The Effect of Advertising Spending on Brand Loyalty Mediated by Store Image, Perceived Quality and Customer Satisfaction: A Case of Hypermarkets

First, the study aims to investigate the effect of advertising spending on brand loyalty. Second, to test the mediating role of store image, perceived quality and customer satisfaction in above mentioned relationship. A quantitative approach was employed, using 15-item, 5-point Likert scale questionnaire administered to 360 participants. Data from the departmental stores are used to examine the direct and indirect impact of advertising spending on brand loyalty. Confirmatory Factor Analysis and Structural Equation Modeling were carried out using Amos-18 to evaluate the results. The study found complex results of advertising spending on brand loyalty. The results showed insignificant direct impact of advertising spending on perceived quality and brand loyalty where as significant direct impact on customer satisfaction and store image. However, store image and perceived quality are confirmed as mediators in the relationship between advertising spending and brand loyalty. The inclusion of brand trust variable can better describe the relationship of advertising spending with brand loyalty. The results of the study highlights the value of advertising spending in building brand loyalty that suggests managers of the departmental stores should give emphasis on advertising spending and on store image that lead towards brand loyalty. It contributes to the current body of knowledge by improving the understanding of advertising spending and brand loyalty. The major contribution of this study is the comparison of theoretical model with mediating model to evaluate the effect of advertising spending on brand loyalty along store image, perceived quality and customer satisfaction.


INTRODUCTION
Internet and e-commerce has paved the way for globalization and in the individuals living in the current age have access to wide range of global products and services. Due to the advancement in the technology, it is now possible for the firms to produce goods according to the requirements of the customers. Brand provides opportunity to the firms to capitalize on its competences and unique organizational structure to get a place in the minds and hearts of the consumers (Keller, 1993). A brand is defined as a name, term, sign, symbol or features which identify one seller's goods and services from the others (Bennett, 1995). The product only provides functional benefits whereas brand provides functional as well as emotional benefits that lead towards satisfaction (Hankinson and Cowking, 1996). Intensity of competition in the present world has made marketers to practice more focused branding strategies, to build and sustain loyal customer's base by offering them superior value. Organizations dedicate billions of dollars every year on the promotional activities. Most of the promotional budget comprises of advertising, as it aims to aware and inform the customers about new and existing products launched by the company.
Brand image is defined as the set of beliefs held about a particular brand (Kotler, 1988). It is a set of associations usually organized in some meaning full mode (Aaker, 1992). The strong brand actually influences consumer perception and brand loyalty primarily focuses on marketing mix variables (Yoo et al., 2000). Nowadays due to the proliferation of media, it made difficult for one advertisement to get people's attention. Hence, winning the market share with the help of advertising has become a challenge for the firms, competing in the current market age. There is a general assumption in literature, the favorable brand image have positive impact on consumer behavior. The advertising spending helps to protect brands and firms loyal customers from competing brands (Agrawal, 1996). Advertising spending has direct as well as indirect effect on brand loyalty. Advertisements enhance consumer knowledge and perception pertaining to the brand and ultimately strengthen the brand loyalty . Brand loyalty is positively associated with profitability (Hallowell, 1996). Brand loyalty is a deeply held commitment to rebuy a preferred brand or service consistently in future (Oliver, 1999). In order to establish a link among these variables, there are various studies which are conducted from time to time. Different firms are using different strategies to get to know about the perception of their brand in the eyes of the customers. Retail industry is the growing industry of the country and focusing on advertising strategies to create awareness among people. In the current age, the loyalty towards the retail stores has also been observed and the customers tend to purchase the desired products from certain preferred retail outlets. In this regard, competition among the retail stores is also increasing.
Trends in the market are always changing and show the influence of competition in the market, the price war among the firms/brands always stress on giving the best product in best suitable price. With the improvement in the technology, different firms use different ways to attract new customers and to increase awareness among them. Firms and brands actually use the goodwill and image, to gain the trust and commitment of their customers and to make them loyal. Advertising is considered as defensive strategy, through that brand loyalty builds; it helps to retain the loyal customers. Beside this, price promotions are considered as offensive strategy, which attract loyal customers away from the competing brands (Agrawal, 1996). Due to the proliferation of advertising options and lots of media choices, in addition to the conventional media, it is tough for the companies to choose a strategy that can emotionally hit consumers.
The study will check the effect of advertising spending on brand loyalty in retail outlets mediated by store image, perceived quality and customer satisfaction. The present study is an attempt to check whether the proposed theoretical model of advertising spending could apply to emerging market economy such as: Pakistan. It is important to check model in different cultures in order to generalize the results. This study extends the body of literature by comparing two models: proposed theoretical model and mediating model. Beside this, the relationship between advertising spending and customer satisfaction will also be explored. The objectives of the study are to contribute towards a contemporary issue of brand loyalty in retail outlets in Pakistan. The study will analyze the impact of advertising spending on brand loyalty that how it will contribute to make customers loyal. The study empirically evaluates the following construct relationships: • To find out the direct effect of advertising spending on brand loyalty • To find out the indirect effect of advertising spending on brand loyalty mediated by perceived quality, store image and customer satisfaction • To find out the direct effect of store image, perceived quality and customer satisfaction The study will begin with the literature to define the constructs and to establish support for the relationships. Next, theoretical model presented, which is tested to establish causality among the variables. Finally, results and implications for the managers along limitations are discussed.

