This article was downloaded by: [Universite De Paris 1]

On: 06 August 2013, At: 12:55

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Total Quality Management

Publication details, including instructions for authors
and subscription information:
http://www.tandfonline.com/loi/ctqm19

SERVQUAL: A tool for measuring

patients' opinions of hospital
service quality in Hong Kong
Simon S. K. Lam
Published online: 25 Aug 2010.

To cite this article: Simon S. K. Lam (1997) SERVQUAL: A tool for measuring patients'
opinions of hospital service quality in Hong Kong, Total Quality Management, 8:4, 145-152,
DOI: 10.1080/0954412979587

To link to this article: http://dx.doi.org/10.1080/0954412979587

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information
(the “Content”) contained in the publications on our platform. However, Taylor
& Francis, our agents, and our licensors make no representations or warranties
whatsoever as to the accuracy, completeness, or suitability for any purpose
of the Content. Any opinions and views expressed in this publication are the
opinions and views of the authors, and are not the views of or endorsed by
Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and
Francis shall not be liable for any losses, actions, claims, proceedings, demands,
costs, expenses, damages, and other liabilities whatsoever or howsoever caused
arising directly or indirectly in connection with, in relation to or arising out of the
use of the Content.

This article may be used for research, teaching, and private study purposes.
Any substantial or systematic reproduction, redistribution, reselling, loan, sub-
licensing, systematic supply, or distribution in any form to anyone is expressly
 forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
Downloaded by [Universite De Paris 1] at 12:55 06 August 2013
 T O TA L Q U ALIT Y M AN AG EM EN T , VO L . 8 , N O . 4 , 1 997 , 145± 1 52

SERVQ U AL: A tool for m easuring patients’

opinions of hospital service quality in H ong
Kong

S IM ON S. K. L AM
School of Business, U niversity of Hong Kong, Pokfulam, H ong K ong
Downloaded by [Universite De Paris 1] at 12:55 06 August 2013

A bstract The service quality m easurement scale (SERV QUAL) has been widely used in research
to measure quality of service. T he aim of this paper is to demonstrate the use of SER VQU AL for
m easuring patients’ perceptions of health care quality in H ong K ong. T he paper also exam ines the
validity, reliability and predictive validity of SERV QUAL and analyzes its applicability to the
health care sector in H ong Kong. The results indicate that SER VQU AL appears to be a consistent
and reliable scale to m easure health care service quality. However, the proposed ® ve dimensions of
SER VQU AL are not con® rmed. T he results also indicate that perceived health care service
performance generally falls short of expectations except in the physical elem ents of service quality.
T im ely, professional and competent services are what patients expect from health care providers.

Introduction

D uring th e past decade, concern for service quality reached unprecedented levels. The
present `quality revolution’ has been ® red by exacerbated com petition and m any com panies
have now accepted the challenge of improving service quality. In the health care sector, th e
past few years have witnessed an increasing concern regarding the quality of prim ary health
care in H ong Kong. The health care service can be broken down into two quality dim ensions:
technical quality and functional quality (Donabedian, 1980; G ronroos, 1984). Technical
quality in th e health care sector is de® ned prim arily on the basis of the technical accuracy of
the m edical diagnoses and procedures or the conform ance to professional speci® cations.
Functional quality refers to the m anner in which the health care service is delivered to th e
patients. Q uality, from the professionals’ perspective, is technical and has been operational-
ized in term s of three constructs: structure, process and outcom e (D onab edian, 1980).
Structure pertains to whether the health care providers have the know ledge, skill and
resources to diagnose and treat the patients’ health condition properly. Process concerns
whether diagnostic and therapeutic interventions are ap plied appropriately. O utcom e re¯ ects
whether professional and technical resource inputs produce the expected effect.
R esearch has shown that technical quality falls short of being a truly useful m easure for
describing how patients evaluate the quality of a m edical service encounter (Bow ers et al.,

E-m ail: sim onlam @hkucc.hku.hk.

