers who are frequent users of each of the four types of retail businesses under study: health clubs, convenience stores, medical clinics, and fast food burger outlets. Information is accepted only from consumers who buy from the specific type of retail outlet: non-users are excluded. The data are from self-administered questionnaires. Twelve classes are randomly selected for inclusion in the study from the business school offering at the university. Each class is assigned to one specific retailer type: three classes to each retailer type. This process results in the following number of usable respondents: health clubs: 79, convenience stores: 89, fast-food hamburger outlets: 96, and medical clinics: 69.

Many stores (retailer bands such as Wendy's) in each category are included in the study. The retailers are identified by speaking with the buyers and looking through the yellow pages to locate outlets within the city limits. An 'others' category is included to catch those retailers not specifically listed on the questionnaire. The retailer types themselves are selected for two reasons. First, it was necessary to find a type of business that was used by the target group under study. Second, it was important to select a variety of retailers in order to test the model, as we would expect to find differences based on retailer-types (Chaudhuri & Hollbrook 2001, Murphy and Enis 1986).

For health clubs there are sixteen original clubs included on the questionnaire. Eleven of these clubs are eliminated after data collection due to small numbers of users or small market shares. It seems that the student market is focused on only a few health clubs in the vicinity of the university. This resulted in five health clubs for inclusion in the analyses: club#2, club#9, club#3, club#10, and club#13, all with market shares above five percent.

For convenience stores there are twelve original outlets in the vicinity of the market that are included on the questionnaire. Only four of these stores are eliminated after data collection due to small numbers of users or small market shares. This resulted in eight convenience stores for inclusion in the analyses: store#8, store#10, store#4, store#6, store#3, store#5, store#2, and store#9, all with market shares above five percent.

For medical clinics there are twelve clinics in the general area that are included on the questionnaire. Six of these clinics are eliminated after data collection due to either a small number of users or a small market share. This resulted in six medical clinics that are included in the analysis. In order of market share from lowest to highest, the clinics are: clinic#1, clinic#4, clinic#6, clinic#8, clinic#2, and clinic#7. All six of the clinics have market shares above three percent.

Six fast food hamburger outlets are included in the study. Each of the six brands has a major presence in the town with multiple locations. Most of the outlets are part of national franchises, with only two being regional. All six outlets are kept for the analyses and all six have market shares above five percent.

MEASUREMENT

The study includes one indicator of market share, two indicators of buyer loyalty, and one indicator of consumer satisfaction. The measures are identical for each of the retail types: summated/aggregate percentages for both loyalty and market share and sample averages for satisfaction. The general statistics of the indicators are shown in Table1, for each of the relevant constructs in each retail type. The variables are described below.

Market-Share (MSHA) is defined as the visits (uses) for a retail store divided by the total visits (uses) for all stores in that category. It is calculated for each brand in each retail type. The respondents are asked how many 'times' they visit each retailer per month. These 'times' are summed for each store and overall within each store type. Thus, MSHA(A) = (times for store A)/ (summation of times for stores A, B, C... N).

Loyalty-%-of-total-use (L%TOT) is defined as the percentage of total times (uses) the respondent uses each store if they are users of that store. It is calculated for each respondent for each store used. Then an aggregate sample percentage is calculated. Thus, for respondent X who uses stores A and B: L%TOT(X,A) = timesA/(timesA + timesB + ... + timesN) and L%TOT(X,B) = timesB/(timesA + timesB + ... + timesN), L%TOT(X,C) =0, and etc.

Loyalty-Most-used-% (L%MST) is defined as the percentage that each store is used as primary store in the category. Respondents indicate the number of 'times' they use each store. The store is the most used for a respondent when the largest number of visits is indicated. Thus, for respondent X, if timesA > timesB, timesC ..., timesN, then storeA is assigned to respondent X as most used store. The indicator is the summation for each store of those respondents most using each store divided by the total respondents in that retailer-type.

The study also includes one indicator of consumer satisfaction (SATF), which is comprised of four measurement items. Each of the four questions is measured using consumer ratings on a scale from very satisfied [7] to very dissatisfied [1]. The four satisfaction items are factor analyzed using principal axis analysis for each type of retailer. In each of the four retail consumer groups the four items exhibited a single dimension. The overall indicator of SATF is constructed by summing the four items into an overall score. Across the sample SATF has a possible range from four to twenty-eight. For health clubs, SATF has a mean of 20.45, a standard deviation of 4.6, and a coefficient alpha of 0.917. For medical clinics, SATF has a mean of 21.43, a standard deviation of 5.4, and a coefficient alpha of 0.949. For convenience stores, SATF has a mean of 19.18, a standard deviation of 4.5, and a coefficient alpha of 0.834. For burger outlets, SATF has a mean of 20.65, a standard deviation of 4.1, and a coefficient alpha of 0.887.

