ArticlePDF Available

Perceived Electronic Service Quality: Some Preliminary Results From a Cross-National Study in Mobile Internet Services

  • Alba Graduate Business School, The American College of Greece
  • Quryon Korea

Abstract and Figures

Work on how consumers evaluate electronic service quality is both topical and important due to the well accepted criticality of electronic channels in selling products and services. However, the extant research on electronic research quality is preoccupied with the web site internet context and most of the studies are single-country studies, inhibiting conclusions of robustness and generalizability. Theoretically rooted in the Nordic Model of perceived service quality the study uses an e-service quality scale to measure mobile internet service quality and most importantly it does so in different national settings. Consistent with the extant e-service quality literature, results indicate that e-service quality is a second-order factor, with three reflective first-order dimensions: efficiency, outcome, and customer care. Most importantly, cross-validation investigations using samples drawn from Korean, Hong-Kong and Japanese mobile internet user populations support the factorial structure invariance of the construct. Following Cheung and Reynolds' (2002) suggestions, we tentatively examine factor means differences between the three countries contributing to the scarce cross-national electronic service quality literature. Findings imply that though consumers in different countries use the same dimensions so as to evaluate mobile internet services, importance weightings assigned on these dimension are not the same.
Content may be subject to copyright.
Electronic copy available at: copy available at:
Work on how consumers evaluate electronic service quality is both topical and
important due to the well accepted criticality of electronic channels in selling products
and services. However, the extant research on electronic research quality is preoccupied
with the web site internet context and most of the studies are single-country studies,
inhibiting conclusions of robustness and generalizability. Theoretically rooted in the
Nordic Model of perceived service quality the study uses an e-service quality scale to
measure mobile internet service quality and most importantly it does so in different
national settings. Consistent with the extant e-service quality literature, results indicate
that e-service quality is a second-order factor, with three reflective first-order dimensions:
efficiency, outcome, and customer care. Most importantly, cross-validation investigations
using samples drawn from Korean, Hong-Kong and Japanese mobile internet user
populations support the factorial structure invariance of the construct. Following Cheung
and Reynolds’ (2002) suggestions, we tentatively examine factor means differences
between the three countries contributing to the scarce cross-national electronic service
quality literature. Findings imply that though consumers in different countries use the
same dimensions so as to evaluate mobile internet services, importance weightings
assigned on these dimension are not the same.
Perceived Electronic Service Quality, Mobile Internet, Cross-National, Multi-
Sample Analysis
Electronic copy available at: copy available at:
Perceived electronic service quality constitutes a well-established construct in the
e-commerce literature (e.g., Liao, Palvia and Lin 2006). According to Zeithaml et al.
(2002, p.135) electronic service quality can be defined “as the extent to which a web site
facilitates efficient and effective shopping, purchasing and delivery”.
Research on e-service quality has just started gaining a momentum and the main
research question that all relevant studies try to address pertain to the factorial structure
of the construct and measurement issues (e.g. Wolfinbarger and Gilly 2003). However,
though research on measurement issues is quite advanced, cross-national considerations
of the electronic service quality construct are scarce in the literature. Wolfinbarger and
Gilly (2003) explicitly recognize this research gap and call for more research in the
investigation of electronic service quality vis-à-vis international populations.
Additionally, discussions in the service quality literature questioning the
generalizability of service quality dimensions across different countries (Tsikriktsis
2002), renders the investigation of the stability of the e-service quality dimensionality in
different countries as topical and important.
The present research contributes to the extant e-service quality in two important
ways: First, addressing the call of Wolfinbarger and Gilly (2003) investigates e-service
quality in a cross-national context. Does the factorial structure of the construct is the
same across different nationalities? If this is the case, do consumers in different countries
assign the same rankings of importance on different e-service quality dimensions?
Second so as to further enhance the external validity of the e-service quality literature the
study investigates the above mentioned research questions in the context of an alternative
electronic channel of services provision, namely the wireless mobile phone internet
channel. An online survey serves as the empirical vehicle of the study, while exploratory
and confirmatory factor analyses (CFA) are used to tackle the research questions under
Theoretically, the present study builds from the Nordic Model of traditional
services perceived service quality so as to investigate the electronic service quality
construct (Brady and Cronin 2001). The Nordic model conceptualizing perceived service
quality in traditional people-oriented services differentiates between the “what” (i.e. what
the consumer receives as a result of his interaction with a service firm/technical quality)
and “how” (i.e. how he/she gets the outcome resulting from his/her interaction with the
seller/functional quality) components of the buyer-seller interaction. For example the
“what” component of the Nordic Model, is addressed using the notions of aesthetics (i.e.,
how enjoyable and visually attractive is to use the service), whereas the “how”
component is addressed using the notions of ease of use, and customer service among
We do not formally hypothesize a priori propositions relating service quality
dimensions with the countries under investigation. However, building from the scarce
cross-national e-service quality literature (e.g. Tsikriktsis 2002), suggesting culture as
influencing e-service quality dimensions, we hypothesize the existence of cross-national
differences in service quality perceptions. This is consistent with the extant traditional
service quality literature (e.g. Furrer, Liu and Sudharshan 2000). Arguably, the
theoretical foundation for proposing cross-national electronic service quality is not strong
enough. Therefore, we view the potential for cross-national differences as a tentative
position that may be explored with the data at hand. If the results are promising,
researchers may be encouraged to theorize these cross-national differences rigorously.
Besides contributing to the scarce cross-national e-service quality literature this
article is novel in that it uses the e-service quality construct so as to measure perceptions
of service quality in an alternative e-commerce channel namely the mobile phone internet
services channel.
Mobile Internet, defined as the wireless access to Internet content via mobile
devices, such as mobile phones and personal digital assistants, has advanced
astonishingly both in terms of user population and technology developed (Kim et al.
2002). Wireless Internet services via mobile phone devices became available in Japan,
Korea and Hong-Kong in 1998 (Kim et al 2004) and debuted in Europe in 2002, mainly
through NTT DoCoMo’s i-mode and Vodafone’s Live!, and are rapidly gaining end-user
acceptance throughout the world.
Currently, only few studies have investigated consumers’ reactions to mobile phone
internet services. These studies have not directly investigated these consumer reactions in
the realms of the e-service quality literature. Chae et al. (2002), employing an on-line
survey and structural equation analysis, found four second-order factors of information
quality for wireless Internet services: connection quality, content quality, interaction
quality and contextual quality. However, their research focuses on perceived information
quality rather than perceived service quality, which is a wider construct. Bruner and
Kumar (2005), employing the Technology Acceptance Model, proposed and tested an
extended consumer TAM for wireless Internet. The core model constructs are those of
usefulness, ease of use and fun. The authors have found that usefulness and fun (directly)
and ease of use (indirectly) influence attitudes toward adopting mobile commerce
On the other hand the literature pertaining to web site wireline internet service
quality is much more advanced. Zeithaml, Parasuraman and Malhotra (2002) suggests
that electronic service quality can be decomposed into four dimensions, namely
efficiency, fulfillment, privacy and technical reliability. Loiacono et al. (2007) propose
twelve dimensions and 3 higher-order constructs, that is ease of use, usefulness and
entertainment. Finally, Wolfinbarger and Gilly (2003) conceptualize electronic service
quality using four dimensions, namely fulfillment/reliability, website design,
privacy/security and customer service.
Admittedly, most research efforts made to measure consumers’ evaluations of
electronic service quality, conclude in giving us extended measurement scales that though
content valid, are practically difficult to use due to their excessive length (e.g. e-
SERVQUAL, consists of 22 items). This is especially the case in the context of the
present study namely handheld internet services, where limited input and output
resources of access devices inhibit the use of extended consumer evaluation instruments.
The present study proposes an abbreviated consumer evaluation instrument that is
practically useful, managerially relevant and psychometrically sound, theoretically
building from the well accepted dual-factor model of perceived service quality (Brady
and Cronin 2002). The study uses seven measures, so as to measure electronic service
quality in mobile internet services. These measures relate to Ease of Use, Usefulness,
Aesthetics, Content, Privacy, Customization and Customer Service (see Table 1).
Conceptually these measures, tap the theoretical dimensionality of perceived service
quality, namely the functional and technical service quality components proposed by the
Nordic Model (i.e. the “what” and “how” components) (Brady and Cronin 2001).
-Insert Table 1 about here-
The work presented in this paper is part of a wider research project conducted by a
worldwide research consortium, known as the Worldwide Mobile Internet Survey
(WMIS). This paper reports results from WMIS in Korea, Japan and Hong Kong.
First, we develop the service quality scale in the Korean sample employing a split-
sample analysis procedure. We explore the structure of the service quality construct
employing exploratory factor analysis (in the first split sample) and then validate this
structure in the second split sample using confirmatory factor analysis. So as to
investigate the generalizibility of the findings we then move on to investigate the stability
of the service quality dimensionality in two fresh samples collected in two additional
Asian countries, namely Hong-Kong and Japan. We selected these three Asian countries
for two reasons: due to high penetration rates of mobile phone devices in these countries,
(77.2% in Hong Kong, 79.3% in Japan, and 39.8% in Korea) and due to high penetration
rates for mobile internet services usage (e.g. Japanese and Korean markets accounted for
73% and 23% of the total mobile internet market in Asia, respectively (Kim et al 2004)).