0954 ± 4127 /97/040145 ± 08 $7.0 0 Ó 1997 Carfax Publishing Ltd

 146 S . S . K . LAM

1994). A lthough technical quality has high priority with patients, m ost patients do not have
the know ledge to evaluate effectively the quality of the diagnostic and therapeutic interven-
tion process. D espite various techniques suggested for evaluating technical quality, like peer
review or m edical protocols, this inform ation is not generally understood or available to th e
patients. C onsequently, patients base their evaluation of quality on interpersonal and en-
vironm ental factors, which m edical professionals have always regarded as less important.
M oreover, m ost patients cannot distinguish betw een the `caring’ (functional) perform ance
and the `curing’ (technical) perform ance of m edical care providers (W are & Snyder, 1975).
Patients’ inability to evaluate the technical aspects of care m eans that m ost of them base their
evaluation of the m edical care process on th e functional perform ance of the m edical care
providers. For exam ple, th e heart attack patients m ay not be com petent to criticize th e
m edications chosen, but surely can evaluate the responsiveness and kindness of staff when
conducting the m edical procedures. Th e exp ressive type of quality is de® ned as `quality in
perception’ Ð the subjective quality as the service receiver sees it (Friedm an, 1979). D uring
Downloaded by [Universite De Paris 1] at 12:55 06 August 2013

the service process, functional quality produced and felt by the patient is frequently an even
m ore important variable in¯ uencing patients’ perceptions of service quality. Q uality in
perception is the result of the patient’ s com parison of his/her perception of the m edical
service encounter experience with his/her pre-encounter expectations (Gronroos, 1984). A
m edical service encounter achieves quality in perception when it m eets or exceeds the level
of th e patients’ expectation. These value perceptions, in turn, affect patients’ satisfaction and
intention to purchase additional services (Zeitham l, 1988). W hile the importance of func-
tional quality to the health care provider is obvious and unequivocal, its m easurem ent and
explanation have presented problem s to health care researchers and m anagers. C learly,
functional quality is m uch m ore dif® cult to evaluate (Zeitham l, 1988). U nlike technical
quality, for which there are objective m easurem ent instrum ents, patients have fewer objective
cues and have relied on their subjective evaluation to judge the level of functional quality.
The developm ent of the SE RVQ U AL scale by Parasuram an et al. (1988) has provided
an instrum ent for m easuring functional service quality ap plicab le across a broad range of
services. In its original form , SE RVQ U AL contains 22 pairs of Likert-type item s. O ne-half of
these item s m easure the resp ondents’ expected levels of service for a particular service
industry. The other 22 m atching item s m easure the perceived level of service provided by a
particular organization within that service industry. Service quality is m easured by th e
difference in scores (the gap scores) betw een the perceived level and the expected level of
service provided. Parasuram an et al. (1988) also contend th at th ere are ® ve dimensions of
service quality that are applicable to service-providing organizations in general. These
dimensions are:

(1) Tangibles: physical facilities, equipm ent and appearance of personnel.

(2) R eliability: ability to perform the prom ised service reliable and accurately.
(3) R esponsiveness: willingness to help custom ers and provide prom pt service.
(4) Assurance: knowledge and courtesy of employees and their ability to insp ire trust and
con® dence.
(5) Empathy: caring, individualized attention provided to custom ers.

The SE RVQ U AL scale has been widely used in research and its psychom etric properties have
also been exam ined by a num ber of studies (Babakus & B oller, 1991; C arm an, 1990; Finn
& Lam b, 1991). A lthough the collective ® ndings by and large provide som e support for th e
validity, reliability and predictive validity of the scale, the factor-loading patterns in th e
original ® ve dimensions are inconsistent across these studies. The usefulness of SE RVQ U AL
rem ains unresolved. There is also little research regarding the validity of SE RVQ U AL in a
 SER VQU AL 147

hospital environm ent outside the U S. The aim of this paper is to dem onstrate the use of an
instrum ent (SER VQ UA L) for m easuring patients’ perceptions of health care quality in Hong
K ong. The paper also exam ines the validity, reliability and predictive validity of SE RVQ U AL
and analyzes its applicability in the health care sector in H ong K ong. The outcom e of th e
m easurem ent process is to provide m anagem ent with additional inform ation for enhancing
service quality in health care organizations.

