Int. J. Services and Standards, Vol. 4, No. 1, 2008 81
Measuring supermarket service quality: proposal for a scale
Eduardo Torres-Moraga and Luis Jara-Sarrua Faculty of Economics and Business Studies University of Chile Diagonal Paraguay 257, Santiago, Chile E-mail: [email protected] E-mail: [email protected]
Jose M. Moneva* Department of Accounting and Finance University of Zaragoza Gran Via 2, Zaragoza, Spain E-mail: [email protected] *Corresponding author
Abstract: The success of the retail industry and especially that of supermarkets is largely determined by the services they offer. Even though service quality scales have been developed in different sectors, because of the unique characteristics of supermarkets, it seems convenient to create a specific scale for them. The results of this study indicate that supermarket service quality is a multidimensional construct (made up of reliability, personal attention, assurance, hygiene, tangibility and accessibility) that is different from the standard constructs previously proposed. These results provide an important tool which enables the assessment and management of supermarket service quality.
Keywords: service quality; supermarkets; satisfaction; structural equations; standards.
Reference to this paper should be made as follows: Torres-Moraga, E., Jara-Sarrua, L. and Moneva, J.M. (2008) ‘Measuring supermarket service quality: proposal for a scale’, Int. J. Services and Standards, Vol. 4, No. 1, pp.81–96.
Biographical notes: Eduardo Torres-Moraga holds a PhD in Business Administration from the Faculty of Economics and Business Studies at the University of Zaragoza, Spain. He is an Assistant Professor of Marketing at the University of Chile, Chile. His main research lines are service marketing and e-banking.
Luis Jara-Sarrua is a PhD candidate in Economic and Business Sciences for the University of Zaragoza, Spain. At present, he is an Assistant Professor in the Faculty of Economy and Business of the University of Chile. His current research interests include corporate social responsibility and the retail industry.
82 E. Torres-Moraga, L. Jara-Sarrua and J.M. Moneva
Jose M. Moneva is an Assistant Professor of Accounting in the Department of Finance and Accounting at the University of Zaragoza (Spain). He is the Vice-Dean for Teaching Affairs at the Faculty of Economics of this university. His current research projects are focused on social and environmental accounting and reporting, and corporate social responsibility.
1 Introduction
Stiff competition in today’s supermarket industry, the limited differentiation which can be attained by means of products sold and increasingly demanding customers have forced supermarkets to concentrate their efforts on providing a range of additional quality services for their customers. Service quality has become a fundamental issue for supermarkets, not just from an operational point of view but also from a strategic perspective.
Well aware of this challenge, supermarket managers have centred their efforts on determining which factors best represent how customers perceive the service quality offered by these establishments and, at the same time, on using this information to improve service processes in order to provide satisfaction and eventually achieve customer loyalty (Abdul et al., 2007).
Although related literature has proposed a series of models and scales to be used for determining service quality in different lines of business (Sasser et al., 1978; Grönroos, 1982; Rust and Oliver, 1994), it is the scale proposed by Parasuraman et al. (1988; 1991) which has been most widely accepted by researchers and marketing managers.
Notwithstanding the fact that this scale (Servqual) has been used to assess supermarket service quality (Morales, 1999), it is known that its application cannot be generalised to all kinds of services (Carman, 1990). In fact, it needs to be adapted to the conditions of each industry, especially when it comes to the supermarket sector. Given the fact that supermarkets commercialise food, their dimensions could differ considerably from those which make up part of the Servqual scale, whose dimensions have proven to be highly relevant for other kinds of services.
Thus, we believe it could be necessary to create a scale specially designed to assess supermarket service quality, that is to say, a scale which contains the specific items and dimensions to be applied to any company in this industry.
The present study aims to design a scale with a high degree of reliability, validity and dimensionality to enable the assessment of supermarket service quality. By means of this instrument, supermarket managers will be able not only to assess the quality of the services they offer, but also to compare them with those of their competitors. In this way, the necessary actions may be taken in order to improve management and, consequently, to meet present and potential customer satisfaction.
In order to achieve this objective, we first carry out an exhaustive analysis of the industry to construct a specific scale for supermarkets. Then, using the concept of the latent variable, we carry out a psychometric analysis of the data, which allows us to design an optimum scale to measure the service quality of supermarkets.
Measuring supermarket service quality: proposal for a scale 83
2 Service quality models
Different models for assessing service quality have been proposed in the literature. In the 1970s, Sasser et al. (1978) proposed a model which integrated some factors that go beyond basic attributes. It considers that consumers translate their expectations into attributes associated not only with the base service but also with adjacent services.