Korean Sample
To collect data efficiently and increase the validity of the empirical data, we
constructed a research consortium consisting of every mobile operator in Korea, major
Internet portals and mobile Internet application developers. Μobile operators verified the
survey data and provided funding to our project.
Data were collected employing a large scale online survey and potential
respondents were given participation incentives. Two major concerns in Internet-based
surveys are the respondents filling out the survey multiple times and “random walk-ins”
(Deutskens et al., 2004; Ilieva et al., 2002). Mobile operators checked whether the phone
numbers self-reported were legitimately registered and whether the owners of the phone
numbers had accessed the mobile Internet at least once in the past.
A total of 15,516 people participated in the survey. Those who did not pass the test
were deleted from the data set. The number of the final effective respondents was 8,912.
This data-collection procedure increases the external validity of the results, since
participants belong in the actual customer base of large mobile operators. The sample is
almost split between men and women, though women are slightly more represented than
men. The age of respondents ranged from 12 to 80 years, with a median of 24 years. Most
of the respondents were in their early 20s (50.3%). In terms of gender, age, and
occupation distributions, our sample may be considered as representative of mobile
Internet users in Korea (Sir et al. 2003).
Hong-Kong Sample
So as to collect data from the Hong-Kong sample we employed an online survey
methodology-as was the case with the Korean sample. The questionnaire was
administered on a non-profit public website run by the Hong Kong government. An e-
mail soliciting participation in the survey was sent to registered members of the website.
Also, a banner advertisement of the survey was made available on the website over a
period of four weeks. To reduce the possibility that a respondent participated in the
survey more than once, each respondent was required to provide his/her mobile phone
number in the survey. To encourage participation, incentives of the latest models of
mobile phones and MP3 players were offered as lucky draw prizes. A total of 1826 valid
responses were collected from the current user group. In total, there were 8941
respondents who successfully completed the questionnaires; of which 7045 were
potential users and 1826 were current users. 817 respondents were males (44.7%) and
1009 were females (55.3%). The age of respondents ranged from 13 to 76 years, with a
median of 25 years. Most of the respondents were in their 20s (53.1%) and 30s (18.0%).
The length of experience with using mobile Internet ranged from 1 to 44 months, with 15
months as the median, and 17.2 months as the mean.
Japanese Sample
A research center administered the data collection process in Japan. Online panel
members were solicited via email requests. More specifically the questionnaire was
uploaded on the homepage of MIN1, and e-mails were sent to ECOM2 members-along
with requests to other relevant parties through ECOM- and to MIN monitors (e.g., i-mode
monitors), asking to access the questionnaire page.
1 Marketing Interactive Network-a marketing website that enables the creation of online panels consisting
of mobile internet (e.g. MIN i-mode monitors) and stationary internet users. MIN is an official research
partner of the Electronic Commerce Promotion Council (ECOM) of Japan.
2 Electronic Commerce Promotion Council of Japan
A total of 3,310 people participated at the survey. The effective number of
respondents was 2,151 (number of respondents self-reporting currently using mobile
internet services). The length of experience with using mobile Internet was 79 months
(median value), indicating Japanese respondents as quite experienced with mobile
internet services. The sample was somewhat balanced between men (47%) and women
(53%) and the age of respondents ranged from 14 to 73 years old with a median of 34
years. Three out of ten respondents reported being less than 30 years old. Therefore
compared to the Korean and Hong-Kong data sets, Japanese respondents reported
belonging in older age groups.
Internal Analyses
A random split-sample approach was employed (Babakus et al. 2004;
Diamantopoulos and Siguaw 2000). In the first half sample (N= 4,456) Exploratory
Common Factor Analysis was used as a first step in identifying the factor structure of the
electronic service quality construct. Then, in the second half of the sample alternative
first-and second-order factor models were tested through Confirmatory Factor Analysis
to examine which model of perceived electronic service quality is superior in fitting the
sample data.
Measurement Invariance
Measurement invariance determines if items used in survey-type instruments mean
the same things to members of different groups (Cheung and Rensvold 2002). We
conduct tests for both category 1 and category 2 invariance (Cheung and Rensvold 2002).
We conduct configural, metric, scalar and invariance tests so as to investigate the
generalizibility of the results found in the Korean sample, in two more samples and then
tentatively examine the existence of latent mean differences in these three countries
employing mean structure analysis procedures (MACS) (Arbuckle and Woethke 1999)
Issues in Assessing Measurement Invariance
An important issue in measurement invariance tests relates to the choice of criterion
to be used so as to assess measurement invariance. Two categories of criteria exist (Chen,
Sousa and West 2005): the first one utilizes purely statistical criteria namely the χ2
difference test-whereas the second criterion involves practical criteria namely fit indices.
Currently the most widely used criterion is the χ2 difference test. However the
likehood ratio test is sensitive to non-normality and has substantial power in large
samples to detect small discrepancies between groups that may be of no practical
importance (Chen, Sousa and West (2005).
Currently the best available guidelines for the usage of practical fit indices in
testing for measurement invariance are those proposed by Cheung and Rensvold (2002).
They concluded that a difference of larger than .01 in CFI would indicate a meaningful
change in model fit when testing for measurement invariance.
Power Analysis
Statistical power is the probability of rejecting a false model when it is false
(McQuitty 2004). If models do not have adequate power, then their contribution to
knowledge is uncertain . In general an accepted level of power is 0.80 (McQuitty 2004) .
So as calculate statistical power we used a stand-alone DOS program called NIESEM
written by Paul Dudgeon3. The NIESEM program is based on the work of
MacCallum,Browne and Sugawara (1996). Power is close to unity, indicating almost
zero probability of conducting type II error.
Exploratory Factor Analysis (Korean Sample)
Common Factor Analysis with varimax rotation was employed to determine the
factor structure of the perceived electronic service quality construct (Gorsuch 1990)..
However, Component Analysis with varimax rotation was also tested to confirm the
emerging factor structure.
We tested a two-, three-, and four-factor model. Based on the percentage of
variance criterion (Hair et al. 1998), the analysis revealed the three factor solution as
more appropriate, explaining 52.5 percent of the total variance (see Table 2). All
measures load clearly to the three factors extracted. Component analysis revealed the
same factor structure, explaining 73 percent of the total variation, while all seven factor
loadings were greater than .75.
-Insert Table 2 about here-
The first factor, explaining 23 percent of the total variance, constitutes the
Efficiency Quality dimension and is related to the ease of use and the usefulness of the
wireless Internet service. Zeithaml et al. (2003, p.365) suggest that efficiency constitutes
a dimension of electronic service quality, defining efficiency as “the ability of the
customers to get to the web site, find their desired product and information associated
with it and check out with minimal effort”. In the human computer interaction (HCI)
literature, efficiency quality refers to whether the consumer perceives that the task is
3 available at
performed without making mistakes or putting too much effort (Sing 2003). In both
definitions it is suggested that efficiency has to do with ease of use and usefulness.
The second factor, accounting for 16 percent of the total variance, is Outcome
Quality. This factor encompasses emotional benefits (i.e. enjoyment), visual
attractiveness (i.e., aesthetics) and content variety (i.e., functional benefits) provided by
the use of wireless Internet services. Outcome quality reflects the “product” of the service
act itself, or in other words what the customer receives in the service encounter (Brady
and Cronin 2002).
The third factor, Customer Care Quality, explains 13 percent of the total variance.
Customer care quality relates to adapting the wireless service to user preferences (level of
personalization), to minimization of personal data provided (privacy issues), and to the
customer service provided, especially when consumers experience a problem while using
wireless Internet services. Gounaris and Dimitriadis (2003) propose the dimension of
customer care in their work related to service quality in business-to-consumer Internet
portals. They found that this dimension encompasses issues like privacy of shared
information and customer service.
Based on these findings, the proposed research model is depicted in Figure 1.
Perceived wireless Internet electronic service quality is suggested to be a second-order
factor, with three first-order factors, namely efficiency quality, outcome quality, and
customer care quality. The proposed research model is strengthened by the growing
stream of evidence conceptualizing perceived service quality as a multidimensional
construct of hierarchical nature (Brady & Cronin 2001, Dabholkar, Thorpe and Rentz et
al. 1996).
-Insert Figure 1 about here-
Confirmatory Factor Analysis (Korean Sample)
Following the methodology employed by Doll, Xia and Torkzadeh (1994) and
Sommers et al. (2003) we tested the proposed second-order factor model against three
other possible alternative factor structures (see Figure 2). More specifically, we tested the
proposed research model against a one first-order factor model, a three-factor model with
orthogonal factors, and a three-factor model with correlated factors.