ANALYSIS/RESULTS

The number of observations (the stores in each type of retailer) is minimally sufficient to allow the necessary observations per variable, usually a minimum of eight to ten per variable (c.f. Stevens 1986). The analysis proceeds in two stages: (i) comparative regressions and (ii) path analysis tests. This will allow the computation of a [X.sub.2] statistic to determine if each of the relevant paths shown in Figure 1 should be included or removed from the model.

First, regressions are run for each of the dependent variables with all predictors included. This is referred to as the fully-recursive analysis. This is Ha for our purposes for each path. Then the variable/path of interest is removed and a regression is again run. This is Ho for our purposes. The explained variance from the two regressions is compared to arrive at the explained variance ([R.sub.2]) for the path of interest. This procedure is repeated for each path in the model plus two paths involving the covariate, retailer-type. For each path, the regression equations are shown in Equations one through six below. The findings of this comparative regression are summarized in Table 2 below.

 For path H3: satisfaction-market share

Equation 1a Ho: MSHA% = TYPE + L%TOT + L%MST + error
Equation 1b Ha: MSHA% = TYPE + SATF + L%TOT + L%MST + error

 For path H1: satisfaction-loyalty

Equation 2a Ho: SATF = TYPE + error
Equation 2b Ha: SATF = TYPE + L%MST + error
Equation 2c Ho: SATF = TYPE + error
Equation 2d Ha: SATF = TYPE + L%TOT + error

 For path H2: loyalty-market share

Equation 3a Ho: MSHA% = TYPE + SATF + error
Equation 3b Ha: MSHA% = TYPE + SATF + L%TOT + L%MST + error

 For path: type-market share

Equation 4a Ho: MSHA% = SATF + L%TOT + L%MST + error
Equation 4b Ha: MSHA% = TYPE + SATF + L%TOT + L%MST + error

 For path: type-loyalty

Equation 5a Ho: TYPE = SATF + error
Equation 5b Ha: TYPE = SATF + L%TOT + L%MST + error.
Equation 5c Ho: TYPE = SATF + error
Equation 5e Ha: TYPE = SATF + L%TOT + L%MST + error.

 For path: type-satisfaction

Equation 6 Ho: SATF = TYPE + error

Next, the current study uses the path analysis methodology to test the overall relationships among the identified research constructs. Path analysis is a relevant technique for testing causal models (e.g. Pedhazur 1982). The technique allows the researcher to determine if the hypothesized model is represented in the data; in other words, whether the causal model is consistent with the variable intercorrelations (Pedhazur 1982). As previously explained, the hypothesized model to be tested is identified in Figure 1

To test a model using path analysis, our hypothesized model is compared with a fully-recursive model using Equation seven identified below (Pedhazur 1982, p. 619). The calculated test statistic, 'W', is distributed as a chi-square, with 'd' degrees of freedom. If the null hypothesis is rejected, then support is provided for the hypothesized model. This means that if a null hypothesis is rejected, then the tested path cannot be excluded from the model. On the other hand, if Ho is not rejected, then no evidence is found to include the tested path in the model. The results of the path analyses tests are also shown in Table 2 and the actual calculations are shown in the Appendix.

Equation 7: W = -(n-d)[ln(1-[R.sub.2]a)/(1-[R.sub.2]o)]

Where

W = [X.sub.2] statistic

d = d.f. = model paths hypothesized to be zero

ln = natural log

[R.sub.2]a = 1-[(1-[r.sub.2]i)(1-[r.sub.2]ii)(1-[r.sub.2]iii)(etc.)] for fully recursive model

[R.sub.2]o = 1-[(1-[r.sub.2]i)(1-[r.sub.2]ii)(etc.)] for model with path excluded

First to test proposition H1, that consumer satisfaction is related to buyer loyalty. It is noted in Table 2 that these constructs exhibit a significant common variance with an [R.sub.2]=.149 (p<.050). Thus, the satisfaction-loyalty path cannot be excluded from the model. Also, Table 2 reveals that the relationship is negative in direction: as satisfaction increases, then loyalty decreases.