Later, Grönroos (1982) proposed a model which based service quality on two fundamental factors:
1 functional quality, which comes from the way service is provided to the customer
2 technical quality, which is based on characteristics inherent to the service (for example, opening hours, check-out queues, product variety).
Grönroos (1982) considers functional quality a more determining factor than technical quality. In fact, he stated that interaction between the service provider and the customer builds the foundation of quality. Subsequently, Rust and Oliver (1994) added a third dimension to this model, which they called service environment.
In addition to these models, we should highlight one of the most famous scales created to assess the quality of service (Parasuraman et al., 1988; Parasuraman et al., 1991). The Servqual scale considers service quality to be fundamentally made up of two elements: customer perception of service quality and the expectations that consumers have about a specific service. In an earlier study, Parasuraman et al. (1985) proposed ten dimensions made up of 97 items in their construct. All of them were used to assess both consumer expectations and perception. In a subsequent study, the ten dimensions were reduced to five (reliability, responsibility, empathy, assurance and tangibility), made up of a total 22 items or indicators (Parasuraman et al., 1991).
Finally, Cronin and Taylor (1992) developed the Servperf scale, which is a modification of the Servqual scale, since the latter was not suited to assessing service quality. In their study, four different opinions were analysed in order to assess service quality based on variables proposed by the Servqual scale. The results enabled them to conclude that the Servperf scale is superior to the Servqual scale in several aspects. This conclusion has been further supported by subsequent studies (Brown et al., 1993; Boulding et al., 1993; Hartline and Ferrell, 1996; Avkiran, 1999).
3 Analysis of the Servqual scale
The real effectiveness of the Servqual scale has been questioned in the literature (Robinson, 1999). Some authors have argued that this scale only assesses delivery processes and not service results, which is what should really be assessed (Buttle, 1996).
Other authors have focused their criticism on the validity of the dimensions making up the scale, arguing that they have not been universally proven (Babakus and Boller, 1992) and that they are not as generic as the scale’s creators have proposed (Carman, 1990).
Other criticisms have focused on expectations as a key factor in the assessment of service quality. Boulding et al. (1993) propose that performance alone should be used to measure service quality and not the gap that exists between performance and expectations. This criticism of the Servqual scale is directly related to the findings of
84 E. Torres-Moraga, L. Jara-Sarrua and J.M. Moneva
Teas (1993) and Babakus and Manglod (1992). Both studies determined that there was little theoretical and empirical evidence to support the idea that customers value service quality in terms of the gap between performance and expectations and, in addition, determined that it is consumer perception (performance) which should be considered when it comes to assessing service quality. This result is backed up by the proposal of Avkiran (1999). According to this author, customers’ expectations are always higher than their perceptions, which turns the gap into something unavoidable, and therefore it would make no sense to use it to assess service quality.
Some authors have been even more specific in their analysis, pointing out that what renders the Servqual scale inappropriate is its low discriminatory validity, a result of the high correlation it presents between expectations and performance (Brown et al., 1993). This does not occur with the Servperf scale. In addition, these authors argue that expectations, as opposed to performance, show a high average and low standard deviation, which restricts the variance of the Servqual scale. This was supported by Cronin and Taylor (1994), who found that assessment using the Servperf scale explains greater variance than the Servqual scale.
Finally, it is relevant to mention that in addition to this criticism, the use of expectations by the Servqual scale has generated some problems even for questionnaire applications. The lengthy content of the questionnaire because of the separation of expectations and performance means that the person being interviewed is asked the same questions twice (Brown et al., 1993). This is not the case with the Servperf scale, since it reduces the number of questions by 50% (Teas, 1993) and thus facilitates the pooling of data and their effectiveness (Newman, 2001).
4 Methodology
In order to obtain a scale with enough reliability and validity, the procedure suggested by Deng and Dart (1994) was applied. The first stage consists of building an assessment scale and trying to ensure its content validity, in order to subsequently develop the questionnaire and pool data for a representative sample. The second stage consists of using the previously collected data to analyse the psychometric properties of the assessment scales. This procedure aims to identify the most suitable scales for assessing supermarket service quality.