-Insert Figure 2 about here-
We used the Maximum Likelihood (ML) estimation method to estimate the
parameters of the models, since ML-based fit indices outperform those obtained from
other methods (Hu and Bentler 1998 ).Based on this guideline, the CFI, Delta 2, RMSEA,
chi-square and standardized RMR fit indices for all four alternative factor structures are
reported in Table 3( Schumacker and Lomax 2004, Hu and Bentler 1999)
Insert Table 3 about here
Based on the fit indices, the one first-order factor model is far from being
acceptable. The uncorrelated three-factor model is also not supported by the data
covariance matrix. Finally, the correlated three-factor model has the same fit indices with
the second-order factor model, indicating adequate fit. However, theory in the domain of
perceived service quality suggests the existence of a second-order factor that accounts for
the common variance of the first-order factors. This model seems to be theoretically more
interesting than the correlated three-factor model4 Doll, Xia and Torkzadeh (1994). With
the exception of the chi-square statistic, the proposed second-order factor structure fits
the data reasonably well (χ2 (11) = 223.91 and p=0.00, CFI=.97, Delta 2=.97, SRMR=.03
and RMSEA=.066).
The internal structure of the proposed model was also examined. First, we
examined the parameter estimates and the accompanying tests of significance (Bagozzi
and Yi 1988). Convergent validity is implied by the magnitude of the factor loading of
each measure on its suggested latent variable (Mathwick et. al 2001; Dabholkar, Thorpe
and Rentz 1996). All λ’s are greater than the .60 level proposed by Bagozzi and Yi
(1988), except λ42 (aesthetics measure) which is marginally below .60. Moreover, all λ’s
are significant since all t-values are above the |2.00| level suggesting convergent validity.
The results are summarized in Figure 3.
-Insert Figure 3 about here-
Discriminant validity can be demonstrated by calculating covariance confidence
intervals (plus or minus two standard deviations) around the factor covariances. In our
case all confidence intervals computed do not include the value of 1.00 suggesting
discriminant validity (Mathwick et. al 2001; Dabholkar,Thorpe and Rentz1996).
Regarding measurement model fit, we used the Composite Reliability (ρc) and
Average Variance Extracted (AVE) criteria. Bagozzi and Yi (1988) propose that ρc and
AVE should be greater than 0.6 and 0.5 respectively. These cut off points hold for the
4 Further support for the superiority of a second-order factor model can be found in the structural equation
modeling literature. Chen, Sousa and West (2005) point to the next set of advantages: (a) a second-order
model can test whether a hypothesized second-order factor can actually account for the pattern of relations
between the first-order factors, (b) puts a structure on the pattern of covariance between the first-order
factors, (c) separates variance due to specific factors (these specific factors are represented by the
disturbance of each first-order factor), leading to a theoretically error-free estimation of the specific factors
and (d) can provide useful simplification of complex multitrait-multimethod models
three constructs, except from the Outcome Quality construct with an AVE of .44. The
composite reliability of Outcome Quality is .61 exceeding the suggested cut-off point (see
Table 4)
-Insert Table 4 about here-
Confirmatory Factor Analysis (Japanese and Hong-Kong Samples )
Japanese Sample
CFA results for the Japanese sample indicate acceptable fit indices values with the
exception of the RMSEA index (see Table 4). More specifically χ2 (11)= 284, 41 (p<.00,
CFI and Delta 2 equal .945, marginally less than the cut-off criterion of .95 suggested by
Hu and Bentler (1999)(see Table 5). SMRS is .045 less than the established .05 level.
RMSEA equals .11 indicating poor fit5.
Altogether these results suggest that the measures used are unidimensional. All λ’s
are greater than the .60 level (Bagozzi and Yi 1988). Moreover, all λ’s are significant
since all t-values are greater than the |2.00| level indicating convergent validity (see Table
6). Additionally covariance confidence intervals computed do not include the value of
1.00 indicating discriminant validity6 ( Schumacker and Lomax 2004) . Composite
Reliability, and AVE are greater than 0.6 and 0.5 respectively with the marginal
5 Regarding the high RMSEA value, we follow Parasuraman, Zeithaml and Malhotra (2005) and the
structural equation modeling literature and point that the interpretation of any fit index in isolation could be
problematic because trade-offs between Type I and Type II errors call for the interpretation of
combinations of indexes in various model contexts. Another related issue is statistical power which have to
be taken into account when interpreting fit indices. In studies where power is overly great (i.e., > 0.9-as is
the case with the present study may require a more relaxed interpretation of fit than is typical. Conversely,
a more stringent interpretation of fit statistics is required when power is low (McQuitty 2004). The high
statistical power of the present study and the acceptable values for the CFI and SRMR indices, seem to
mitigate the somewhat high root mean square error of approximation (RMSEA) values (Parasuraman,
Zeithaml and Malhotra 2005)
6 We also checked for discriminant validity employing the most stringent criterion of Fornell and Larcker
(1981) namely we tested whether AVE from each latent is greater than its shared variance with the other
two latents (γ2).Results indicate that for each pair of latents AVE> γ2 , though the AVE of “customer care”
quality is marginally greater than squared correlation of “customer care” quality and “efficiency”.
exception of the “customer care” factor (AVE=.49) (see Table 6). Figure 4 depicts the
standardized, second-order mobile internet perceived service quality path diagram for the
Japanese sample.
-Insert Figure 4 about here-
Hong-Kong Sample
Confirmatory factor analysis results for the Hong-Kong sample indicate a
unidimensional and reliable measurement model (see Table 4). Fit indices are in the
established acceptable ranges indicating adequate fit between the model and the data at
hand (χ2 (11) = 148, 21 (p<.00), CFI=.97, Delta 2=.97, SRMR=.039) (see Table 5), with
the possible exception of the RMSEA index, which is marginally greater than the .08
criterion indicating reasonable fit Schumacker and Lomax (2004). Pertaining to the
somewhat high RMSEA value, the acceptable values of CFI and SRMR in association
with the high statistical power of the present study seems to mitigate this problem
(Zeithaml, Parasuraman and Malhotra 2005). Altogether these results indicate acceptable
The composite reliability index for the three first-order factors is greater than .70
surpassing the .60 criterion suggested by Bagozzi and Yi (1988) and AVE is greater than
the established .50 cut-off value. Additionally correlation confidence intervals (plus or
minus two standard deviations) computed for the three first-order factors do not include
the value of 1.0 indicating discriminant validity. Altogether these results are indicative of
reliability, convergent and discriminant validity (see Table 6). Figure 5 depicts the
second-order mobile internet service quality construct for the Hong-Kong sample.
-Insert Figure 5 about here-
Measurement Invariance
So as to test for the measurement invariance of the second-order service quality
model across the three countries we follow the general procedures suggested by Byrne
(2004) and Chen, Sousa and West (2005), essentially testing for a series of increasingly
constrained hierarchical nested models.
Configural Invariance (Model 1)
Testing for this form of invariance requires the specification of an unrestricted
baseline model. The simultaneously estimated model provides the baseline value against
which all subsequently specified (increasingly constrained models) are compared. This
multi-sample analysis yields only one set of fit statistics (Byrne 2004). Altogether fit
indices for this unconstrained model indicate acceptable fit (χ2 (33) = 650,390 (p<.00),
CFI=.97, Delta 2=.97, RMSEA=.047, SRMS=.028). These results support the validity of
the hypothesized three-factor service quality model across Korea, Japan and Hong-Kong.
Invariance of first-order loadings (Model 2)
In testing for this level of factorial invariance, all first-order loadings were
constrained to be equal across the three countries. This model is nested within the fully
unconstrained model (model 1). The chi-square difference test is significant (∆χ2 (df=8)
=56,320 (p=.00)), indicating non-invariance of the first-order factor loadings across the
three countries. However, given that the test was based on a large sample size, and due to
no substantial difference in CFI (CFI=.003, .966 vs. .963) we concluded that there was
no appreciable difference between the unconstrained model and the first-order
measurement weights constrained model (Chen, Sousa and West 2005; Cheung and
Rensvold 2002).
Invariance of second-order factor loadings (Model 3)
Testing for this level of invariance necessitates that all first-order and second-order
factor loadings to be constrained to be equal across the three groups. This model is nested
within model 2. The chi-square difference test is significant (∆χ2 (df=4) =18,622
(p=.00)), indicating non-invariance of the second-order factor loadings across the three
countries. However, the difference in CFI was not substantial (CFI=.001, .963 vs.
.962), therefore we concluded that there was no appreciable difference between model 2
and 3.
Invariance of intercepts of observed variables (Model 4)
In model 4, in addition to the constraints imposed on first-order and second-order
factor loadings in model 3 the intercepts of the observed variables were constrained to be
equal across the three countries. This is a prerequisite for comparing latent means across
groups Chen, Sousa and West (2005).The fit of this model is not good (∆χ2 (df=14)
=2626,280 (p=.00), CFI=.146). Following Steenkmamp and Baumgartner (1998) and
Ueltchy et al. (2004) we examined for partial scalar/intercept invariance. We do so
examining each pair of the three participating countries. As Steenkamp and Baumgartner
(1998, p. 81) point: “…at least one item besides the marker item has to have…invariant
intercepts in order for cross-national comparisons of factor means to be meaningful”.