Next to test the proposition H2, that buyer loyalty is related to market share. It is noted in Table 2 that these constructs exhibit a significant common variance with an [R.sub.2]=.636 (p<.001). Thus, the loyalty-market share path cannot be excluded from the model. Also, Table 2 reveals that the relationship is positive in direction: as loyalty increases, then market share increases.

Next to test the proposition H3, that consumer satisfaction is related to market share. It is noted in Table 2 that these constructs do not exhibit a significant common variance with an [R.sub.2]=.004 (p<.800). Thus, the satisfaction-market share path should be excluded from the model.

Next to determine if the covariate, industry type, has an influence on market share. It is noted in Table 2 that these constructs do not exhibit a significant common variance with an [R.sub.2]=.008 (p<.700). Thus, the retailer type-market share path should be excluded from the model. It appears that market share does not differ across these service categories.

Next to determine if the covariate, industry type, has an influence on buyer loyalty. The regression for this path has two dependent variables. Thus, the explained variance is the total of two comparative regressions divided by two. It is noted in Table 2 that these constructs exhibit a significant common variance with an [R.sub.2]=.463 (p<.001). Thus, the satisfaction-loyalty path cannot be excluded from the model. It appears that buyer loyalty differs across categories.

Next to determine if the covariate, industry type, has an influence on consumer satisfaction. This path is tested using simple Anova, as there are no paths to eliminate for path analysis testing. It is noted in Tables 2 that these constructs exhibit a significant common variance with an [R.sub.2]=.249 (p<030.). Thus, the retailer type-satisfaction path cannot be excluded as a covariate in the model. It appears that consumer satisfaction differs across categories.

DISCUSSION/IMPLICATIONS

The purpose of the study was to determine if the general model presented in Figure 1 is valid in the four retailer-types under investigation in this preliminary study. The results indicate that this generally accepted theoretical model is not as broad-ranging as might be expected, as one of the three proposed relationships is not supported. Additionally, the covariate retailer-type (category) is shown to have a significant impact on two of the three constructs in the model.

The findings do confirm the proposed relationship between loyalty and market share. This is congruent with many other studies across industries (Reinartz & Kumar 2002). Thus, as expected, increased buyer loyalty leads to increases in market share.

The authors also find a significant relationship between satisfaction and loyalty. But, the direction is not as expected, with increases in satisfaction associated with decreases in loyalty. It could be that, with these types of services, repetitive usage of the same brand (loyalty) may lead to lower levels of satisfaction through boredom. Or possibly, the initial euphoria with a brand (satisfaction) only decreases through time and repetitive usage or loyalty.

The absence of a relationship between satisfaction and market share is puzzling but not unexpected. Situational or socio-cultural influences might lead a satisfied buyer to purchase other brands on a regular basis. In university settings, for example, emphasis is placed on reference groups as well as social status. It may be possible that a buyer is pressured by friends to use a specific retailer because everyone else in the group shows patronage there. Or, it may be that one retailer has a much better reputation leading the buyer to switch even though there is satisfaction with a lower status retailer. Or, the prevalence of multiple retail services offering satisfying purchase experiences may overcome this expected relationship.

The lack of a direct effect of satisfaction with market share does not mean that satisfaction has no influence on market share. It could be that loyalty and satisfaction interact, thus indicating satisfaction might be a moderator. We cannot test this idea in the current study. An alternative may be that loyalty acts as a mediator between satisfaction and market share. To recall, the main effect of loyalty on market share is 63.6%, with 9.48% (.636 *.149) of that effect being from loyalty acting as a mediator between satisfaction and market share. However, this 9.48% is not a significant amount in the current study. Path analysis seven in Table 2 and the appendix reveals that we should eliminate the loyalty-as-mediator path (p<.15). Therefore, satisfaction simply does not have either a significant main effect or a significant indirect effect on market share in this study.

The covariate, retailer-type or category, is shown to be related to both consumer satisfaction and buyer loyalty, but not market share. The market shares would most likely vary across categories if the categories exhibited differing levels of concentration. This is likely not the case in these services retailers seemingly in long-term stable markets. However, it is not surprising that both satisfaction and loyalty differ across categories. As each type of retailer seems to be distinct from the others regarding many factors related to both loyalty and satisfaction (such as level of service, skills required to compete, tangibility, etc), it is not surprising that the categories differ on these two important constructs.