LITERATURE REVIEW
Brand loyalty is defined as a deeply held commitment to re-buy a preferred product or service consistently in the future, thereby causing repetitive same brand or same brand set purchasing, despite situational influences and marketing efforts, having the potential to cause switching behavior (Oliver, 1999). There are different perspectives in the literature through which brand loyalty can examine. Brand loyalty is investigated with the help of two perspectives: attitudinal loyalty and behavioral loyalty (Dick and Basu, 1994). Brand loyalty is positively associated with profitability (Hallowell, 1996). Loyalty is one of the most important factor which describe the success of business organizations, brand loyalty is directly linked with the customer satisfaction and brand image (Kandampully and Suhartanto, 2000). There is a positive link between the retail brand and loyalty of the customers (Binninger, 2008).
Advertising spending: Advertising spending is studied as consumer perception of advertising frequency and expenditure. Advertising spending is found to have direct and positive impact on brand loyalty, store image and perceived quality. There is a significant relationship between advertising spending and perceived quality. Moreover, the primary effect of advertising is to increase brand name recognition (Moorthy and Zhao, 2000). Millions of dollars spend every year on advertising that leads towards the brand loyalty and results in creating loyal customers attached to particular brand or firm (Chioveanu, 2008). Advertising is important tool in creating awareness among customers and intense advertising spending is the indicator of high quality and shows, company is investing in brand/product. Therefore, there is a positive and significant association between advertising spending and perceived quality (Aaker and Jacobson, 1994). Furthermore, the literature indicates, brand loyalty and perceived quality both have positive influence on the brand, the more advertising for the brand; the more consumers have high perceived quality and loyalty for the brand (Yoo et al., 2000). Advertising is the most important and traditional way of awaring and informing the customers. Radio, television and newspapers are the most traditional ways for creating image of brands and help in generating more sales and revenues. With the passage of time, there were several technologies which become a part of marketing environment and facilitated the marketers for achieving their objectives (Howard, 2010). The advertising spending has direct as well as indirect impact on brand loyalty . The study has five factors, in which the effect of advertising spending has checked on all other variables. In the study, following hypotheses are proposed as: H1: Advertising spending has direct and positive impact on brand loyalty. H2: Advertising spending has direct and positive impact on store image. H3: Advertising spending has direct and positive impact on perceived quality. H4: Advertising spending has direct and positive impact on customer satisfaction.