The Servqual scale proposed by Parasuraman et al. (1991) was used as the basis to develop a scale with a suitable degree of content validity. This scale was subsequently adjusted by means of an exhaustive analysis recommended by De Wulf and Odekerken-Schröder (2003). A series of analyses was used consisting of successive focus groups made up of customers from different sectors of the city of Santiago, Chile, as well as interviews with the marketing managers of different supermarkets. These activities permitted the analysis of the different dimensions and indicators included in the initial construct proposed by this study (Bagodi and Mahanty, 2007). This analysis enabled the addition of dimensions and/or indicators which more suitably represent supermarket service quality. It also enabled the readjustment and/or elimination of dimensions and/or indicators which are conflictive or redundant for the concept studied.
The scale was finally made up of six dimensions and not the five proposed in the Servqual scale. To be precise, the dimensions of empathy and responsiveness (now called personal attention) turned out to belong to a single dimension and not separate
Measuring supermarket service quality: proposal for a scale 85
ones as proposed in Parasuraman et al. (1991). Furthermore, two new dimensions (hygiene and accessibility), not considered in the Servqual scale, were incorporated. These new dimensions, especially hygiene, were very important for the participants in the focus groups.
Lastly, we carried out a quantitative pretest on a random sample of 45 customers of different supermarkets, and with the data obtained, we carried out a factorial exploratory analysis and calculated the Cronbach alpha for each of the resulting dimensions. This prior analysis confirmed the existence of each of the dimensions that came from the focus groups. As a result, the definitive questionnaire was made up of 31 questions or items grouped into the dimensions of reliability, personal attention, assurance, hygiene, tangibles and accessibility (see the Appendix). These dimensions differ from those contained in the Servqual scale that has been used in previous studies to measure service quality in supermarkets (Morales, 1999).
All the items of the scale were presented in a positive sense, and those surveyed were required to answer using a Likert scale from 1 (completely disagree) to 7 (completely agree). Service quality was assessed in keeping with the concept proposed by Cronin and Taylor (1992) in their Servperf scale. This consists of assessing service quality considering only consumer perception and not expectations.
The field study was performed during June and July 2006 on a sample of 405 customers. The sample was representative of the market quotas of the supermarkets that operate in Chile. Of the people surveyed, 34.6% were between 25 and 35 years old and 40.5% between 36 and 55 years old. Fifty-four percent were women and 66.8% had some level of university studies.
5 Analysis of the scales and results
5.1 Sifting of the measurement scale and its properties
A psychometric analysis of data was performed once the interviews were completed. In order to create a construct which would enable the assessment of service quality as a latent variable that could also produce a suitable degree of reliability, validity and dimensionality, the study used the analysed data in three stages. The first was based on an exploratory study which consisted of maintaining exclusively those items which enabled the analysis of other dimensions or factors with a suitable degree of reliability or unidimensionality. The second stage was a confirmatory study which discarded those items which did not enable suitable dimensionality for the entire construct known as ‘supermarket service quality’. In the last stage, the construct which came out of this entire process was subjected to final validity and reliability analysis.
5.2 Exploratory analysis 5.2.1 Analysis of initial reliability
During this first stage, all items which hampered the scales from reaching a suitable degree of reliability were eliminated. In all cases, the Cronbach alpha statistic produced values greater than 0.6, which is the minimum recommended level for exploratory analysis (Hair et al., 1998), and levels even higher than 0.7, which is recommended for confirmatory analysis (Nunnally, 1978; Hair et al., 1998). In fact, the reliability
86 E. Torres-Moraga, L. Jara-Sarrua and J.M. Moneva
dimension produced a value of 0.74, personal attention 0.90, assurance 0.72, hygiene 0.90, tangibles 0.82 and accessibility 0.80. Notwithstanding these good results, one of the scales did not meet the minimum items-total subscale correlation of 0.3 established for this analysis (Nurosis, 1993). The ASSU4 variable, which belongs to the assurance dimension, showed a correlation of 0.29, below the determined limit. Therefore, this variable was eliminated in order to increase the internal consistency of the assurance dimension.
5.2.2 Unidimensionality analysis of each of the subscales
This analysis was used to test whether each of the six subscales or dimensions which make up the service quality construct showed a suitable degree of unidimensionality. An exploratory factorial analysis was carried out on each of the subscales considered, applying principal components and, when necessary, Varimax rotation. The latter was implemented alone to identify the indicators which are less associated with a determined factor (Hair et al., 1998).
As can be seen in Table 1, each of the subscales showed a suitable degree of unidimensionality, with the high variance explained and with factorial burdens which largely surpassed the limit of 0.25 proposed by Hair et al. (1998) for these to be significant.1 According to these results, the elimination of the subscale indicators analysed was considered unnecessary.