LIMITATIONS AND FUTURE CONSIDERATIONS

The readers must wonder if the current findings are indicative of general tendencies or simply a characteristic of this limited student-based study of four retailer-types in a single university town. Larger studies with more respondents taken over time are probably needed to truly identify the scope of the outlined model in services retailing. Future research might include both different target respondents as well as different retailer-types. Additionally, future efforts might address these issues in cross-national studies to determine where and under what conditions the satisfaction-loyalty-share model applies globally. Finally, the question of life cycle issues might also be relevant for this area of study: is the model supported at various stages of the industry life cycle.

An unanswered question is whether there is an interaction between satisfaction and loyalty in their effects on market share. The number of observations in this study was too small to include interactions in the regression analyses and, therefore, moderating effects could not be determined. A problem in this study is the relatively small number of major competitors in a category. This limitation would have to be addresses in order to obtain a large enough aggregate sample size of stores or brands in order to test this possibility.

Also, the question of measures may also be relevant. The use of other performance measures, such as profitability, as well as alternative indicators of each construct may add variety to the findings. Specifically, one of these measures is the share of wallet which represents the total spending of customers in any particular category (Cooil et al 2007). Share of wallet is effectively the share of a buyer's spendable income and may be a more logical proxy for share than those used in this study. Also, different types of loyalty might be included (true, spurious, latent, cognitive, affective, conative, and action) in order to capture a better explanation of the relationship among loyalty, customer satisfaction and market share (Oliver 1999, Dick & Basu 1994).

APPENDIX

W =-(n-d)[ln(1-[R.sub.2]a)/(1-[R.sub.2]o)]

Where

W = [X.sub.2] statistic

d = d.f. = model paths hypothesized to be zero

ln = natural log

[R.sub.2]a = 1-[(1-[r.sub.2]i)(1-[r.sub.2]ii)(1-[r.sub.2]iii)] for all paths included in full model: i, ii, iii

[R.sub.2]o = 1-[(1-[r.sub.2]i)(1-[r.sub.2]ii)] for i and ii included in model less path(s)

PATH ANALYSIS #1: loyalty--market share

[R.sub.2]Loy-MSh: .636

N=25: number of observations

d.f.=d=1 : hypothesized paths which equal zero

[R.sub.2]o = 1-[(1-.004)(1-.008)] = 1-[(.996)(.992)] = 1-.988032 = .011968

[R.sub.2]a = 1-[(1-.636)(1-.004)(1-.008)] = 1-[(.364)(.996)(.992)] = 1-.359644 = .640356

W =-(25-1)[ln (1-.011968) / (1-.640356)] = -24(-1.0106) = 24.25443

'p' = <.001

Conclusion: must reject proposed model--thus, loyalty is a significant predictor of market share and the path should be included in the model

PATH ANALYSIS #2: satisfaction--loyalty

[R.sub.2]Sat-Loy: .298/2 = .149

N=25: number of observations

d.f.=d=1 : hypothesized paths which equal zero

[R.sub.2]o = 1-[(1-.463)] = 1-.537 = .463000

[R.sub.2]a = 1-[(1-.149)(1-.463)] = 1-[(.851)(.537)] = 1-.456987 = .543013

W =-(25-1)[ln (1-.543013) / (1-.463000)] = -24(-.16134) = 3.872236

'p' = <.05

Conclusion: must reject proposed model--thus, satisfaction is a significant predictor of loyalty and the path should be included in the model

PATH ANALYSIS #3: satisfaction--market share

[R.sub.2]Sat-MSh: .004

N=25: number of observations

d.f.=d=1 : hypothesized paths which equal zero

[R.sub.2]o = 1-[(1-.636)(1-.008)] = 1-[(.364)(.992)] = 1-.361088 = .638912

[R.sub.2]a = 1-[(1-.636)(1-.004)(1-.008)] = 1-[(.364)(.996)(.992)] = 1-.359644 = .640356

W =-(25-1)[ln (1-.640356) / (1-.638912)] = -24(-.00401) = .096169

'p' = <.800

Conclusion: cannot reject proposed model - thus, satisfaction is not a significant predictor of market share and the path should be excluded from the model

PATH ANALYSIS #4: type--market share

[R.sub.2]Type-MSh: .008

N=25: number of observations

d.f.=d=1 : hypothesized paths which equal zero

[R.sub.2]o = 1-[(1-.636)(1-.004)] = 1-[(.364)(.996)] = 1-.362544 = .637456

[R.sub.2]a = 1-[(1-.636)(1-.004)(1-.008)] = 1-[(.364)(.996)(.992)] = 1-.359644 = .640356