Store image:
Brand image is one of the most significant assets that is positively related with satisfaction of the customers (Porter and Claycomb, 1997). Store image could be achieved with the help of sustained financial performance. Perceived quality of a brand is related positively to the extent to which the brand is distributed through stores with a good image (Yoo et al., 2000). Furthermore, store image leads to satisfaction and it leads to store/brand loyalty (Martenson, 2007). Store image develops with the perception and beliefs about particular store and formed over time. A store with a good image reflects that the brands it have are of high quality and customers are loyal with it. A store image is related to the advertising spending and perceived quality (Miller and Berry, 1998). There is difference of perception about store image among domestic and international market. Store atmosphere is very important that leads to customer satisfaction. In addition to this, the perceived quality of services is identified as a critical factor, leading to the satisfaction of the customer's by means of overall credibility and reputation of the store (Chang and Luan, 2010). There is a direct relationship between brand image and customer loyalty, as the brand image of the firm/product or brand increases in the eyes of the customers, the loyalty of the customers also flourishes (Ogba and Tan, 2009). The favorable store image is derived from shopping attributes which leads to customer patronage intentions. These intentions further leads to the loyalty of customers towards the e-tail store (Yun and Good, 2007). The attributes of store are considered as basis for building the relathionship with retailers (Wang and Ha, 2011). To check the relathionship, following hypotheses are stated as: H5: Store image has direct and positive impact on brand loyalty. H6: Store image has direct and positive impact on perceived quality. H7: Store image mediates the relationship of advertising spending and brand loyalty.
Perceived quality: Previous researches have shown a positive association between perceived quality and brand loyalty and positive relationship of perceived quality with customer satisfaction. Perceived quality is actually the customer's perception about the overall quality of the brand. There is important function of perceived quality on behavior of consumer that actually identifies the effect that different points and levels of perceived quality have on satisfaction and purchase intentions (Tsiotsou, 2005). Perceived quality have direct and positive impact on brand loyalty and customer satisfaction and when the perceived quality of the brand increases, the customer satisfaction also raises, thus leads to brand loyalty (Ha et al., , 2009. The customers who are satisfied are more loyal and the customers who have high perceived value have strong relationship among customer loyalty and customer satisfaction as compare to the customers who have low perceived value (Chang and Wang, 2010).
The perceived service quality with customer satisfaction and customer trust affect customer loyalty. There is significant relationship among perceived quality and customer satisfaction. Word of mouth communication played very important role in generating the loyalty (Kassim and Abdullah, 2010). To examine the relationship, following hypotheses are proposed as: H8 : Perceived quality has a direct and positive impact on brand loyalty. H9 : Perceived quality has direct and positive impact on customer satisfaction. H10 : Perceived quality mediates the relationship of advertising spending and brand loyalty.
Customer satisfaction: Satisfaction is a performance indicator and previous researches have shown positive relationship between these variables. Consumers who have different level of perception are also different in their level of satisfaction (Tsiotsou, 2005). Consumer satisfaction is considered as fundamental determinant and most of the studies investigated the impact of satisfaction on post consumption evaluation such as attitudinal loyalty (Cooil et al., 2007). Earlier studies have shown such relationship between satisfaction and brand loyalty. Satisfaction has direct effect on brand loyalty (Ha et al., 2009;Selnes, 1993) and loyalty can be an outcome of customer satisfaction (Oliver, 1997). Satisfaction act as a mediator among the relationship of brand loyalty and brand equity. The satisfaction is a key indicator to make customers loyal in long run and it is a determinant of long term business. The consumers develop loyalty with the brand because the particular brand experience fits with the lifestyle of the person. The physical surroundings have positive impact in making customers loyal thus creating brand loyalty (Nam et al., 2011). The mediating affect of satisfaction on brand loyalty have positive relationship with each other and have positive impact on brand loyalty.
Perceived quality also positively affect satisfaction and brand loyalty . loyalty is one of the most important factor which describes the success of business organization (Kandampully and Suhartanto, 2000). Positive relationship among the quality dimensions and customer satisfaction were found which positively effects customer loyalty (Jamal and Anastasiadou, 2009). Following hypotheses are stated as:

H11
: Customer satisfaction has a direct and positive impact on brand loyalty. H12: Customer satisfaction mediates the relationship of advertising spending and brand loyalty.

METHODOLOGY
In this section, the development of scale and data collection procedure is discussed. In the next section, Confirmatory Factor Analysis (Amos-18) is explained to establish the validity of the construct. Lastly, Structural Equation Modeling is used to compare the theoretical model and mediated model.
The research design is based on primary data and it is collected through questionnaire. The questionnaire with instructions of how to complete is distributed among customers. Convenience sampling is used to collect the data from the customers of departmental stores. There are approximately 10 large and medium sized departmental stores in Islamabad and Rawalpindi and six departmental stores are selected on the basis of average customer visit frequency per day. Those stores are selected who have customer visit frequency more than 300/day. The sample size of 360 customers is used to collect the data that are geographically spread in order to overcome the problem of generalizability. The sample size is enough for this type of analysis Nguyen et al., 2011;Chang and Wang, 2010;Nam et al., 2011). Total 334 usable questionnaires are found to analyze the results. Invalid questionnaires were removed from the sample.