Japan- Hong-Kong
The specific strategy employed so as to test for partial scalar invariance is the one
suggested in Byrne (2004, p. 285). Tests for scalar invariance pertaining to these two
countries indicate partial scalar invariance based on Steenkamp and Baumgartner (1998).
More specifically we find partial scalar invariance for the “customer care” and “outcome
quality” factors (for both factors only one intercept of an observed variable-customization
and content correspondingly-is found to be invariant). The comparison of the model
having constraints for the first-order and second-order factor loadings with the model
further imposing invariance constraints on two intercepts indicates a .01 difference in CFI
(∆χ2 (df=2) =94,183 (p=.00), the level Cheung and Renvold (2002) suggested as
indicative of practical invariance. These results indicate that we are entitled to test for
difference in factors means between these two countries but only for the customer care
and outcome quality constructs, since we do not find partial scalar invariance for the
efficiency factor.
Japan- Korea
Assuming model 3 to be correct (invariance of first-order and second-order factor
loadings) the model with contstraints on the full list of measured variables invariance do
not fit the data (∆χ2 (df=7) =2117,86 (p=.00), CFI=.161). Once again we start
investigating for partial invariance following Byrne (2004). Tests for the “customer care”
subscale (all three intercepts are posed to invariant) are not good (∆χ2 (df=3) =1083,380
(p=.00), CFI=.082). Following these results we started relaxing intercept invariance
constraints. Results are not good for the “customer service” observed variable (∆χ2
(df=2) =624,598(p=.00), CFI=.047), though the results for CFI are less than the
benchmark (<.05) suggested by Little (1997), but high above the suggestions of Cheung
and Rensvold (2002). Increasingly relaxing constraints do not improve the CFI criterion
to be less than the Cheung and Rensvold (2002) cut-off. Therefore results suggest that we
are not entitled to conduct a means difference test for the customer care latent in this pair
of countries, following Cheung and Rensvold (2002) but we can do so if we rely on Little
(1997). Results for the outcome quality latent, indicate partial scalar invariance.
Constraining both observed variables indicate intercept invariance following Little (1997)
since CFI=.024<.05. Relaxing one of the observed variables indicates a CFI less than
.01. Results for the efficiency factor suggest a CFI equal to .013 marginally greater than
the Cheung and Rensvold (2002) criterion.
Hong-Kong- Korea
Results for fully constraining intercepts of measured variables for this pair of
countries indicate non-invariance (∆χ2 (df=7) =265,335 (p=.00), CFI=.02). However
CFI equals.02, which is greater than the Cheung and Rensvold (2002) criterion
(CFI<.01) but much less than Little’s (1997) suggestion (<.05).Results for subscales
indicate partial scale invariance for the customer care factor ((∆χ2 (df=2) =77,347
(p=.00), CFI=.006) and the outcome quality (∆χ2 (df=4) =98,923 (p=.00),
CFI=.008). Results for the efficiency factor are marginal (∆χ2 (df=5) =139,401
(p=.00), CFI=.011)
In summary, results for the invariance of measured variables intercepts (along with
results pertaining to first-order factor loadings), indicate that we are entitled to compare
all factor means in the Hong-Kong –Korea pair, outcome quality and customer care
means for the Japan- Hong-Kong pair and the “outcome quality” means for the Japan-
Korea pair. Whatsoever in the latter pair, due to the marginality of results we will
tentatively report the means differences tests for the efficiency factor too.
Means Structure Analyses
Means structure analyses are required so as to investigate latent mean differences
between groups. So as to obtain estimates of the differences between the first-order
factors in the three groups, in each pair of countries one was chosen as a reference or
baseline group and its first-order factor means was set to zero. Then the latent means of
the other group was estimated; this value is the difference between the factor means of
the two groups/countries. The significance test (z value) indicates whether there is a
statistically significant difference in the latent means of the two countries analyzed.
So as to directly compare first-order factor means between pairs of countries we
specified a correlated first-order factor model of perceived service quality. As Chen,
Sousa and West (2005, p. 485) note: “the first-order factors means are conditional on the
higher-order factor mean (s) in a hierarchical model, and thus cannot be directly
compared.” We considered the possibility of second-order mean comparisons, but such a
test was inappropriate due to non-invariant second-order intercepts. Results are reported
in Table 5.
Hong-Kong- Korea
Invariance of first-order factor loadings and intercepts of measured variables was
imposed on the Korean and Hong-Kong samples. The Korean sample was chosen as the
baseline group and its latent mean was set to zero. There was a significant mean
difference between the two countries in all three factors. More specifically results
indicate that Hong-Kong scores lower on the importance of efficiency (-.24, z=-8,43)
p<=.00), outcome quality (-.32, z=-9,90, p<=.00), and customer care (-.23, z=-6,30,
Japan- Hong-Kong
Based on the invariance test conducted, factor means difference tests for this pair of
countries was conducted for the customer care and outcome quality constructs. There was
a significant mean difference between the two countries only in the “customer care”
factor. More specifically results indicate that Hong-Kong scores higher on the importance
of customer care (.417 z=9,74, p<=.00). There was no difference on the outcome quality
factor (.02, z=,50, p<=.62).
Japan- Korea
Based on the invariance test conducted, factor means difference tests for this pair of
countries was conducted for the “outcome quality” latent. More specifically results
indicate that Koreans-compared to Japanese-believe the outcome quality factor as being
more important when experiencing mobile services (.12 z=,40, p<=.out 00). Due to the
marginal results obtained when investigating intercept invariance for the efficiency
factor, we tentatively report the means difference results for this factor too. It seems that
Koreans assign significantly less important than Japanese in the efficiency factor (.13 z=-
6,58, p<=.00)
-Insert Table 5 about here-
This study uses the established e-service quality literature so as to measure
perceived service quality in the mobile phone internet services context. Next, following
discussions in the service quality literature questioning the generalizibility of service
quality dimensions across different countries and that technology acceptance may differ
across countries (Straub, Keil and Brenner 1997), it iinvestigates the stability of the
proposed dimensionality in two new samples drawn from different countries.
Results imply that the same dimensionality holds for the three countries
investigated. Configural invariance results imply that participants belonging in the three
countries investigated conceptualize the construct of service quality in the same way.
Finally an important contribution of this research effort pertains to the differing relative
importance that different countries assign on service quality dimensions in the context of
mobile internet services.
Though we did not formally hypothesize, a priori specific propositions relating
service quality dimensions with the countries under investigation, our results confirm
findings in the scarce cross-national e-service quality literature (Tsikriktsis 2002),
suggesting culture as influencing e-service quality dimensions. The results indicate that
companies should take into account these different importance weightings when
allocating resources for improving service quality in different countries.
Even though the services literature suggests information-based services as easier to
standardize across nations7, our results indicate that this may not be the case. The reader
should take into account though, that we did not directly account for the influence of
cultural dimensions on perceived mobile e-service quality dimensions. In the context of
this study, countries are considered as cultural characteristics proxies. This logic, is
strengthened from the worqk of Hofstede (1980), and evidence purporting the three
countries sampled in this study as scoring differently in three dimensions of national
7 Compared to people-processing services and possession-processing services (Furrer, Liu and Sudharshan
culture (1980), namely masculinity, individualism and uncertainty avoidance (see Kim et
al. 2004) for a discussion on these specific scores).
Though we expect the relative importance of mobile e-service quality dimensions
to be different across the three Asian countries investigated we consider our results as
tentative on this matter and call for more research involving strong a priori hypothesis
linking specific dimensions as more or less influenced by differing cultural
Whatsoever, we believe that a post-hoc effort to explain differences found on
relative importance assigned on different mobile e-service quality dimensions is
worthwhile. This strategy has precedence in the literature (e.g. see Straub, Keil and
Brenner 1997). Therefore using post-hoc explanations, we build our discussion on the
work of Kim et al. (2004) who tried to explain differences in the usage of mobile internet
services in the Japan, Hong-Kong and Korea based on cultural dimensions (masculinity,
uncertainty avoidance and individualism) and economic factors (gross national income,
internet penetration rates, broadband internet penetration rates).
In this study, so as to provide explanations for means differences found, we make
use of the uncertainty avoidance index and of the reported differences in economic
factors characterizing the three countries (Kim et al. 2004)
We start the discussion with the findings indicating Koreans as assigning more
importance in all three factors when compared with Hong-Kong respondents. The finding
that Koreans assign more importance on ease of use and usefulness issues might be
explained by differences found in economic factors (lower income and maturity of the
broadband stationary internet) (Kim et al. 2004). More specifically, this can be explained
by the high penetration of broadband internet in Korea, compared to Hong-Kong, and
more specifically on the notion of relative advantage (Kim et al. 2004). Mobile internet
via handheld devices was less readily adopted in Korea compared to Hong-Kong (and
Japan) due to the relative advantage of the stationary internet (i.e. much richer
information environment at a less cost). Therefore, one can hypothesize that Koreans
would like to have mobile internet services that are more easy to use based on the
following reasoning: difficult to use services might increase the cost of using such
services (at least in the case of a time-based revenue business model) and cost is a much
more important factor for Koreans, due to lower gross-national income and the cheaper
stationary internet.