W =-(25-1)[ln (1-.640356) / (1-.637456)] = -24(-.00803) = .192772

'p' = <.70

Conclusion: cannot reject proposed model--thus, retail store type is not a significant predictor of market share and the path should be excluded from the model

PATH ANALYSIS #5: type--loyalty

[R.sub.2]Type-Loy: .926/2 = .463

N=25: number of observations

d.f.=d=1 : hypothesized paths which equal zero

[R.sub.2]o = 1-[(1-.149)] = 1-[(.851)] = .149000

[R.sub.2]a = 1-[(1-.149)(1-.463)] = 1-[(.851)(.537)] = 1-.456987 = .543013

W =-(25-1)[ln (1-.543013) / (1-.149000)] = -24(-.62176) = 14.92217

'p' = <.001

Conclusion: must reject proposed model--thus, retail store type is a significant predictor of loyalty and the path should be included in the model

PATH ANALYSIS #6: type--satisfaction

[R.sub.2]Type-Sat: .249

N=25: number of observations

d.f.=d=1 : hypothesized paths which equal zero

'p' = <.03

Note: this path [R.sub.2] is found through an Anova, not comparative regressions Conclusion: must reject proposed model--thus, retail store type is a significant predictor of satisfaction and the path should be included in the model

PATH ANALYSIS #7: loyalty as mediator between satisfaction and share

[R.sub.2]Loy-MSh:. .636-(.636 *.149) = .541

[R.sub.2]Sat-MSh through Loy: .095

[R.sub.2]Sat-MSh: .004

N=25: number of observations

d.f.=d=1 : hypothesized paths which equal zero

[R.sub.2]o = 1-[(1-.541)(1-.004)(1-.008)] = 1-[(.459)(.996)(.992)] = 1-.453507 = .546493

[R.sub.2]a = 1-[(1-.541)(1-.095)(1-.004)(1-.008)] = 1-[(.459)(.905)(.996)(.992)] = 1-.410424 = .589576

W =-(25-1)[ln (1-.589576) / (1-.546493)] = -24(-.09982) = 2.395668

'p' = <.15

Conclusion: cannot reject proposed model--thus, loyalty does not act as a significant mediator between satisfaction and market share.

REFERENCES

Bitner, M.J. (1990) Evaluating Service Encounters: The Effects of Physical Surroundings and Employee Responses, Journal of Marketing, 54 (April), 69-82.

Bloemer, J. & de Ruyter, K. (1998) On the relationship between Store Image, Store Satisfaction, and Store Loyalty, European Journal of Marketing, 32, 499-513.

Brody, R.P. & Cunningham, S.M. (1968) Personality variables and the consumer decision process, Journal of Marketing Research, 10, 50-57.

Caminal, R. & Vives, X. (1996) Why market shares matter: An information-based theory, Rand Journal of Economics, 27 (summer), 221-39.

Chaudhuri, A. & Hollbrook, M.B. (2001) The chain of effects from brand trust and brand affect to brand performance: The role of brand loyalty, Journal of Marketing, 65 (2), 81-93

Ching, W., Ng, M.K. & Wong, K. (2004) Hidden Markov models and their applications to customer relationship management, Journal of Management Mathematics, 15 (1), 13-24,

Churchill, G. A., Jr. & Surprenant, C. (1982) An investigation into the determinants of customer satisfaction, Journal of Marketing Research, 19 (Nov), 491-504.

Colombo, R.A. & Morrison, D.G. (1989) A brand switching model with implications for marketing strategies, Marketing Science, 8 (1), 89-99.

Cooil, B., Keiningam, T., Aksoy, L, & Hsu, M. (2007) A longitudinal analysis of customer satisfaction and share of wallet: Investigating the moderating effect of customer characteristics, Journal of Marketing, 71, 67-83.

Day, G.S. & Wensley, R. (1988) Assessing advantage: A framework for diagnosing competitive superiority, Journal of Marketing, 52 (2), 1-20.

Dick, A.S. & Basu, K. (1994) Consumer loyalty: Toward an integrated framework, Journal of the Academy of Marketing Science, 22 (2), 99-113.

Dong, Mo Koo (2003), Inter-relationship among store images, store satisfaction, and store loyalty among Korea discount retail patrons, Asia Pacific Journal of Marketing and Logistics, 15 (4), 42-72.