Table 1 Exploratory factorial analysis of service quality scales
In this stage, a more exhaustive analysis of the supermarket service quality assessment scales was performed by means of a confirmatory factorial analysis using the structural equations method.
Considering the latent variables which represent the service quality construct, an improvement process was performed using a model development strategy (Hair et al., 1998), which consists of eliminating the indicators (or variables) which are less suitable for achieving proper adjustment. This variable elimination process generates successive models until it reaches the model which provides the best adjustment measures, dimensionality and a suitable number of variables for each subscale (Ding et al., 1995).
The process was carried out considering the three criteria proposed by Jöreskog and Sörbom (1993). The first criteria of weak convergence would eliminate indicators that did not have a significant factorial regression coefficient t-student > 2.58 (p = 0.01).
The second criteria of strong convergence would eliminate those indicators that were not substantial, i.e., those whose standardised coefficient (λ) was less than 0.5.
Lastly, it has been proposed to eliminate the indicators that least contribute to the explanation of the model, considering the cut-off point as R2 < 0.3.
The structural equation method was used for this analysis (Bentler, 1992). The first criterion was not applied to eliminate indicators because they all indicated strong convergence in their corresponding latent variables, in all cases surpassing a t of 2.58. However, when it came to the second criterion, it was necessary eliminate the REL4, REL3 and ACCE1 indicators, since they did not meet the limit established for this parameter (λ > 0.5) (see Table 2).
When the third criterion was applied, the ACCE6 variable was eliminated from the study because it presented an R2 which was at the limit of 0.3, leading to a fifth version of the model (see Figure 1). Although we attempted to eliminate other variables higher than 0.3, the model did not show significant improvements in its adjustment measures to make such a decision. Therefore, it was decided that the process should be stopped at this fifth version, which represents the best model for assessing supermarket service quality (see Table 2).
88 E. Torres-Moraga, L. Jara-Sarrua and J.M. Moneva
Measuring supermarket service quality: proposal for a scale 89
5.4 Reliability and dimensionality analysis
Three analyses were performed in order to determine the reliability of service quality: Cronbach alpha, composite reliability coefficient (Jöreskog, 1971) and the average variance extracted (Fornell and Larcker, 1981). In both the Cronbach’s alpha analysis and the composite reliability coefficient, a scale is considered to be reliable when it gives values equal to or greater than 0.7 (Hair et al., 1998). The average variance extracted requires values equal to or greater than 0.5.
The results indicate that, in all cases, Cronbach’s alpha indexes and composite reliability coefficients greatly surpass the minimum recommended level of 0.7 (Table 3). In addition, the elimination of some indicators enabled the improvement of Cronbach’s alpha indexes for the reliability and assurance dimensions. However, observing the indexes produced by the variance extracted analysis, we can see that all dimensions show indexes below 0.5, but much higher than 0.4, which indicates that in most dimensions this parameter is relatively reliable.
Table 3 Reliability of the subscales of service quality
Reliability Personal attention Assurance Hygiene Tangibles Accessibility
Cronbach alpha
0.77 0.85 0.73 0.77 0.76 0.75
Composite reliability coefficient
0.71 0.85 0.73 0.77 0.76 0.75
Average variance extracted
0.45 0.45 0.48 0.46 0.45 0.43
A rival model strategy (Hair et al., 1998; Flavian and Guinaliu, 2006) was implemented in order to determine whether the service quality variable in a supermarket context is multidimensional. This compared two alternative models, a first-order model and a second-order model (Steenkamp and Van Trijp, 1991). The analyses indicate that the second-order model (which is the model in which the service quality construct is represented by six dimensions) shows better adjustment levels than the first-order model (in which all indicators are weighted into just one factor). This analysis confirms the multidimensionality of the service quality variable in a supermarket context.
5.5 Validity analysis
In our study, to measure the scale validity, the scheme outlined by Nunnally (1978) was followed. According to this author, for validity to exist, it should be checked that the scale used has content validity, construct validity and criterion-related validity.
Content validity is largely guaranteed because the scales used to assess supermarket service quality have been designed based on a detailed analysis of the related literature. In addition, the scales have been subjected to both the judgement and discussion of different experts as a subsequent sifting process.