Most importantly Koreans, compared to Hong-Kong residents, score higher in the
uncertainty avoidance cultural dimension. This entails ease of use as more important
since, easy to use services reduces the possibility of service failure and therefore
underscores higher confidence levels for the service used.
In the same vein, Koreans assign more importance on usefulness (i.e. a service that
satisfies given task), due to cost reasons (i.e. they are not that willing to pay for mobile
internet in the case it does not provide useful content, since they can satisfy their needs
cheaper using stationary internet). Generally speaking one can explain greater importance
assigned on all three factors of service quality from Koreans on the fact that they have
greater service quality expectations due their prior experience with high-speed mature
stationary internet services (e.g., they seem to assign more importance on the outcome
quality -namely the content depth and width-along with aesthetic appeal- when compared
to Hong-Kong respondents, something that might be due to their prior experience with a
much richer internet environment both in terms of content variety and visual elements).
Finally a possible explanation supporting the greater importance Koreans assign on
customer care may be found on the higher uncertainty avoidance scores of Koreans (Kim
et al. 2004). Uncertainty avoidance is the extent to which, the member of a culture feel
threatened by uncertain or unknown situations (Hofstede 1991). Therefore in these
cultures uncertainty associated with a possible service failure has to be reduced by the
guarantee of a quick solution to the problem” (Furrer, Liu and Sudharshan 2000, p. 360.
Therefore the existence of a customer service department, though admittedly important in
both countries, might be more important for cultures exhibiting higher levels of
uncertainty avoidance. The same reasoning might be employed for the privacy observed
variable. One could expect cultures with high uncertainty avoidance, to exhibit higher
wariness levels when it comes to privacy concerns.
Finally, greater relative importance imposed on customization might be also
explained by economic factors. Providing customization mechanisms in a mobile internet
services context is important , since it allows for a more efficient way of fulfilling desired
tasks and therefore requires less expenses (i.e. in terms of money paid for navigating the
service-time based revenue business model-and in terms of system resources, i.e., battery
We now move on to discuss the significant differences found in the Japan-Korea
pair of countries. Mean structure analysis indicated Japanese as assigning more
importance than Koreans on the efficiency factor, whereas it seems that Koreans assign
more importance on the core-product factor. Pertaining to the core-product factor and
continuing the line of reasoning explicated previously, one possible explanation for such
a state is the extensive prior experience of Koreans (when compared with Japanese) with
stationary broadband internet services. Fast stationary internet connections enable the
provision of content services that are wider both in terms of width and breadth.
Additionally Koreans seem to more favorably rate outcome quality due to their
current mobile services usage pattern. As Kim et al. (2004) point, Koreans (and Hong-
Kong residents as well) seem to prefer using mobile services that are more of a hedonistic
rather than a utilitarian character. For this kind of services it seems reasonable to say that
content depth and width as well as visual elements (aesthetics) are more important. On
the other hand Japanese, seem to more frequently use utilitarian mobile services (e.g., e-
mail, buying train tickets).To put it more bluntly, mobile services preferred by Koreans
(i.e., hedonistic services, for example downloading music content), are primarily
evaluated with criteria like content depth and width and visual/presentation elements,
therefore having a prominent status in Koreans importance weighting schemes.
On the other hand services preferred by Japanese, namely utilitarian services (e.g.
reading news, stock exchange information sending e-mails and booking train tickets), are
primary evaluated with criteria pertaining to the reliability and accuracy of the
information and not that much by presentation issues (Chae et al. 2002).
Pertaining to the greater importance assigned on ease of use and usefulness (i.e., the
efficiency factor) by Japanese when compared with Koreans, a logical assumption is that
such a state holds due to the higher-levels of uncertainty avoidance characterizing
Japanese. Hofstede (1980) argued that uncertainty avoidance relates to a general feeling
of anxiety when confronted with problems or challenges. Easy to use mobile services
reduce the possibility of confronting problems therefore reducing anxiety levels and
enhancing cognitions of confidence.
Further theoretical support for the relationship between uncertainty avoidance and
ease of use can be found using transaction cost theory (Devarai, Fan and Kohli 2002).
Uncertainty seems to constitute a form of transaction cost and ease of use is posited as a
mechanism for reducing such a transaction cost. Recently, Hwang (2004) found a
positive relationship between uncertainty avoidance and ease of use.
Pertaining to usefulness, it seems that Japanese may view it as more important due
to higher masculinity levels. In masculine-like societies, performing is highly valued and
useful services (i.e., services that enhance one’s performance (Davis 1989)) seem to be a
mechanism for attaining higher performance in everyday life activities.
Finally, pertaining to the Hong-Kong- Japan pair, our results imply that Hong-Kong
nationals assign more importance on customer care than Japanese. Based on the higher
uncertainty avoidance scores of Japanese when compared to Hong-Kong nationals, one
would expect a different sign in this difference. However, a potential explanation for such
a difference might be also found on the higher expectations that Hong-Kong nationals
might have for electronic service quality due to their extensive prior experience with
likewise stationary internet services. Additionally according to Kim et al. (2004), Hong-
Kong nationals seem to primarily use mobile internet services for commercial rather than
communication exchanges. It is expected therefore that due to the potential of economic
loss in their transactions, facets of service quality like privacy and customer service are
prompted as more important.
To sum up with it seems that mobile service providers with an active presence in
these three countries should not be guided by simplistic rules when investing resources
for improving service quality. Though, all service quality factors are important so as for
consumers to infer high service quality assumptions, the relative importance of these
factors is differential and managers should try to localize their resource allocation
strategies in the quest for high service quality ratings. Standardizing service quality
investment programs might be tempting due to cost advantages but this strategy may not
be on the right track.
This study is not without limitations. However these limitations present
opportunities for future research. First, the reader should take into account that our results
pertaining to mean structure analysis, and more specifically on measurement invariance,
heavily depend upon the criterion used so as to infer measurement invariance. Chen,
Sousa and West (2005) point, that currently the methodological literature is armed with
two measurement invariance criteria, namely the likelihood ratio criterion and the CFI
criterion. The former should be considered as too conservative whereas the latter should
be considered as a liberal test of measurement equivalence. This research study follows
Cheung and Rensvold (2002), who find CFI as the best performing index8 for
investigating measurement invariance.
Whatsoever, test statistics and fit indices should no replace sound judgment and
substantive expertise (Bollen 1993). Prior research on the relationship between culture
and e-service quality (e.g., Tsikriktsis 2002) along with the different patterns of using
mobile services in the three countries examined (Kim et al. 2004; Lee at al. 2002),
8 in terms of not being overly sensitive to small errors of approximation
increase our confidence of the results found employing the CFI criterion: relative
weighting schemes pertaining to mobile services quality dimensions are different across
Another important limitation involves employing an online survey design which
introduces self-selection bias problems (Chae et al. 2002). Self-selection bias might
create problems of sample representativeness. However, we are confident that due to the
screening procedure employed (e.g., in Korea almost half of the primary respondents
were deleted from the data set), our sample is consisted of real mobile internet users, and
based on their self-reported demographics they are representative of the mobile internet
user population in Korea (Sir et al. 2003). The problem of self-selection bias was
somewhat mitigated in the Hong-Kong and Japanese samples where along with banner
advertisements of the survey, e-mails were sent to registered users of specific web sites
that agreed to participate in the survey.
Once again, a convenience sampling methodology entails representativeness bias
concerns, but this was due to budget constraints and due to the complexity and cost of
simultaneously managing data collection in three countries. However, one should take
into account, the exploratory nature of the present study, since it is probably one of the
first rigorously investigating electronic perceived service quality differences in more than
one national markets.
Another important issue that merits discussion, due to its potential threat to our
study’s validity is concerns about the content validity of instrument used. We did not
employ measures pertaining to the technical reliability of mobile internet services (e.g.
times a mobile internet site crashes e.t.c). The dimension of technical reliability, relates to
QoS issues (i.e., network performance), and admittedly the measurement of such issues is
much more objective than the measurement of other potential service quality dimensions
(e.g., perceptions of usefulness). To put it differently, in this research study we consider
technical reliability as a given, as a pre-condition for a good mobile internet service. This
has precedence in the stationary internet service quality literature (Loiacono, Goodhue
and Watson 2007).
Likewise, an important research question that requires investigation is the role of
device quality perceptions in overall perceived service quality scores. To put it differently
is/or should device quality be a part of a perceived service quality in a mobile internet
context? Device quality manifestations may pertain to technical features but also to visual
elements (feel and look of the device), and one could argue that all these influence or
enable the provision of high service quality in a mobile internet context.
This study confirms the complexity of managing service quality perceptions and
does so by providing theoretical and empirical evidence for a) the multidimensionality
and the hierarchical nature of the construct and b) for the existence of significant
differences between countries in the relative importance assigned on certain perceived
service quality dimensions.