Ehrenberg, A. & Goodhardt, G.J. (2002) Double jeopardy revisited, again, Marketing Research, 14 (1), 40-42.

Fader, P.S. & Schmittlein, D.C. (1993) Excess behavioral loyalty for high-share brands: Deviations from Dirichlet model for request purchasing, Journal of Marketing Research, 30 (Nov), 478-493.

Farley, J.U. (1964) Why does brand loyalty vary over products?, Journal of Marketing Research, 1, 9-14.

Gounaris, S. & Stathakopoulos, V. (2004) Antecedents and consequences of brand loyalty: An empirical study, Journal of Brand Management, 11 (4), 283-307.

Keith, R.J. (1960) The marketing revolution, Journal of Marketing, 24, 35-38.

Kerin, R.A., Hartley, S.W., Berkowitz, E.N. & Rudelius, W. (2006) Marketing, 8th Edition, Boston: The McGraw-Hill Companies Inc.

Knox, S.D. & Denison, T.J. (2000) Store loyalty: Its impact on retail revenues - An empirical study of purchasing behavior in the UK, Journal of Retailing and Consumer Services, 7, 33-45.

Leavitt, T. (1960) Marketing myopia, Harvard Business Review, (July-August).

Lehmann, D.R. & Winner, R.S. (1997) Analysis for Marketing Planning, 4th edition, Burr Ridge, IL: Irwin.

Leverin, A & Liljander, V. (2006) Does relationship marketing improve customer relationship satisfaction and loyalty?, International Journal of Bank Marketing, 24 (4), 232-251.

McDowell, W.S. & Dick, S.J. (2001) Using TV daypart 'double jeopardy effects' to boost advertising efficiency, Journal of Advertising Research, 41 (6), 43-51.

Mitchell, V.W. & Kiral, R.H. (1998) Primary and secondary store-loyal customer perceptions of grocery retailers, British Food Journal, 15 (7), 312-319.

Murphy, P.E. & Enis, B.M. (1986) Classifying products strategically, Journal of Marketing, 50 (July), 24-42.

Oliver, R.L. (1999) Whence consumer loyalty?, Journal of Marketing, 63, 33-45.

Oliver, R.L. (1997) Satisfaction: A behavioral perspective on the consumer. New York, NY : Irwin McGraw-Hill.

Oliver, R. L. (1980) A Cognitive model of antecedents and consequences of satisfaction decision, Journal of Marketing Research, 17, 460-469.

Pedhazur, E.J. (1982) Multiple Regression in Behavioral Research, Fort Worth, Texas: Holt, Rhinehar, and Winston.

Pleshko, L.P. & J. Cronin, J.J. (1997) An investigation into the relationship between consumer satisfaction, marketing performance, and firm performance in a service environment, Journal of Customer Service in Marketing & Management, 3 (1), 7-17.

Pleshko, L.P. & Heiens, R.A. (1997) An empirical examination of patient loyalty: An application of the customer loyalty classification framework to the healthcare industry, Journal of Customer Service in Marketing and Management, 11 (2), 105-114.

Pleshko, L.P. & Heiens, R.A. (1996) Categories of customer loyalty: An application of the customer loyalty classification framework in the fast-food hamburger Market, Journal of Food Products Marketing, 3 (1), 1-12.

Raj, S.P. (1985) Striking a balance between brand popularity and brand loyalty, Journal of Marketing, 49 (winter), 53-59.

Reichheld, F.F. (2003) The one number you need to grow, Harvard Business Review, 82 (12), 46-54.

Reichheld, F.F. & Sasser, W.E., Jr. (1990) Zero defections: Quality comes to services, Harvard Business Review, 68 (5), 105-111.

Reinartz, W.J. & Kumar, V. (2002) The mismanagement of customer loyalty, Harvard Business Review, 80 (7), 4-12.

Robinson, J. (1979) Brand strength means more than market share, Journal of Advertising Research, 19 (5), 83-87.

Rosenberg, L. & Czepial, J. (1983) A marketing approach to customer retention, Journal of Consumer Marketing, 2, 45-51.

Shankar, V., Smith, A. & Rangaswamy, A. (2003) Customer Satisfaction and Loyalty in Online and Offline Environments, International Journal of Research in Marketing, 20 (2), 153-75.