Construct validity is formed by two fundamental categories of validity: convergent and discriminatory. The model proposed will fulfil the condition of convergent validity if the latent variables that make it up are strongly and significantly correlated. Observing
90 E. Torres-Moraga, L. Jara-Sarrua and J.M. Moneva
Figure 1, which represents the model coming from the sifting process applied by means of confirmatory factorial analysis, we can state that the correlations between the latent variables which make up the service quality scale are, in all cases, relatively high and significant at 0.01 (t > 2,58). The results, therefore, confirm the presence of convergent validity.
The proposed scales will have discriminatory validity if the factors used to assess them are not valid to assess other constructs or, in other words, if they are more correlated between themselves than with other factors or latent variables which are less related to the studied concept. The Chi-squared difference test and the confidence interval test (Anderson and Gerbing, 1988) were used to determine the presence of discriminatory validity.
The first test consists of comparing Chi-squared values between the model coming from the confirmatory factorial analysis and different alternative models. Each of the 15 alternative models which were compared with the model coming from the confirmatory factorial analysis were made up of the same dimensions contained in the resulting model, but with the difference that a perfect correlation was assigned to two of these dimensions. Discriminatory validity will exist when the Chi-squared difference produced by this comparison is significant. As indicated in Table 4, the differences between the theoretical model produced by confirmatory factorial analysis and the different alternative models are highly significant in the cases studied.
Table 4 Discriminatory validity test for the service quality construct
Constructs Construct pairs Test differences χ2 (d.f.); p-value Confidence intervals
In the case of the confidence interval test, there is discriminatory validity between the two latent variables when the confidence intervals produced by determining the correlation between latent variables do not include the value ‘1’. Table 4 shows that in no case does the value ‘1’ occur. In addition, in all cases, the correlations were considerably distant from this value.
Measuring supermarket service quality: proposal for a scale 91
Therefore, with these results, we can state that the service quality model proposed
shows discriminatory validity between the latent variables which make it up. Finally, criterion-related validity analysis aims to determine whether the concept
analysed is related to the criteria values, which theory indicates they should be related to. Give that the data was pooled at the same time, a special kind of criterion-related validity will be applied, concurrent validity.
To determine whether this type of validity is fulfilled, the hypothesis that service quality might influence customer satisfaction was proposed (Taylor and Bullard, 1993; Kristensen et al., 1999; Martensen et al., 2000; Subramony et al., 2004; Li et al., 2004; Liu and Yun, 2005; Sadig and Saeed, 2005; Türkyilmaz and Özkan, 2007). This hypothesis was tested using the structural equations model.2 For this analysis, we used the service quality construct which came from the previous psychometric analysis of data, as well as a scale which was created to assess supermarket customer satisfaction.
As can be seen in Figure 2, there is an evident cause-and-effect relationship between perceived supermarket service quality and customer satisfaction. In addition, both the explanatory capacity of the model (R2 = 0.80) and its adjustment levels (CFI 0.929; RMSR 0.084; IFI 0.930; NFI 0.911; NNFI 0.902) are high. This enables us to deduce that the supermarket service quality construct proposed shows adequate concurrent validity and, therefore, that the criterion-related validity of this concept is satisfactorily backed up.
Figure 2 Concurrent validity. Relationship between Service Quality (SQ) and Satisfaction (SAT)
0.44R2=0.80
SAT1
SAT2
SAT3
0.42= 0.82
0.55
R2
0.42
R2 0.91*
0.91*
0.83*SAT
0.44R2 = 0.80
SAT1
SAT2
SAT3
R2
R2 = 0.70
R2 = 0.83
SAT
0.77*
0.73*
0.61*
0.74*0.62
R2 = 0.62
0.64R2 = 0.59 REL
PER
ASSU
HYG
TANG
0.64
R2 = 0.59
ACC0.68
R2 = 0.54ACC
0.68
R2
0.68
R2 = 0.530.80
R2 = 0.37 0.77*
0.79*
SQ 0.90*
Note: *Significant at the 0.01 level.
Finally, considering all the analyses performed in this study, it can be concluded that the subscales finally proposed to assess supermarket service quality provide a suitable level of reliability, validity and dimensionality (see the Appendix).
6 Conclusions, managerial implications and future research lines
This paper strives to contribute to the literature on the assessment of supermarket service quality, as perceived by consumers, in order to establish a starting point which will enable the managers of such organisations to establish closer relationships with their customers.
From this point of view, the objective of this study was to develop a valid and reliable scale to assess supermarket service quality as perceived by consumers, different from the Servqual scale (Morales, 1999) used in previous research.
92 E. Torres-Moraga, L. Jara-Sarrua and J.M. Moneva
Thus, the Servqual scale has only been considered as a starting point, incorporating adaptations suggested by experts and customers who justify the need to design a new scale in order to assess service quality for the industry.