Research instrument:
There were five constructs in the study, advertising spending, store image, perceived quality, customer satisfaction and brand loyalty. All constructs were measured with the items adapted from different previous studies. Three items for advertising spending is adapted from Kirmani and Wright (1989) and Yoo et al. (2000), 3 items for perceived quality is adapted from Yoo et al. (2000), 3 items for store image is adapted from Stern et al. (1977), 2 items for satisfaction is adapted from Ragunathan and Irwin (2001) and 4 items for brand loyalty is adapted from Sirgy and Simli (1985). It has fifteen items that measures all the variables of researcher interest. The response scale for the study is 5 point Likert scale, anchored by 1 strongly agrees and 5 strongly disagree. The questionnaire has translated into Urdu language and bilingual questionnaire is used for the ease of the customers. To test the hypothesized relationships presented in theoretical model were conducted using Amos-18. First, the measurement model was accessed by Confirmatory Factor Analysis (CFA) and hypothesized relationships by Structural Equation Modeling (SEM).
Demographic statistics: Demographic statistics shows out of 334, 59% represents male respondents where as 41% represents female respondents. On participant's age, 38% represent age group of 16-25, 28% are representing age group of 26-35, 18% represents age group of 36-45 and 15% represents age group of 46-55 where as 1% of respondents represent 55 and above age group. On marital status, 63% respondents are married, 34% are single where as 3% are divorced. On educational level, 2% of respondents has education till primary, 10% have education till secondary, 42% belong to bachelor's level while 46% pursuing degree of higher education. On income level, 16% of respondents belong to 20,000 or below income level, 23% of respondents belongs to 20,001-40,000 income level, 21% belongs to 40,001-60,000 income level, 17% belongs to 60,001-80,000 income level, 11% belongs to 80,001-100,000 income level and 12% belongs to 100,000 or above income level. On occupational level, 33% of respondents belong to government service, 29% of respondents belong to private service and 21% respondents are housewives while 17% respondents are business man. On question of how often you visit to store, 20% respondents answered that they visit store on weekly basis, 21% of respondents visit store on fortnightly basis, 53% of respondents visit store on monthly basis and 6% of respondents visit store on quarterly basis. On question of, how you first time know about store, 4% of respondents answered that they know through print ad, 3% of respondents know through T.V ad, 8% of respondents through magazine ad, 5% respondents through pamphlet, 6% of respondents through poster, 16% of respondents through store signage, 8% of respondents through banner of the store, 5% of respondents through billboard, 2% of respondents through broacher and 43% of the respondents know about the store through reference of family/friend. Before starting analysis, the normality and reliability of the data is checked. The data is normally distributed and reliability of overall instrument is Cronbach's alpha 0.770.
Correlation statistics: Correlation illustrates the direction and strength of linear relationship between two variables. Table 1 shows the correlation statistics. Correlation analysis was employed to access the possible association between each variable: advertising spending, store image, perceived quality, customer satisfaction and brand loyalty. The outcome of the analysis shows that advertising spending and perceived quality has positive small correlation r = 0.27 and statistically significant at p = 0.01, advertising spending and store image has small correlation r = 0.144 (Meenaghan, 1995) thus significant at the same level, advertising spending and customer satisfaction has significant relationship r = 0.125, whereas advertising spending and brand loyalty has medium correlation r = 0.260, as the advertising spending increases, store image, perceived quality, customer satisfaction and brand loyalty increases. The perceived quality and store image has large correlation r = 0.572, when perceived quality increases, there is increase in store image, perceived quality and customer satisfaction also has large correlation r = 0.577 (Tam, 2011), with the increase in perceived quality, customer satisfaction also rises, where as perceived quality and brand loyalty has medium correlation r = 0.431 (Nguyen et al., 2011). The store image and customer satisfaction has large correlation r = 0.669 (Martenson, 2007), it means both have positive effect on each other, the boast in store image leads to increase in customer satisfaction and store image and brand loyalty also has large correlation r = 0.581 (Miller and Berry, 1998), when store image increases, brand loyalty also increases. The customer satisfaction and brand loyalty has medium correlation r = 0.492, when customers are satisfied, they are more loyal (Binninger, 2008). Thus, the results show the change in one variable due to the change in other variable. The data shows positive correlation among all variables. The mean of a data ranges from 2.2 to 3.2 which shows that majority of people are agreed with the statements asked about the departmental stores. The data do not show high standard deviation. The main reason of that, majority of the people have somehow same opinion except customer satisfaction because, for being satisfied every individual have different opinion and perception of every individual also varies.