Stevens, J. (1986) Applied Multivariate Statistics for the Social Sciences, Hillsdale, New Jersey: Lawrence Erlbaum Associates

Story, J. & Hess, J. (2006) Segmenting customer-brand relations: Beyond the personal relationship metaphor, Journal of Consumer Marketing, 23 (7), 406-413.

Tellervision (2006) Banks face big challenges in satisfying customers, Aspen Publisher, # 1345, May.

Turan, S. (2002) An exploratory analyses of customer satisfaction, Journal of the American Academy of Business, 2 (1), 177-184.

Twedt, D.W. (1967) How does brand awareness-attitude affect marketing strategy?, Journal of Marketing, 31 (October), 64-66.

Verhoef, P.C. (2003) Understanding the effect of customer relationship management efforts on customer retention and customer share development, Journal of Marketing, 67 (4), 30-45.

Zeithaml, V.A. (2000) Service quality, profitability, and the economic worth of customers: What we know and what we need to learn, Journal of the Academy of Marketing of Science, 28 (1), 67-85.

Larry P. Pleshko, Kuwait University Samar M. Baqer, Kuwait University

Table 1: General Statistics: Aggregate Derived Indicators
by Store Type

Store Type Store # MSHA% L%TOT L%MST SATF

Health HC2 5.6 100.0 3.7 24.0
Clubs HC9 9.4 85.8 7.4 22.2
 HC3 11.4 69.5 7.4 19.1
 HC10 12.2 74.5 9.9 22.0
 HC13 48.0 89.1 40.7 18.7
Convenience CS8 6.6 27.0 5.7 20.9
Stores CS10 7.1 32.4 8.0 18.8
 CS4 7.5 29.5 10.3 22.7
 CS6 8.7 29.1 10.3 20.4
 CS3 9.0 35.0 12.6 20.3
 CS5 9.5 23.6 6.9 18.0
 CS2 12.0 20.8 9.2 19.1
 CS9 19.1 33.5 16.1 18.7
Medical MC4 3.6 46.8 3.0 24.7
Clinics MC1 3.6 58.3 4.5 22.6
 MC6 4.4 47.0 3.0 24.0
 MC8 16.6 54.4 17.9 21.9
 MC2 17.4 56.4 22.4 24.1
 MC7 42.2 66.0 46.3 19.2
Fast-Food FF3 4.7 18.7 2.2 18.3
Burger FF2 10.1 25.5 6.5 22.8
Outlets FF6 13.1 28.0 16.1 19.8
 FF1 18.3 27.9 21.5 21.2
 FF5 18.7 29.8 16.1 21.2
 FF4 34.8 41.6 37.6 19.9

Table 2: Path Analysis Statistics

Path Path [R.sup.2] Sign
Analysis Eliminated Path

I H2: Loy-MSh .636 +
II H1: Sat-Loy .149 -
III H3: Sat-MSh .004 -

IV Type-MSh .008 n/a

V Type-Loy .463 n/a

VI Type-Sat .249 n/a
VII Loy as mediator .095 n/a

Path Path [R.sup.2 [R.sup.2
Analysis Eliminated .sub.o] .sub.a]

I H2: Loy-MSh .011968 .640356
II H1: Sat-Loy .463000 . 543013
III H3: Sat-MSh .638912 .640356

IV Type-MSh .637456 .640356

V Type-Loy .149000 .543013

VI Type-Sat n/a n/a
VII Loy as mediator .546493 .589576

Path Path [R.sup.2] W
Analysis Eliminated Path

I H2: Loy-MSh .636 24.25441
II H1: Sat-Loy .149 3.872236
III H3: Sat-MSh .004 .096169

IV Type-MSh .008 .192772

V Type-Loy .463 14.92217

VI Type-Sat .249 n/a
VII Loy as mediator .095 2.395668

Path Path 'p' Conclusion
Analysis Eliminated

I H2: Loy-MSh <.001 keep H2 path
II H1: Sat-Loy <.050 keep H1 path
III H3: Sat-MSh <.800 discard
 H3 path
IV Type-MSh <.700 discard
 Typ-MSh
V Type-Loy <.001 keep
 Typ-Loy
VI Type-Sat <.029 keep Typ-Sat
VII Loy as mediator <.150 no mediation

Notes: (a) for all tests, N=25 and d=1; (b) Path VI is
derived from Anova; (c) [R.sup.2] Path for Path II and Path V is
derived by dividing total [R.sup.2] by two due to two loyalty
indicators