We believe that constructs pertaining to electronic service evaluation (e.g. service
quality, satisfaction, value)- either that be a wireline internet service or a mobile internet
service- are not that different at least in terms of factorial structure. We suggest that the
electronic service context (wireline internet and mobile internet contexts) should be better
treated as a moderator variable that weakens, strengthens or makes insignificant
relationships pertaining to structural relationships (i.e. relationships between evaluation
constructs). For example one could hypothesize that ease of use-though salient both in a
wireline internet and a mobile internet context-is perceived as more important in
explaining an outcome variable (i.e. intention to use) in a mobile internet context due to
the well known limitations of handheld devices and due to situational characteristics of
mobile internet services consumption (i.e., consumption of mobile internet services
compared to stationary internet services seems to be an “on the run” activity”).
J. Arbuckle, W. Wothke, AMOS 4.0 User’s Guide. Chicago, IL: Smallwaters Corp.:1999
E. Babakus, C. Bienstock, J. Scotter, Linking Perceived Quality and Customer
Satisfaction to Store Traffic and Revenue Growth, Decision Sciences 35(4),
2004, pp. 713-737
R. Bagozzi, Y. Yi, On the Evaluation of Structural Equation Models, Journal of the
Academy of Marketing Science 16(1), 1988, pp.74-94
H. Baumgartner, C. Homburg, Applications of structural equation modeling in marketing
and consumer research: A review, International Journal of Research in
Marketing 13 (2), 1996, pp. 139-161
K. Bollen, J. Long, Testing structural equation models. Newbury Park, CA:1993 Sage.
M. Brady, J. Cronin, Some New Thoughts on Conceptualizing Perceived Service Quality:
A Hierarchical Approach, Journal of Marketing 65 (July), 2001, pp. 34-39
M. Browne, R. Cudeck, Alternative ways of assessing model fit. Sociological Methods
and Research, 21 (), (1992), pp. 230-258.
G. Bruner, A. Kumar, Explaining consumer acceptance of handheld Internet devices,
Journal of Business Research, 58 (5), 2005, pp. 553-558
B. Byrne, Testing for multigroup invariance using AMOS Graphics: A road less traveled.
Structural Equation Modeling, 11(2), 2004, pp. 272-300.
M. Chae, J. Kim, H. Kim, H. Ryu, Information Quality for Wireless Internet Services,
Electronic Markets,12(1), 2002, pp. 38-46
M. Chae, J. Kim, What’s So Different about the Wireless Internet, Communications of
the ACM 46(12), 2003, pp.240-247
F. Chen, K. Sousa, S. West, Testing Measurement Invariance of Second-Order Factor
Models, Structural Equation Modeling, 12 (3), 2005, pp. 471-492
G. Cheung, R. Rensvold,. Evaluating goodness-of-fit indexes for testing measurement
invariance. Structural Equation Modeling, 2002, 9 (2), pp. 233 – 255.
T. Childers, C. Carr, J. Peck, S. Carson, Hedonic and utilitarian motivations for online
retail shopping behavior. Journal of Retailing, 77(4), 2001, pp. 511-535.
J. Cox, B. Dale, Key Quality factors in Web Site design and use: an examination,
International Journal of Quality & Reliability Management 19(7), 2002, pp.
P. Dabholkar, D. Thorpe, J. Rentz, A Measure of Service Quality for Retail Stores: Scale
Development and Validation, Journal of the Academy of Marketing Science
24 (1), 1996, pp. 3-16
F. D. Davis, “Perceived usefulness, perceived ease of use and user acceptance of
information technology,” MIS Quart., vol. 13, no. 3, 1989, pp. 319–340
S. Devaraj, M. Fan, R. Kohli, Antecedents of B2C Channel Satisfaction and Preference:
Validating e-Commerce Metrics, Information Systems Research, 13(3), 2002,
pp. 316-333.
E. Deutskens, K. Ruyter, M. Wetzels, P. Oosterveld, Response Rate and Response
Quality of Internet-Based Surveys: An Experimental Study, Marketing
Letters 15(1), 2004, pp. 21-36
Diamantopoulos, J. Siguaw, Introducing LISREL. London: Sage Publications, 2000
W. Doll, W. Xia, G. Torkzadeh, A Confirmatory Factor Analysis of the End-User
Computing Satisfaction Instrument, MIS Quarterly18(4), 1994, pp. 453-461
O. Furrer, B. Liu, D.Sudharshan, The Relationships between Culture and Service Quality
Perceptions: Basis for Cross-Cultural Market Segmentation and Resource
Allocation, Journal of Service Research, 2 (4), 2000, pp. 355-71.
R. Gorsuch, Common Factor Analysis versus Component Analysis: Some Well and Little
Known Facts, Multivariate Behavioral Research 55(1), 1990, pp.33-39
S. Gounaris, S. Dimitriadis, Assessing Service Quality on the Web: Evidence from
Business-to-Consumer Portals, Journal of Services Marketing,17(5), 2003,
J. Hair, R. Anderson, R. Tatham, W.Black, Multivariate Data Analysis, Fifth Edition,
New Jersey: Prentice-Hall, Inc., 1998
G. Hofstede, Culture’s Consequences: International Differences in Work-Related Values.
Newbury Park, CA: Sage, 1980.
Hofstede, G. (1991), Cultures and Organizations, HarperCollins, London, 1991
L. Hu, P. Bentler, Cutoff criteria for fit indexes in covariance structure analysis:
Conventional criteria versus alternatives, Structural Equation Modeling 6(1),
1999, pp.1-55
Y. Hwang, An empirical examination of individual level cultural orientation as an
antecedent to ERP systems adoption, 2004, research-in-progress [Available
online at]
J. Illieva, S.Baron, N. Healey, Online Surveys in Marketing Research: Pros and Cons,
International Journal of Market Research 44(3), 2002, pp.361-376
M. Jun, S. Cai, The key determinants of Internet banking service quality: a content
analysis, International Journal of Bank Marketing 19(7), 2001, pp. 276-291
S. Kaynama, C. Black, A Proposal to Asses the Service Quality of Online Travel
Agencies, Journal of Professional Services Marketing 21(1), 2001, pp. 63-68
H. Kim, J. Kim, Y. Lee, M. Chae, Y. Choi, An empirical Study of the Use Contexts and
Usability Problems in Wireless Internet, In Proceedings of the 35th Hawaii
International Conference on System Sciences, 2002
J. Kim, I. Lee, Y. Lee, B. Choi, Exploring E-business Implications of the Mobile Internet:
A Cross-national Survey of Hong Kong, Japan and Korea, International
Journal of Mobile Communication, 2 (1), 2004, pp.1-21
T. Koivumaki, Consumer Attitudes and Mobile Travel Portal. Electronic Markets 12(1),
2002, pp. 47-57
Lee, Y., Kim, J., Lee, I. & Kim, H., A Cross-cultural Study on the Value Structure of
Mobile Internet Usage: Comparison Between Korea and Japan, Journal of
Electronic Commerce Research, 3 (4), 2002
T. Little, Mean and covariance structures (MACS) analyses of cross-cultural data:
Practical and theoretical issues. Multivariate Behavioral Research, 32 (1),
1997, pp.53-76.
C. Liao, P. Palvia and Hong-Nan Lin (2006), The roles of habit and web site quality in e-
commerce, International Journal of Information Management, 26 (6), 2006,
C. Liu, P. Arnett, Exploring the factors associated with Web site Success in the content of
electronic commerce, Information & Management 38(1), 2000, pp. 23-34
E. Loiacono, R. Watson, D.Goodhue, WebQual: An Instrument for Consumer
Evaluations of Web Sites, International Journal of Electronic Commerce 11
(3), 2007, forthcoming
R. MacCallum, M. Browne, H. Sugawara, Power Analysis and Determination of Sample
Size for Covariance Structural Modeling, Psychological Methods; 1996 (1)1,
C. Mathwick, N. Malhotra, E. Rigdon, Experiential Value: conceptualization,
measurement and application in the catalog and Internet shoppimg
environments. Journal of Retailing 77(1), 2001, pp. 39-56
S. McQuitty, Statistical Power and Structural Equation Models in Business Research,
Journal of Business Research 57 (2), 2004, pp.175-183
D. Reibstein, What Attracts Customers to Online Stores, and What Keeps them Coming
Back? Journal of the Academy of Marketing Science 30 (4), 2002, pp. 465-
R. Schumacker, R. Lomax, A Beginner’s Guide to Structural Equation Modeling, 2nd
Edition, Lawrence Erlbaum Associates, Mahwah, 2004
C. Sing, The Measurement, Analysis and Application of the Perceived Usability of
Electronic Stores. Singapore Management Review 26 (2), 2003, pp. 49-64
J. Sir, J. Yoo, C. Cho, H. Yang, 2003 Survey on the Wireless Internet Use, Korea
Network Information Center, 2003, Seoul, Korea.
T. Sommers, K. Nelson, J. Karimi, Confirmatory Factor Analysis of the End-User
Computing Satisfaction Instrument: Replication within an ERP Domain.
Decision Sciences 34(3), 2003, pp. 595-621
S. Srinivasan, R. Anderson, K. Ponnavolu, Customer loyalty in e-commerce: an
exploration of its antecedents and consequences, Journal of Retailing 78(1),
2002, pp. 41-50
J. Steenkamp, H. Baumgartner, Assessing measurement invariance in cross-national
consumer research. Journal of Consumer Research, 25 (1), 1998, pp. 78-90.