M ethod

Subjects and procedure

The sam ple consisted of tw o groups of patients attending a 2-day sem inar in Hong K ong.
A ltogether 38 patients attended the ® rst sem inar and another 46 patients attended the sam e
Downloaded by [Universite De Paris 1] at 12:55 06 August 2013

program m e 6 m onths later. The tw o sam ples showed no m arked differences in the m easures
and hence were com bined. The subjects have been hospitalized in H ong Kong for at least one
day in the past 6 m onths. They ranged in age from 20 to 42 with a m ean age of 27.4 years;
71% were m ale.
The subjects were asked to ® ll in a self-com pletion questionnaire. Eighty-four question-
naires were distributed, of which 83 were com pleted and returned. All but one of th e
returned questionnaires were usable. The sam ple showed no signi® cant differences in th e
respondents’ sex and age. Thus, a ® nal sam ple of 82 was obtained and used in the analysis.

M easures

The m ain section of th e questionnaire contained the SE RVQ U AL scale, with 22 statem ents
relating to patients’ expectations of the quality of the service that excellent hospitals should
offer and 22 corresponding item s relating to perform ance perceptions of th e quality of service
actually delivered. A ® ve-point Likert response form at (ranging from `strongly disagree 5 1’
to `strongly agree 5 5’ ) was adopted instead of the original seven-point scale form at. This
m odi® cation was based on the questionnaire pre-test experience with 12 patients, which
indicated that th e ® ve-point form at wo uld reduce the frustration level of the respondents, and
would increase the response rate and the quality of th e responses. The pre-test also indicated
that the m ixture of negatively and positively wo rded statem ents created confusion and
frustration on the part of respondents. Therefore, the negatively worded statem ents con-
tained in the original scale were rephrased and converted to positive ones. The instrum ent
also contained a question about the patient’ s overall rating of hosp ital service quality, and a
question about whether or not he/sh e intended to return to the sam e hospital if additional
needs were to arise in the future. A ® nal set of questions pertained to the respondent’ s
dem ographic characteristics.

R esults

The adequacy of th e SE RVQ U AL scale for assessing patients’ perceptions of service quality
was exam ined in accordance with the recom m endations provided in the m easurem ent
literature (N unnally, 1978). The analyses conducted related to the scale’ s reliability, under-
lying dim ensionality and predictive validity.
 Downloaded by [Universite De Paris 1] at 12:55 06 August 2013
148

Table 1. Expectation and performance perceptions means scores (N 5 82)

Perceptions (P) Expectations (E) Service Quality

Items in each dim ension X SD Item -to-total X SD Item -to-total (P 2 E)

correlations correlations
S . S . K . LAM

Tangibles a 5 0.711 a 5 0.645

(1) Has up-to-date equipm ent 3.63 0.31 0.58 2.45 0.73 0.51 1.18
(2) Physical facilities are visually appealing 3.41 0.84 0.65 2.45 0.43 0.49 0.96
(3) Em ployees are neat in appearance 4.01 0.77 0.62 2.54 0.61 0.52 1.47
(4) Materials are visually appealing 3.98 0.72 0.78 3.06 0.59 0.51 0.92

Reliability a 5 0.723 a 5 0.755

(5) W hen prom ises to do something, it does so 3.61 1.03 0.68 3.56 0.77 0.58 0.05
(6) Shows sincere interest to solve your problem s 2.22 0.85 0.68 4.01 0.47 0.61 2 1.79
(7) Performs the service right the ® rst tim e 4.21 0.87 0.66 4.84 0.92 0.60 2 0.63
(8) Provides services at the tim e it prom ises 3.98 1.25 0.64 4.05 1.20 0.71 2 0.07
(9) Keeps accurate records 3.94 0.52 0.88 4.02 0.66 0.62 2 0.08

Responsiveness a 5 0.902 a 5 0.777

(10) Tells you when the services will be perform ed 3.32 1.03 0.76 3.65 0.67 0.63 2 0.33
(11) Gives prompt services 3.90 0.87 0.70 4.60 0.86 0.57 2 0.70
(12) Alw ays w illing to help 3.11 1.20 0.88 4.52 0.42 0.70 2 1.41
(13) Never too busy to respond to your requests 2.87 0.56 0.83 3.67 0.67 0.67 2 0.80