MEASUREMENT MODEL
Measurement model is drawn to carry out the Confirmatory Factor Analysis. Maximum Likelihood Method is used to calculate the parameters. Table 2 shows the CFA Model and Table 3 shows the fit indices of CFA. Confirmatory Factor Analysis is done rather exploratory because established measurement scale is used, it check the validity and reliability of the instrument even it is adopted model, because the model is being tested in new environment, in new industry with new sample. It describes how well the fifteen items calculated the five constructs. Four analyses were accomplished to access the measurement model.
Firstly, CFA was performed using Amos-18 to examine the appropriateness of the proposed theoretical model. The overall fit of the model is acceptable. The chi square statistics (3.2) is significant in the sample (p<0.01) because it is highly susceptible and sensitive to the sample size. The Goodness of Fit (GFI) estimates were 0.926, respectively, Comparative Fit Index (CFI) estimates were 0.902 provides a good fit to data.
Secondly, unidimensionality was accessed before examining the validity and reliability. For this purpose, CFI was conducted for each of five variables on measurement model. The CFI of every construct is between 0.86 to 1.00 shows good fit, thus signifies evidence of unidimensionality (Hair et al., 1995;Sureshchandar et al., 2002).   Thirdly, to test the construct reliability; squared multiple correlations ( $ ) for every measurement item, composite reliability and average variance extracted for each factor is used in the study. The ( $ ) for every item is ranges between 0.37 to 0.62, thus shows good reliability (Holmes-Smith, 2001), it shows variance in variable provided by latent construct. Thus, all constructs show good fit, as well as are unidimensional. The Cronbach's alpha values were calculated to measure the reliability of the constructs. All values lie between 0.611 to 0.74, thus shows good indication of internal consistency and reliability. All composite reliabilities ranges between 0.611 to 0.74 ensure the validity of the measures (Anderson and Gerbing, 1998).
Fourthly, the convergent validity of the measurement items was obtained by the factor loadings and composite reliabilities. The standardized factor loadings are between 0.36 to 0.79 and are greater than suggested level of 0.35 and are significant, these are actually good indicators of CFA (Hair et al., 1995). CFA reveals, all items significantly loaded on individual constructs (p<0.05), except one item of brand loyalty variable, that is removed for the further analysis (Anderson and Gerbing, 1998). Discriminant validity of the construct was evaluated by calculating the shared average variance extracted between the pair of constructs, its value ranges between 0.80 to 1.2 (Fornell and Larcker, 1981). The discriminant validity was also assessed with a method of Campbell and Fiske (1959), value below 0.85 suggests discriminant validity. On the basis of these criteria, it is concluded that study exhibits convergent and discriminant validity. Table 4 shows the fit indices of SEM. The absolute fit indices actually test the fitness between recommended and observed variance-covariance matrix. The results of proposed theoretical model showed, the chi square for goodness of fit is 2.858, CMIN 151.452 with 53° of freedom, it is a first statistic that gives information about the fit of data with the model. Its values should be less than 5. CMIN is a chi square statitic, it is an indication of how much the data is fit with the model. The Goodness of Fit Index (GFI) tells the amount of variance in the sample variance co-variance matrix accounted for by the model; in this case its value is 0.939, quite good. RMSEA is root mean square error of approximation; its value is 0.075, also fine. RMR is root mean square residual; it is an index that shows difference between the estimated variance and covariance from the observed variance, smaller is better. In particular case its value 0.06 is acceptable.