D. Straub, M. Keil, W. Brenner, Testing the Technology Acceptance Model across
Cultures: A Three-Country Study, Information & Management, 33 (), 1997,
pp. 1-11.
J. Swinder, J. Trocchia, P. Gwinner P, Consumer Perceptions of Internet Retail Service
Quality, International Journal of Service Industry Management 13(5), 2002,
pp. 412-431
N. Tsikriktsis, Does culture influence web site quality expectations? An empirical study.
Journal of Service Research, 5(2), 2002, pp.101-113.
L. Ueltschy, M. Laroche, R. Tamilia, P. Yannopoulos , Cross-Cultural Invariance of
Satisfaction and Service Quality Measures, Journal of Business Research, 57
(8), 2004, pp. 901-912
V.Venkatesh, F. Davis, A Theoretical Extension of the Technology Acceptance Model:
Four Longitudinal Field Studies. Management Science 46 (2), 2000, pp.186-
M. Wolfinbarger, M. Gilly, eTailQ: dimensionalizing, measuring and predicting e-tail
quality, Journal of Retailing 79 (3), 2003, pp.183-198
K. Woo, K. Fock, Customer Satisfaction in the Hong Kong Mobile Phone Industry, The
Service Industries Journal 19 (3), 1999, pp.162-174
Z. Yang, R. Peterson, L. Huang, Taking the Pulse of Internet Pharmacies, Marketing
Health Service (Summer), 2001, pp. 5-10
V. Zeithaml, A. Parasuraman, A. Malhotra, Service Quality Delivery Through Web
Sites: A Critical Review of Extant Knowledge, Journal of the Academy of
Marketing Science 30(4), 2002, pp. 362-375
Table 1. Definitions of e-SQ Measures/Constructs
Measure Definition Relevant Studies
Ease of Use
The extent to which a person
believes that using wireless
Internet services will be free of
effort and easy to learn
Loiacono, Goodhue and
Watson(2007); Sing
(2003);Wolfinbarger and Gilly
(2003);Cox and Dale
(2002);Koivumaki (2002); Zeithaml,
Parasuraman and Malhotra
(2002);Chae et al. (2002);Yang and
Peterson (2001); Jun and Kai
(2001);Mathwick, Malhotra and
Rigdon (2001); Venkatesh and
Davis (2000)
The extent to which a person
believes that using wireless
Internet services will enhance his
or her performance
Zeithaml, Parasuraman and
Malhotra (2003);Chaet al.
(2002);Liu and Arnett (2000);
Venkatesh and Davis (2000)
The extent to which wireless
Internet services are visually
attractive and pleasant
Loiacono, Goodhue and
Watson(2007); Chae and Kim
(2003); Zeithaml, Parasuraman and
Malhotra (2003); Swinder, Trocchia
Gwiner (2002);Jun and Kai (2001) ;
Yang and Peterson (2001)
The variety of content offered
(depth & width) through wireless
Internet services
Srinivasan, Anderson and
Ponnavolu (2002);Kayanama and
Black (2001)
The respect of personal
information shared through
wireless Internet services
Wolnfinabrager and Gilly (2003);
Zeithaml, Parasuraman and
Malhotra (2003); Koivumaki
(2002); Yang and Peterson (2001)
The ability to adapt and
personalize wireless Internet
services to individual preferences
Wolfinbarger and Gilly (2003);
Srinivasan, Anderson and
Ponnavolu (2002); Kayanama and
Black (2001)
Responsive and helpful service
that responds to customer
inquiries quickly
Wolfinbarger and Gilly (2003);
Zeithaml, Parasuraman and
Malhotra (2003);Koivumaki
(2002);Reibstein (2002) ];Woo and
Fock (1999)
Table 2. Perceived Mobile Internet Services Quality:
Exploratory Factor Analysis, N= 4, 456
Factors Measures Loadings
F1: Efficiency
(23.2 % of
Ease of use (how easily can I learn to use the service) .66
Usefulness (how useful the service offerings are to me) .67
F2: Outcome
(15.8% of
Aesthetics (how enjoyable and visually attractive is to use
the service) .63
Width/Depth of Content (for example, number of items
available to download) .57
F3: Customer
Care Quality
(13.5% of
Privacy (minimization of personal data that I need to
disclose to the service provider) .65
Level of personalization (whether I can personalize the
service to my tastes) .66
Customer service (whether the provider is able to support
me effectively in problems I might face) .75
Figure 1. The Perceived Service Quality Construct
Ease of Use
Customer Care
Outcome Quality
Efficiency Quality
Service Quality
Content Variety
Figure 2. Alternative Factor Structures of the Perceived Service Quality Construct
Ease of Use
Customer Care
Outcome Quality
Efficiency Quality
Content Variety
Ease of Use
Customer Care
Outcome Quality
Efficiency Quality
Content Variety
Ease of Use
Content Variety
Service Quality
Three first-order factor model (Uncorrelated) Three first-order factor model (Correlated)
First-order factor model
Table 3. Model Fit Criteria
9 This model was unidentified, and based on the suggestions of AMOS 5.0, two parameter loadings were
fixed to one.
Korea Japan Hong-
point Reference
Three factor
Three factor
and Lomax
(p=.00) 2444.63 (p=.00) 223.91
Df 14 16
11 11 11 11
CFI >0.95 Hu and
Bentler (1999) .61 .70 .97 .97 .945 .97
Delta 2 >0.90 Bagozzi and
Yi (1988) .41 .70 .97 .97 .945 .97
RMR <0.08 Hu and
Bentler 1999) .09 .23 .03 .03 .045 .039
RMSEA <0.06 Hu and
Bentler (1999) .10 .18 .066 .066 .108 .083
Figure 3. Standardized Parameter Estimates (Korean Sample)
Ease of Use
Customer Care
Outcome Quality
Efficiency Quality
Content Variety
Service Quality
Figure 4. Standardized Parameter Estimates (Japanese Sample)
Ease of Use
Customer Care
Outcome Quality
Efficiency Quality
Content Variety
Service Quality
Figure 5. Standardized Parameter Estimates (Hong-Kong Sample)
Ease of Use
Customer Care
Outcome Quality
Efficiency Quality
Content Variety
Service Quality
Table 4. Psychometric Properties of First-Order Factors
Korea Japan Hong-Kong
Loadings and t-
(ρc) AVE Standardized
Loadings and t-values (ρc) AVE Standardized
Loadings and t-values (ρc) AVE
γ11 Service Quality-
Efficiency Quality .80 (*) .87 (*) .86 (*)
γ21 Service Quality-
Outcome Quality .81 (t=20.73) .75 (t=18.07) .67 (t=17.97)
γ31 Service Quality-
Customer Quality .71 (t=19.93) .80 (t=17.49) .85 (t=16.15)
Efficiency Quality .70 .54 .76 .62 .70 .54
y1 Ease of use (how easily
can I learn to use the
.62 (*) .78 (*) .69 (*)
y2 Usefulness (how useful
the service offerings are to
.84 (t=28.53) .79 (t=28.60) .78 (t=22.54)
Outcome Quality .61 .44 .73 .57 .84 .72
y3 Aesthetics (how
enjoyable and visually
attractive is to use the
.74 (t=25.19) .79 (t=23.45) .75 (t=25.00)
y4 Width/Depth of Content
(for example, number of
items available to
.58 (*) .72 (*) ..94 (*)
Customer Care Quality .77 .53 .74 .49 .78 .55
y5 Privacy (minimization of
personal data that I need to
disclose to the service
.67 (t=37.31) .75 (t=23.82) .75 (t=25.28)
y6 Level of personalization
(whether I can personalize
the service to my tastes)
.72 (*) .61 (*) .66 (*)
y7Customer service
provided (whether the
provider is able to support
me effectively in any
problem I might have).
.80 (t=38.59) .73 (t=23.60) .81 (t=26.17)
Table 5. Factor Means Comparisons
Efficiency Outcome Quality Customer Care
Korea- HK Korea>HK Korea>HK Korea>HK
Japan- HK n.a. n.s. HK>Japan
Japan- Korea Japan>Korea Korea>Japan n.a.
Notes: n.a.= not applicable due to non-invariance, n.s.= not statistically significant
... This is also the case for SQ in the electronic and mobile environment (Aladwani, 2002;Francis, 2009, p. Dabholkar et al. (1996) in the context of retail stores and underpinned with empirical evidence by (Brady and Cronin, 2001). Especially the more recent papers on MSQ incorporate this hierarchical approach to assess the different connotations of SQ (e.g., Stiakakis and Petridis, 2014;Vlachos et al., 2011). ...
... The valence dimension is at least partially addressed in several identified papers using var-ious denotations (e.g., Bauer et al., 2006;Cai and Jun, 2003;Loiacono et al., 2007;Vlachos et al., 2011). In line with the positive experience that a high SQ mobile app should create, Fassnacht and Koese (2006) introduce the "emotional benefit" (p. ...