Assurance a 5 0.835 a 5 0.877

(14) Em ployees can be trusted 3.12 0.81 0.65 3.39 0.56 0.72 2 0.27
(15) Feels safe in your interaction with em ployees 4.22 0.98 0.85 4.53 0.46 0.71 2 0.31
(16) Consistently courteous 3.45 1.02 0.95 4.42 0.74 0.72 2 0.97
(17) Has knowledge to answer your questions 3.30 0.68 0.87 4.71 0.97 0.67 2 1.41

Empathy a 5 0.874 a 5 0.728

(18) Gives individual attention 2.43 0.98 0.86 3.46 0.81 0.47 2 1.03
(19) Has operating hours convenient to you 3.00 1.04 0.74 3.64 0.55 0.48 2 0.64
(20) Em ployees give personal attention 2.34 1.45 0.67 3.87 0.78 0.47 2 1.53
(21) Has your best interests at heart 2.01 1.47 0.66 3.56 1.24 0.46 2 1.55
(22) Em ployees understand your speci® c needs 3.56 1.41 0.76 4.55 0.56 0.47 2 0.99
 Downloaded by [Universite De Paris 1] at 12:55 06 August 2013

Table 2. Exploratory factor analyses (N 5 82)

Loadings Loadings Loadings

Expectations Perceptions G ap
scale item s Factor 1 Factor 2 scale items Factor 1 Factor 2 scale item s Factor 1 Factor 2

E1 0.572 0.233 P1 0.645 0.124 G1 0.524 0.023

E2 0.591 0.258 P2 0.553 0.122 G2 0.524 0.023
E3 0.744 0.343 P3 0.734 0.124 G3 0.726 0.178
E4 0.789 0.322 P4 0.789 0.125 G4 0.692 0.013
E5 0.723 0.070 P5 0.766 0.201 G5 0.752 0.012
E6 0.678 0.106 P6 0.744 0.104 G6 0.724 0.110
E7 0.731 0.023 P7 0.749 0.023 G7 0.758 0.013
E8 0.832 0.120 P8 0.914 2 0.120 G8 0.824 0.189
E9 0.823 0.044 P9 0.731 0.122 G9 0.693 0.130
E10 0.763 0.210 P10 0.775 0.013 G 10 0.734 0.204
E11 0.732 0.124 P11 0.841 0.043 G 11 0.634 0.045
E12 0.728 0.165 P12 0.712 0.124 G 12 0.813 0.134
E13 0.628 0.120 P13 0.868 0.044 G 13 0.898 0.044
E14 0.886 0.122 P14 0.784 0.034 G 14 0.812 0.201
E15 0.826 0.240 P15 0.887 2 0.043 G 15 0.639 0.176
E16 0.789 0.127 P16 0.902 0.107 G 16 0.589 0.115
E17 0.698 0.034 P17 0.778 0.025 G 17 0.732 0.013
E18 0.845 0.012 P18 0.774 0.123 G 18 0.765 0.199
E19 0.823 0.055 P19 0.744 0.058 G 19 0.723 0.243
E20 0.789 0.273 P20 0.723 0.050 G 20 0.643 0.134
E21 0.791 0.115 P21 0.832 0.166 G 21 0.730 2 0.134
E22 0.901 0.161 P22 0.723 0.135 G 22 0.860 0.201

Eigenvalue 13.135 1.701 Eigenvalue 12.512 1.081 Eigenvalue 12.545 1.113

Variance explained 59.70% 11.34% Variance explained 56.87% 7.21% Variance explained 57.02% 7.42%
SER VQU AL
149
 150 S . S . K . LAM

R eliability

Item s for each subscale were subjected to reliability assessm ent. Results of the item analysis
for both the expectations and perform ance perceptions scores are presented in Table 1.
The coef® cient alpha values for th e expectations subscales range from 0.645 to 0.877.
N one of the item -to-total correlations for the individual expectation item s was less than th e
0.35 cut-off value (Nunnally, 1978). C oef® cient alpha values for the perform ance percep-
tions subscales ranged from 0.711 to 0.902. A gain, none of the item -to-total correlations for
the perception item s was less th an the 0.35 cut-off value. R eliability for linear com binations
of the ® ve subscales was also com puted to assess the overall internal consistency of th e
expectations and perceptions m easures. The overall coef® cient alpha values were 0.855 and
0.884 for the expectations and perform ance perceptions scores respectively. Th ese results are
consistent with those reported in Babakus and Boller (1991) and Parasuram an et al. (1988)
suggesting that both m easures exh ibit desirable levels of reliability and internal consistency.
Downloaded by [Universite De Paris 1] at 12:55 06 August 2013