Given that the scale emerging from this study shows a good degree of reliability, validity and unidimensionality in each of its dimensions, we should point out some of the differences it has with respect to the Servqual scale, which has been used to measure supermarket service quality (Morales, 1999). One of the fundamental differences is that it incorporates the dimensions of accessibility and hygiene, the latter being especially relevant in this industry.
The set of dimensions included in the scale proposed is completed with reliability, customer service, safety and tangibles. Therefore, the scale which was finally selected was made up of six subscales which are clearly related and integrated in one exclusive construct, demonstrating the latent and multidimensional nature of the supermarket context. In fact, this study proved that service quality can be assessed from different perspectives, all of which are integrated into one exclusive assessment instrument, applied to supermarket customers in this case.
This creates important challenges for supermarket managers, since many factors must be considered at the same time when planning and managing service quality. Keeping these different dimensions in mind and appreciating their relative importance is essential nowadays, when differentiation has become a key consideration for supermarkets. In fact, these companies have made substantial efforts to provide more and better services for their customers, such as a greater number of checkouts, larger parking lots, internet purchasing services and home delivery.
All these aspects contribute to forming a general but at the same time detailed perception of supermarkets in the mind of the consumer. The results of the empirical analysis carried out in this study indicate that supermarket marketing managers should focus especially on service reliability; customer service; safety of the building and surrounding areas; hygiene of the establishment, employees and food handling; physical facilities; and access to products and services provided by the supermarket.
As for service quality management for this sector, it is essential that managers clearly understand the parts played by different elements which make up the quality of service offered by the supermarket and understand how consumer behaviour and trust change in the event of variations in each of the factors analysed.
In short, we could say that, within the increasingly competitive context of supermarkets, managers must be well aware of the consumer perception of service quality when making strategic plans for their companies.
Finally, it is important to highlight that it would be a good idea for future studies to further analyse the validity of this scale, attempting to test its discriminatory validity by carrying out alternative tests related to service quality concepts with other constructs significantly different from those analysed in this study. Similarly, it would be convenient to test the criterion-related validity analysis carried out in this study. This could be done by means of a predictive validity analysis to determine whether the service quality construct analysed is able to predict future variables other than satisfaction. Likewise, it would be interesting to repeat the study in other geographic locations in order to test more reliably the possibilities of scale extrapolation developed in this study.
Measuring supermarket service quality: proposal for a scale 93
Acknowledgements
Jose M. Moneva acknowledges the financial support from the Ministerio de Ciencia y Tecnología, the European Social Fund (research project SIRCARSO-SEJ2006-08317) and Gobierno de Aragón (GESES research group).
References Abdul, M.R., Choy, S. and Lin, B. (2007) ‘Service quality of a local Malaysian bank: customers’
expectations, perceptions, satisfaction and loyalty’, International Journal of Services and Standards, Vol. 3, No. 1, pp.18–38.
Anderson, F.C. and Gerbing, D.W. (1988) ‘Structural equation modelling in practice: a review and recommended two-step approach’, Psychological Bulletin, Vol. 103, No. 3, pp.411–423.
Avkiran, N.K. (1999) ‘Quality customer service demands human contacts’, International Journal of Bank Marketing, Vol. 17, No. 2, pp.61–71.
Babakus, E. and Boller, G.W. (1992) ‘An empirical assessment of the SERVQUAL scale’, Journal of Business Research, Vol. 24, pp.253–268.
Babakus, E. and Mangold, W.C. (1992) ‘Adapting the SERVQUAL scale to hospital services: an empirical investigation’, Health Services Research, Vol. 26, No. 6, pp.767–786.
Bagodi, V. and Mahanty, B. (2007) ‘Micro and macro analysis of service quality of two-wheeler service in India: to capture the potential market’, International Journal of Services and Standards, Vol. 3, No. 1, pp.39–63.
Bentler, P.M. (1992) EQS: Structural Equations Program Manual, Los Angeles: BMDP Statistical Software.
Boulding, W., Kalra, A., Stealing, R. and Zeithaml, V. (1993) ‘A dynamic process model of service quality: from expectations to behavioural intentions’, Journal of Marketing Research, Vol. 30, pp.7–27.
Brown, T.J., Churchill, G.A., Jr. and Peter, J.P. (1993) ‘Improving the measurement of service quality’, Journal of Retailing, Vol. 69, No. 1, pp.127–139.