Structural equation modeling:
The incremental fit indices actually compare the theoretical model with the null model (independent model). Normed Fit Index (NFI), is the difference between the two models chi square divided by chi square of independent model, in this case its value is 0.887 and is acceptable. AGFI is an adjusted GFI; it is an alternate GFI index in which the value of the index is adjusted for the number of parameters in the model. The less the number of parameters in the model, the closer the AGFI will be to the GFI. In this case the values of both are closer to each other, 0.895 and 0.939. CFI is comparative fit index, 0.922 values shows good fit.
Parsimonious fit indices are actually used to test the theoretical model fits with the number of estimated coefficients. It explains that model fitness could not be achieved by "over fitting". As, the initial measurement model shows acceptable fit indices, there in the study modification indices has not applied, shows no chances of over fitting. PNFI is the parsimony normed fit index, it is the product of NFI and PRAITO, its value is 0.602, is acceptable. PGFI is 0.547, also fine. So all these terms are measuring fit of the model and in general terms data is fit with the model. Hypotheses testing: The measurement model was estimated with CFA and research hypotheses were tested on the basis of standardized structural coefficients with SEM. Details is shown in Table 5. H1: Stated that advertising spending has direct and positive impact on brand loyalty. The correlation between advertising spending and brand loyalty is positive and significant small correlation (r = 0.260). The standardized structural coefficients showed advertising spending has insignificant relationship with brand loyalty (β = 0.15, p>0.05), hence H1 is not supported. H2: Hypothesized that advertising spending has direct and positive impact on store image. The results of correlation between advertising spending and store image showed positive and significant small correlation (r = 0.144). The standardized structural coefficients showed advertising spending has significant positive relationship with store image (β = 0.25, p<0.01), hence H2 is supported. H3: Stated that advertising spending has direct and positive impact on perceived quality. The correlation results demonstrated advertising spending and perceived quality has positive and significant small correlation (r = 0.273). The standardized structural coefficients showed advertising spending has insignificant relationship with perceived quality (β = 0.17, p>0.05), thus rejects H3. H4: Hypothesized that advertising spending has direct and positive impact on customer satisfaction. The correlation results showed advertising spending and customer satisfaction has significant small correlation (r = 0.125). But the standardized structural coefficients showed advertising spending has significant association with customer satisfaction (β = 0.68, p<0.01), therefore accepts H4.
H5: Stated that store image has direct and positive impact on brand loyalty. The results of correlation between store image and brand loyalty is positive and significant large correlation (r = 0.581). The standardized structural coefficients showed store image has direct and positive impact on brand loyalty (β = 0.541, p<0.01), hence H5 is supported. H6: Hypothesized that store image has direct and positive impact on perceived quality. The results of correlation between store image and perceived quality showed positive and significant large correlation (r = 0.572). The standardized structural coefficients showed store image has direct and positive impact on perceived quality (β = 0.763, p<0.01), thus accepts H6. H7: Will be discussed in mediating roles of variables in next section. H8: Stated that perceived quality has direct and positive impact on brand loyalty. The results of correlation of perceived quality on brand loyalty showed positive and significant medium correlation (r = 0.431). The standardized structural coefficients showed perceived quality has significant impact on brand loyalty (β = 0.329, p<0.05), thus accepts H8. H9: Stated that perceived quality has direct and positive impact on customer satisfaction. The results of correlation of perceived quality on customer satisfaction, showed large, positive and significant correlation (r = 0.577). The standardized structural coefficients showed perceived quality has significant impact on customer satisfaction (β = 0.90, p<0.01), hence supported H9. H10: Will be discussed in mediating roles of variables in next section. H11: Stated that customer satisfaction has direct and positive impact on brand loyalty. The correlation of customer satisfaction with brand loyalty showed positive medium and significant correlation (r = 0.492). The standardized structural coefficients showed, the bond between customer satisfaction and brand loyalty is insignificant (β = 0.116, p>0.05), thus rejects H11. H12: Will be discussed in mediating roles of variables in next section.

Mediating roles of SI, PQ and CS:
Mediation is tested using Amos-18; two SEM models were developed and compared. The first model was the proposed theoretical model (Fig. 1), had all direct and indirect relationships between the variables and second model was the mediated model (Fig. 2), in which advertising spending influence brand loyalty through store image, perceived quality and customer satisfaction. It has only relationships via mediating variables and don't have any direct relationship with brand loyalty. The approach used in the study is similar to other studies (Yen and  Table 5, both models fit with the data and the chi square difference test indicated no significant difference. Subsequently, the models were evaluated on the basis of following criteria (Morgan and Hunt, 1994;Yen and Gwinner, 2003). The criteria are: the overall fit of the model by CFI, amount of variance explained by R square, parsimony evaluated by parsimonious fit index PNFI and Akaike's Information Criteria (AIC). Both models also fit on the overall statistics. From the nine proposed theoretical model paths, six model paths were supported at (p<0.01 and p<0.05). Beside this, in mediating model, five out of six model paths were supported. The variance in the outcomes was explained by model as measured by R 2 value. The store image and customer satisfaction explained by mediated model is much better than proposed theoretical model (0.802, 0.65) (0.98, 0.82) respectively. On the other hand, the perceived quality explained by proposed theoretical model is better than mediated model (0.67, 0.60). However, the proposed theoretical model is much better at explaining the brand loyalty than mediating model (0.98, 0.95). Moreover, the PNFI of mediated model is greater than proposed research model (0.742, 0.602). The Akaike's Information Criteria (AIC) are a modification of goodness of fit and it is helpful in making comparison between two models (Tabachnick and Fidell, 2000). A model with the lowest AIC value is considered to have good fit and in this case, mediated model have lowest AIC 201.529 is consider as best fit as compared to proposed theoretical model, whose value is 295.858. As, the difference between two models are small, but still, mediated model is more accurate and significant than proposed research model. In addition to this, mediated model accurately explain the relationships between constructs. According to that, H7 and H10 are accepted, that store image and perceived quality mediates the relationship of advertising spending and brand loyalty. It means when the store advertise heavily, it create awareness among customers, establish good store image in the minds of the customers as well as build high perceived quality, than customers become loyal with the store. But H12 is rejected that customer satisfaction mediates the relationship of advertising spending and brand loyalty because mediated model show insignificant relationship between customer satisfaction and brand loyalty. The path coefficients explain whether the path AD to CS and CS to BL is statistically significant or not. As the product of this path is insignificant, it means we do not have evidence for mediation, thus rejects H12.