D imensionality

To exam ine the dimensionality of the scale, an exploratory factor analysis was conducted on
each of the correlation m atrices of the expectations, perform ance perceptions and th e gap
scores (perform ance perceptions m inus exp ectations). Factor analysis results for expecta-
tions, perform ance perceptions and the gap scores are provided in Table 2.
Data on the three correlation m atrices produced very sim ilar results, with one factor
accounting for m ost of the variation in item scores. W hen subjected to oblique rotation, th e
loading did not provide a clear picture of any m eaningful factor structure. A fourth round of
factor analysis was also conducted using expectations and perform ance perceptions item s
together. The results identi® ed one dom inating factor representing expectations and percep-
tions. Factor loadings from the com bined solution did not differ in any signi® cant way from
those obtained with separate analyses of the expectations and perceptions.
These results suggest that the proposed ® ve dimensions of SERVQ U AL cannot be
con® rm ed, that both scales can be treated as unidimensional and that the 22 item s can be
considered as one com posite set of individual m easures.

Predictive validity

Individual item scores for expectations, perceptions and gap scores were added together to
obtain overall scores for each construct. These scores were then correlated with a single-item
behavioural intentions scale (return to the hospital for another service) and another single-
item m easure of overall quality rating. Both the behavioural intentions scale and the overall
quality rating item were on Likert scales with a ® ve-point response form at wh ere a higher

Table 3. Correlations of summated scores (N 5 82)

Items 1 2 3 4 5

(1) Perceptions ±
(2) Expectations 0.431 * ±
(3) G ap scores 0.532 * 0.133 ±
(4) Intentions 0.667 * 0.145 0.441 * ±
(5) Overall ratings 0.711 * 0.174 0.512 * 0.701 * ±

*Signi® cant beyond the 0.01 level.

 SER VQU AL 151

score indicated a m ore favourable response. The resulting correlation m atrix is presented in
Table 3.
A ll these correlation scores were statistically signi® cant in the predicted direction. This
result provides strong evidence of th e predictive validity of the scale. However, it also raises
questions on the role of th e expectations com ponent of the scale. The correlation between
behavioural intentions and perceptions was higher than th e gap scores (0.667 and 0.441).
The sam e was true for the correlation between overall ratings and perceptions and gap scores
(0.711 and 0.512). The correlation of exp ectations with behavioural intentions and overall
quality ratings was very low.
These ® ndings suggest that the expectations scores m ay not be contributing to th e
strength of the relationsh ip betw een service quality and the intention or overall quality rating
variables. Such results raise the question of the relevance of the exp ectations± perform ance
gap as the basis for m easuring service quality and provide support for the sim ple perform -
ance-based m easures of service quality in th e health care industry.
Downloaded by [Universite De Paris 1] at 12:55 06 August 2013

Individual service quality scores

Since the expectations, perform ance perceptions and gap scores have emerged as unidim en-
sional m easures without any m eaningful underlying dim ensions, it was decided that the 22
item s could be analyzed individually. The overall picture of health care quality in Hong Kong
can be sum m arized as follows:

(1) Patients’ exp ectations are highest for com petent and prom pt services. `Perform s th e
service right the ® rst tim e’ , `gives prom pt services’ and `has know ledge to answer
your questions’ received the highest expectation scores com pared to other item s.
This suggests that prom pt and com petent services are the m ost important factors
patients expect of hospitals. It is worthwhile to note that although the perform ance
scores of these item s are on average above 3.5, they still fall short of patient
expectations. Patients are still not satis® ed with services they expect m ost, especially
`has the know ledge to answer your questions’ and `gives prom pt services’ ; both with
a gap of m ore than 1 betw een the expectation and perception. These are the areas
hospitals can improve in order to achieve better service quality.
(2) The ® rst four item s, `has up-to-date equipm ent’ , `physical facilities are visually
appealing’ , `em ployees are neat in appearance’ and `m aterials are visually appealing’ ,
sh ow the lowest score am ong the 22 item s for expectation. These four factors also
sh ow positive gap scores. These ® ndings suggest that ph ysical elem ents are perceived
to be least important and the patients are generally satis® ed with this aspect of
service quality.
(3) The widest gap betw een patients’ perceived perform ance and expectations are th e
factors of `has your best interests at heart’ , `show s sincere interest to solve your
problem s’ and `em ployees give personal attention’ Ð all with a gap score higher th an
1.5. H ospitals in H ong Kong do not seem to be good at giving personal attention
and caring service to their patients. Although expectations of these factors were not
the highest am ong patients, these are the factors that hospitals should concentrate on
in order to improve their service, so that they not only satisfy the basic needs of their
patients but also provide a service which exceeds their expectation.
 152 S . S . K . LAM