Buttle, F. (1996) ‘SERVQUAL: review, critique, research agenda’, Journal of Marketing, Vol. 30, No. 1, pp.8–32.
Carman, J.M. (1990) ‘Consumer perceptions of service quality: an assessment of the SERVQUAL dimensions’, Journal of Retailing, Vol. 66, Spring, pp.33–35.
Cronin, J.J. and Taylor, S.A. (1992) ‘Measuring service quality: a reexamination and extension’, Journal of Marketing, Vol. 56, July, pp.55–68.
Cronin, J.J. and Taylor, S.A. (1994) ‘SERVPERF versus SERVQUAL: reconciling performance-based and perceptions-minus-expectations measurement of service quality’, Journal of Marketing, Vol. 58, January, pp.125–131.
De Wulf, K. and Odekerken-Schröder, G. (2003) ‘Assessing the impact of a retailer’s relationship efforts on consumer’ attitudes and behaviour’, Journal of Retailing and Consumer Services, Vol. 10, No. 2, pp.95–108.
Deng, S. and Dart, J. (1994) ‘Measuring market orientation: a multi-factor, multi-item approach’, Journal of Marketing Management, Vol. 10, pp.725–742.
Ding, L., Velicer, W.F. and Harlow, L.L. (1995) ‘Effects of estimation methods, number of indicators per factor and improper solutions on structural equation modelling fit indices’, Structural Equations Modelling, Vol. 23, pp.119–143.
Flavian, C. and Guinaliu, M. (2006) ‘Consumer trust, perceived security and privacy policy. Three basic elements of loyalty to web site’, Industrial Management & Data Systems, Vol. 106, No. 5, pp.601–620.
94 E. Torres-Moraga, L. Jara-Sarrua and J.M. Moneva
Fornell, C. and Larcker, D. (1981) ‘Evaluating structural equations model with unobserved variables and measurement error’, Journal of Marketing Research, Vol. 18, pp.39–50.
Grönroos, C. (1982) Strategic Management and Marketing in the Service Sector, Swedish School of Economics and Business Administration, Sweden: Helsingfors.
Hair, J.F., Jr., Anderson, R.E., Tatham, R.L. and Black, W.C. (1998) Multivariate Data Analysis, 5th ed., Prentice Hall.
Hartline, M. and Ferrell, O. (1996) ‘The management of customer-contact service employees: an empirical investigation’, Journal of Marketing, Vol. 60, pp.52–70.
Jöreskog, K.G. (1971) ‘Statistical analysis of sets of cogeneric test’, Psycometrika, Vol. 36, pp.109–133.
Jöreskog, K.G. and Sörbom, D. (1993) New Feature in LISREL 8, Chicago: Scientific Software.
Kristensen, K., Martensen, A. and Gronholdt, L. (1999) ‘Measuring the impact of buying behaviour on customer satisfaction’, Total Quality Management, Vol. 10, pp.602–614.
Li, B., Riley, M. and Hsieh, C-T. (2004) ‘Assessing customer satisfaction in parcel delivery industry: an empirical study among university customers’, International Journal of Services and Standards, Vol. 1, No. 2, pp.172–192.
Liu, C-M. and Yun, C. (2005) ‘The multidimensional and hierarchical structure of perceived quality and customer satisfaction’, International Journal of Management, September, Vol. 22, No. 3, pp.426–435.
Martensen, A., Gronholdt, L. and Kristensen, K. (2000) ‘The drivers of customer satisfaction and loyalty: cross-industry findings from Denmark’, Total Quality Management, Vol. 11, pp.544–553.
Morales, M. (1999) ‘Assessing the cross-cultural applicability of a service quality measure. A comparative study between Quebec and Peru’, International Journal of Service Industry Management, Vol. 10, No. 5, pp.449–468.
Newman, K. (2001) ‘Interrogating SERVQUAL: a critical assessment of service quality measurement in a high street retail bank’, International Journal of Bank Marketing, Vol. 19, No. 3, pp.126–139.
Nunnally, J.C. (1978) Psychometric, 2nd ed., New York: McGraw-Hill.
Nurosis, M.J. (1993) SPSS. Statistical Data Analysis, SPSS Inc.
Parasuraman, A., Berry, L. and Zeithalm, V. (1991) ‘Refinement and reassessment of the SERVQUAL scale’, Journal of Retailing, Vol. 64, No. 4, pp.420–430.
Parasuraman, A., Zeithalm, V. and Berry, L. (1985) ‘A conceptual model of service quality and its implications for future research’, Journal of Marketing, Vol. 49, No. 4, pp.41–50.