CONCLUSION AND DISCUSSION
The rationale of the study was to find the effect of advertising spending on brand loyalty mediated by store image, perceived quality and customer satisfaction. The study was specifically designed to measure the effect of mediating variables on brand loyalty. The study examines the impact of advertising spending on brand loyalty and contributes to existing body of literature by considering the mediating impact of store image, perceived quality and customer satisfaction on departmental stores in Pakistan by comparing two models, theoretical model and mediating model. The findings of the study showed number of similarities and differences from previous studies.
Results showed that advertising spending has insignificant direct relationship with brand loyalty. Ha et al. (2011) also found insignificant relationship between advertising spending and brand loyalty, it means, no direct effect of advertising spending on brand loyalty. It is because most of the departmental stores are not advertising heavily and people cannot make difference between the Ads of two departmental stores that which Ad is more expensive, people cannot make difference between the Ads or they cannot relate advertising with brand loyalty, for them to be loyal with the store, there are many other reasons than advertisements. The stores are not running Ads on television; no hoardings etc., as compared to the stores in Western countries, that's why results are insignificant in Pakistan. If they do advertise at large scale, they can inform general public and their customers about the products available at the store in order to make them aware and loyal. It helps to generate sales and create revenues. It implies that advertising spending effects brand loyalty through store image and perceived quality. Advertising spending influence perception of the customers, that plays major role in shaping attitude of the individuals.
It is also found that advertising spending has insignificant impact on perceived quality. There are many reasons of this, as in the study, customer perception of firm advertising is used to measure advertising spending, the frequency and expenditure of Ads. People might not compare frequency and expenditure of Ads with the perceived quality and that's why results were not supported. The perceived quality is not the quality of any product as departmental stores are taken, so in this case, perceived quality is actually the quality of store in terms of environment of the store, ambience, location, parking area etc., which people consider important for predicting the quality of a store. The perceived quality of the stores also does not depend upon the products available in the store but depends upon the services provided by the management, how they interact and deal with the customers and customer's friendly environment.
Results showed that advertising spending has direct impact on store image and customer satisfaction, as the advertising increases, the image about the store in the minds of the customers also rises that leads towards customer satisfaction, because advertising is actually a tool to create awareness and when customers are aware about the store than they have good perception about the store. Martenson (2007) found store image leads to satisfaction and it leads to store/brand loyalty. Store image develops with the perception and beliefs about particular store and formed over time. There are different attributes are associated with the store and it plays very important role in building brand loyalty and perceived quality. Miller and Berry (1998) found that store with a good image reflects that the brands it have are of high quality and customers are loyal with it.
Moreover, it is observed that store image has direct and positive impact on brand loyalty and perceived quality. As the image of the store increases, the perception about the store also enhances and people become loyal with the store. The relationship of store image with perceived quality and brand loyalty is also significant with the literature (Yoo et al., 2000). The study suggests that store image acts as a mediator in explaining the advertising spending and brand loyalty. It explains, people have favorable attitude towards a store.
In addition to this, the results showed that perceived quality has direct and positive impact on customer satisfaction and brand loyalty. As the perceived quality of the store increases, satisfaction of the customers also increases thus leads towards brand loyalty. It is found that perceived quality has significant impact on customer satisfaction and brand loyalty (Nguyen et al., 2011). Nam et al. (2011) explored physical quality has positive impact on customer satisfaction with mediating role of brand experience. Beside this, Tam (2011) found that customer perceived value has positive impact on customer satisfaction and brand loyalty. Chang and Wang (2010) found that customer perceived value has positive impact on customer satisfaction and loyalty. Hong-Youl and Kang-Hee (2012) explained that perceived quality have direct impact on brand loyalty and satisfaction. Yoo et al. (2000) showed the perceived quality of a brand has a positive relationship with a brand that is distributed with a good brand image. It is observed that there is a stronger relationship between perceived quality and customer satisfaction, it could be due to people who have high perceived value for the store in terms of environment of the store and services etc., are more satisfied with the store. When the perceived quality of the customers increases, the satisfaction automatically amplify thus leads to brand loyalty. Perceived quality also acts as a mediator in explaining the relationship of advertising spending with brand loyalty. It means people think about store that it is of high quality and reliable, it shapes attitude of the individuals in positive way.
The contribution of the study is the impact of advertising spending on customer satisfaction; this adds to the body of literature that advertising spending has direct and positive impact on customer satisfaction. Moreover, it is also confirmed from the literature that these variables has positive impact on each other (Martenson, 2007).
Moreover the study explained the mediating roles of store image and perceived quality that plays very important part in explaining advertising spending and brand loyalty. Store image and perceived quality showed significant and positive relationship with brand loyalty. Advertising spending influence brand loyalty via these mediating variables.
Furthermore, the results of customer satisfaction with brand loyalty are contradictory with the previous studies. Customer satisfaction does not lead to brand loyalty, in this way it is not playing part as a mediator in the study. There could be many reasons for that, as the people are different, every individual or customer has its own perception. Satisfied customer is not loyal because of location, might be the individual is living far away from the store that he/she cannot come to store for shopping even he/she is satisfied with the store. It might be the store is not in the access of the customers. The culture of the store or the culture of the country is different that's why it yield insignificant results. The other reason that could be the person satisfied with the branded products available at the store but not with the store itself. Beside this, there could be the price factors; different sales taxes and GST are included at final purchase payments.