D iscussion

In th e present study, based on data from outside the U S, SE RVQ U AL appears to be a

consistent and reliable scale to m easure health care service quality. The scale exhibits reliable
and valid m easures of patients’ exp ectations of health care services and their perceptions of
the health care providers’ perform ance level. However, the expectations, perform ance per-
ceptions and gap scores have emerged as unidimensional m easures without any m eaningful
underlying dimensions. These results m ay be in part due to a different environm ent in a
different country where the study was undertaken, or because the original dimensions m ay
not be applicable to the health care sector. It m ay be that th e scale item s that de® ne service
quality in one country m ay be different in another, or the scale item s de® ned in one particular
industry m ay be different from those of another industry. These results indicate the need for
further work on th e dim ensionality and conceptual level of th e construct.
The results pinpoint areas for attention to improve health care service quality. First, they
Downloaded by [Universite De Paris 1] at 12:55 06 August 2013

indicate that perceived service perform ance generally falls short of idealized expectations
except in the physical elem ents of service quality. Timely, professional and com petent service
is wh at the custom ers expect from health care providers, and although hospitals in Hong
K ong are generally providing go od services in these three areas, improvem ents are still
needed to m eet patients’ expectations. The results also indicate patients’ perception that
hospital staff show not enough caring and that they are not provided with individualized
attention. These are the areas that hospitals should improve in order to build a patient-
focused attitude towards service delivery.

R eferences
B ABAKU S , E. & B OLLER , G.W . (1991 ) An empirical assessm ent of the SERVQU AL scale, Journal of Business
Research, 24, pp. 254± 263.
B OW ERS , M., S W AN , J.E . & K OEHLER , W .F. (1994 ) W hat attributes determ ine quality and satisfaction with
health care delivery?, Health Care M anager Review , 19, pp. 49± 55.
C ARM AN , J.M . (1990 ) Consumer perceptions of service quality: an assessm ent of the SERVQU AL dimen-
sions, Journal of Retailing, 66, pp. 33± 44.
D O NAB EDIAN , A. (1980) Explorations in Quality Assessment and Monitoring. Vol. 1, The De® nition of Quality and
Approaches to Its Assessment (N H ealth Administration Press).
F INN , D.W . & L AM B , C.W . (1991 ) An evaluation of the SERVQUAL scale in a retail setting. In: R.H .
H OLM AN & M . S OLOM ON (E ds) Advances in Consumer R esearch, 18, pp. 312 ± 322.
F RIEDM AN , H.S. (1979 ) Non-verbal com m unications between patients and m edical practitioners, Journal of
Social Issues, 35, pp. 298± 304.
G RONROO S , C.A. (1984 ) Service quality m odel and its m arketing implications, European Journal of M arketing,
18, pp. 36± 40.
N U NNALLY , C.J. (1978 ) Psychometric Theory (N ew York, M cGraw-Hill).
P ARASU RAM AN , A., Z EITHAM L , V.A. & B ERRY , L.L. (1988 ) SERVQUAL: a m ultiple-item scale for measuring
customer perceptions of service quality, Journal of R etailing, 64, pp. 12± 23.
W ARE , J.E . & S NYDER , M.K . (1975 ) Dimensions of patient attitudes regarding doctors and m edical services,
Medica l Care, 26, pp. 669± 673 .
Z EITHAM L , V.A. (1988 ) Consum er perceptions of price, quality, and value: a m eans-end m odel and synthesis
of evidence, Journal of Marketing, 52, pp. 2± 11.