Parasuraman, A., Zeithalm, V. and Berry, L. (1988) ‘SERVQUAL: a multiple-item scale for measuring customer perceptions of service quality’, Journal of Retailing, Vol. 64, Spring, pp.12–40.
Robinson, S. (1999) ‘Measuring service quality: current thinking and future requirements’, Marketing Intelligence and Planning, Vol. 17, No. 1, pp.21–32.
Rust, R.T. and Oliver, R.L. (1994) ‘Service quality: insights and managerial implications from the frontier’, in R. Rust and R. Oliver (Eds.) Service Quality: New Directions in Theory and Practice, Thousand Oaks, CA: SAGE Publications, pp.1–19.
Sadig, M. and Saeed, A. (2005) ‘Measuring service quality at King Fahd International Airport’, International Journal of Services and Standards, Vol. 1, No. 4, pp.482–493.
Sasser, W., Olsen, R.P. and Wyckoff, D. (1978) Management of Service Operations: Test and Cases, Boston: Allyn & Bacon.
Steenkamp, J. and Van Trijp, H. (1991) ‘The use of LISREL in validating marketing constructs’, International Journal of Research in Marketing, Vol. 8, pp.283–299.
Measuring supermarket service quality: proposal for a scale 95
Subramony, M., Beehr, T. and Johnson, C.M. (2004) ‘Employee and customer perceptions of
service quality in an Indian firm’, Applied Psychology: An International Review, Vol. 53, No. 2, pp.311–327.
Taylor, S.A. and Bullard, W. (1993) ‘Recreational service quality in the international setting’, International Journal of Service Industry Management, Vol. 4, pp.68–86.
Teas, R.K. (1993) ‘Expectations, performance evaluation and consumer’s perceptions of quality’, Journal of Marketing, Vol. 57, No. 4, pp.18–34.
Türkyilmaz, A. and Özkan, C. (2007) ‘Development of a customer satisfaction index model. An application to the Turkish mobile phone sector’, Industrial Management & Data Systems, Vol. 107, No. 5, pp.672–687.
Notes
1 According to Hair et al. (1998), for a reliability level of less than 95%, such as our case, and for a sample size of approximately 400 (405 in our case), the minimum value which each factorial loading should show over the main factor is 0.25.
2 We must consider that, in this model, the scale which represents service quality is made up of six indicators. These indicators were calculated as the average of all the indicators which were finally selected after sifting the scales and which represent each of the service quality dimensions.
96 E. Torres-Moraga, L. Jara-Sarrua and J.M. Moneva
Appendix
Definitive scale for assessing supermarket service quality
Reliability When I have a problem, supermarket personnel show true interest in helping me to solve it.
REL1
The supermarket performs the service right the first time. REL2 Product prices are clearly marked on the shelves. REL*** The supermarket respects advertised prices (flyers, gondolas). REL*** Products sold are generally of good quality. REL5 Personal attention Supermarket employees provide fast service. PER1 Supermarket employees are always willing to help me. PER2 The supermarket has personnel available to meet my needs. PER3 Supermarket employees give me personalised attention. PER4 The supermarket really keeps my interests in mind. PER5 Supermarket employees understand my specific needs. PER6 Supermarket employees treat me courteously. PER7 Assurance Supermarket employees are prepared to answer my questions. ASSU1 The supermarket is safe to shop at. ASSU2 There are enough security personnel for the size of the supermarket. ASSU3 The supermarket has clearly indicated emergency exits. ASSU4* Hygiene Products sold by the supermarket are fresh. HYG1 Product packaging dates and best-before dates are clearly labelled. HYG2 The supermarket is concerned with its hygiene. HYG3 Food is handled hygienically at the supermarket. HYG4 Employees are clean and well-groomed. HYG5 Tangibles The supermarket’s physical facilities are visually attractive. TANG1 The supermarket surroundings are attractive. TANG2 There is enough space in the supermarket. TANG3 Environmental conditions at the supermarket are pleasant (smell, background music, decoration and lighting).
TANG4
Accessibility The supermarket has enough parking spaces. ACCE1*** Products are easy to find. ACCE2 The supermarket has enough checkout lanes to make the checkout process fast. ACCE3 The supermarket has affordable prices. ACCE4 The supermarket layout is comfortable. ACCE5 The supermarket features a wide variety of products and brands. ACCE6***
Notes: Each item was measured with a seven-point Likert scale.