LIMITATIONS AND FUTURE RESEARCH
The major limitation of the study was only one industry has adopted to analyze the results. Sample size was not too much big and convenience sampling was used because in Pakistan as mostly people does not provide their information to stores and are not cooperative that's why they don't want to participate in the study. In the study demographic variables are not used to find the effect of advertising spending on brand loyalty. So future research might use these variables to explore the relationships. As the literature identify some other variables such as brand equity and brand trust as important variables that can affect the brand loyalty, should be studied in future research. Future studies might enlarge the sample size that can give more accurate results. Beside departmental stores, shopping malls etc should be used to find the impact of advertising spending on brand loyalty. In future research a comparison study between different industries could be made to access the impact of advertising spending on brand loyalty. As in this study for the advertising spending, perception about the expenditure and frequency of ads are used so in future research actual measures of advertising should be used to access the impact on brand loyalty e.g., which type of Ad is able to attract your attention, which medium is more powerful to create awareness etc.

Managerial implications:
The findings of the study present suggestions for managers of departmental stores through which they can enhance brand loyalty. It is suggested that departmental stores should focus on the advertising because very rare advertising is done that is very important in order to create awareness. From the study it is came to know, majority of people know about store through reference of family and friend, departmental stores should focus on advertising through print Ad, magazine Ad, pamphlets and broachers etc. In this way they can make customers aware about the new products and offerings. As people consider the quality of the store is very high and reliable. Beside this store image is confirmed as mediator in explaining the relationship between advertising spending and brand loyalty. Thus, the store image and perceived quality of the store should be made perfect in the eyes of the customers, so that people think positive about particular store. Finally, as people respond, they will not buy from other store if particular store is available, so they should open new branches and outlets that facilitate people who are living far from the stores.

Advertising spending:
The store (Brand) is advertised intensively. The ad campaigns for my store (Brand) are frequently seen. The ad campaigns for my store (Brand) seem very expensive, compared to campaigns for competing brands.

Perceived quality:
The store (Brand) is of high quality. The likelihood that my store (Brand) is reliable is very high. The likely quality of my store (Brand) is extremely high.