... It measures the customer's perceived feelings of the online purchase and its relation to the trust in the retailer. Vlachos et al. (2011) incorporate aspects of the valence dimension within usefulness. An enhancement of performance when using m-services positively influences a customer's feelings (p. ...
Full-text available
The increasing utilization of mobile apps for shopping leads retailers to provide customers with dedicated mobile shopping companion apps to create an omni-channel shopping experience involving traditional brick-and-mortar, electronic and mobile business. Mobile shopping companion apps extend the traditional and electronic services of brick-and-mortar retailers by an additional mobile channel providing the customer with a digital companion supporting the shopping within and outside the stores using mobile technology. A twofold approach is pursued in this thesis. Firstly, a structured literature review is conducted to identify candidate dimensions for developing a scale for measuring the service quality of mobile shopping companion apps. Secondly, design requirements for improving the service quality of these mobile apps are deduced from online customer reviews of three exemplary mobile shopping companion apps applying a qualitative content analysis. The mobile app service quality of mobile shopping companion apps can be measured using a hierarchical and multi-dimensional scale consisting of three primary dimensions, seven secondary dimensions and 22 related items. The primary dimensions interaction quality, environment quality and outcome quality structure the secondary dimensions responsiveness, information, security and privacy, design, performance, technical reliability and valence. Based on these dimensions, 22 implementation guidelines and 14 service design requirements are derived as potential areas for optimizing the mobile app service quality of mobile shopping companion apps and achieving a high overall service quality. A mobile shopping companion app should include a set of features consisting of 16 features from three different areas. Results show that measuring the service quality of mobile shopping companion apps require for a tailored measurement scale. Equally, design requirements are proposed for this particular category of mobile apps. Retailers should provide a single mobile shopping companion app providing all features and mobile services to the customer.
... Page 10 efficiency, system availability, promise fulfillment, and privacy. Vlachos et al. (2011) proposed seven antecedents of e-service quality including ease of use, usability, aesthetics, content, privacy, customization and customer service. ...
... This explanatory research aims to examine and explain the causal relationship between e-service quality, e-trust, and esatisfaction in the context of online learning. Measurement of e-service quality refers toVlachos et al. (2011). The measurement of e-trust is in accordance with the opinion of Doney & Cannon (1998),Giffin (1967),Luhmann (1988), andMayer & Davis (1999). ...
Online learning was applied during the COVID-19 pandemic in almost all universities in Indonesia. Online learning needs to be evaluated for its effectiveness, especially regarding the responses from students. This study examines the effect of e-Service Quality on e-Trust and e-Satisfaction during the covid-19 pandemic. The study was conducted on 1,212 students who have participated in online learning by distributing online questionnaires. The analytical tools used are SPSS and WarpPLS. The results of this study indicate that e-service quality has a significant positive effect on e-trust, e-service quality has a significant positive effect on e-satisfaction, and e-trust has a significant positive effect on e-satisfaction.
... It is need of time to identify determinants of m-internet, because of high competition in the industry. Previous studies found perceived usefulness as an important determinant of m-Internet service quality (Lu, Zhang, and Wang 2003;Kim, Chan, and Gupta, 2007;Tan and Chou, 2008;Vlachos, Giaglis, Lee, and Vrechopoulos, 2011). Most of the studies also found content quality (Chae, Kim, Kim, and Ryu 2002;Cheong and Park, 2005;Kuo, Wu, & Deng, 2009) and ease of use ( Cheong and Park, 2005;Tan and Chou, 2008) as crucial measures of m-Internet service quality. ...
... Some USA based studies have also explored crucial measures of m-CSQ scale in different contexts ( Lu et al., 2003;Lim et al., 2006;Malhotra and Malhotra, 2012). Some authors have determined the crucial measures of m-CSQ measurement scale through cross-cultural studies ( Lee et al., 2004;Vlachos et al., 2011;Dwivedi et al., 2016). Usefulness, ease of use, responsiveness, security, content, network quality, tangibility, application design, and system availability have been identified as consistent determinants of m-CSQ measurement scale in most of the contexts. ...
Service quality has captured the attention of many service marketing researchers over past three decades. Globalization and information technology development have witnessed the metamorphosis from traditional to digital business and service quality has been replaced by mobile service quality. Mobile commerce (m-commerce) refers to any transaction trough mobile devises. Siau, Ee-Peng, and Shen (2001) defined m-commerce as innovative and new version of electronic commerce, in which transaction are conducted through wireless telecommunication network. This study undertakes a systematic review of mobile commerce service quality (m-CSQ) related articles to critically analyze scale development methodology and to know major determinants of measurement scales in different contexts of mobile commerce. An extensive qualitative literature review of 68 m-CSQ measurement scale related articles have been conducted to fulfill above objectives. Articles were selected from EBSCOhost, ABI/INFORM, Google Scholar, ProQuest Direct, Wiley, Scopus, Science Direct and Emerald Insight databases by manual search. Initially 122 articles were identified by this process. After further refining, a total of 68 articles were included in this study. Data were analyzed in two steps; in first step m-CSQ scale development process adopted by different authors has been explored and in second step different determinants of measurement scale have been identified in four major mobile commerce contexts. Furthermore, origin country of the studies was also explored for better understanding of growth and development of different context of mobile commerce and overall conceptual framework of m-CSQ has been developed. M-commerce was divided into different contexts such as; m-retailing, m-heath, m-internet, m-entertainment, m-telecommunication, m-service, m-banking, and m-payment. Privacy and security, content, responsiveness, efficiency, reliability, ease of use and usefulness were identified as consistent determinants of m-CSQ measurement scales in most of the contexts. This study provides framework of m-CSQ scale to guide managers and scholars to identify crucial determinants of m-CSQ in different contexts.
... Para Chang & Wong (2010) a principal tarefa do e-procurement é identificar, através de um sistema informatizado, os possíveis fornecedores para um determinado material. Pesquisas recentes indicam a importância de se avaliar a percepção do usário do serviço de e-procurement, tanto no de ações governamentais (Aman & Kasimin (2011), Concha, Astudillo, Porrúa, & Pimenta (2012) e Johnson (2011) quanto no contexto de iniciativa privada (Samiee (2008), Walters (2008), Vlachos et al (2011). Andreu et al (2010), Ladhari (2010) e Brandon-Jones, & Carey. ...
Full-text available
Na literatura científica há indícios de que a adoção de ferramentas de e-procurement melhoram o desempenho das organizações em diversos aspectos. Mas, como estas empresas percebem os impactos oriundos da adoção de ferramentas de eprocurement em seus negócios? Essa percepção é semelhante a do fornecedor desse tipo de serviço? Apesar das pesquisas reportadas na literatura, estas questões permanecem pendentes. O presente trabalho apresenta e aplica, um questionário para o mapeamento e contraste dessas percepções, permitindo a identificação de aspectos a serem melhorados para o caso particular estudado.
... The level of customers' needs satisfaction is usually determined on the basis of customer surveys. Processed using mathematical-statistical methods, the results of a questionnaire survey provide a more objective picture of the customer needs and lead to more sound conclusions [12]. It is appropriate applying to specialized market research companies, which carry out such studies for service providers that do not have a chance to do customer surveys themselves. ...
Full-text available
Neither scientific nor special literature contains systematic information about the actions required for maintenance service market penetration, and most of this kind of research is purely theoretical, not linked to practical possibilities and addressing these issues only fragmentary. Procedural model with flexible structure is developed; the model consists of 9 reasoned components. The article provides systematized information about the market penetration, based on modern scientific findings. The proposed model shows all the necessary actions, sequence of the performances, content of the actions and possibilities for using. The maintenance services market penetration model was tested and found to be suitable. This resulted in the emergence of a new maintenance company in the Lithuanian market. The use of the model will help to design a market penetration strategy for maintenance companies. Also, the use of the model can help to penetrate the maintenance market of any country which has a market economy and legitimized competition.
... The customer's ex-post impression of the service delivery. VAL1: Overall satisfaction with the provided service VAL2: Satisfaction with the scope of provided services [22,43,46,47,51,52] Following Pagano and Maalej [39], we first extracted review data from Apple's App Store and Google's Play Store using a paid version of the online service on 2017-07-03. ...
Conference Paper
Full-text available
Shopping companion apps, which assist customers in product search and buying decisions, are an emerging phenomenon in the context of omnichannel retail. These retailer-provided apps link the digital with the physical servicescape of the store, allowing for new forms of online and at the same time physical service. So far, there is no dominant design for this type of information system. Both academia and practice lack empirical information about what customers expect from this kind of mobile app. Drawing from service quality literature as theoretical foundation, we conducted a qualitative content analysis of 1,448 customer reviews of three major shopping companion apps. The analysis yielded 23 aspects that customers expect from shopping companion apps, and that, in turn, can support establishing high mobile service quality. Our results contribute to the knowledge of m-service in retail and quality-driven app design.
... Today, many services are provided by a customer's interaction with self-service technology. Self-service technologies enable customers to get served without a help of human service providers [16,17]. Taco Bell announced a new ordering system that uses a digital application that is 20% more expensive than human cashiers. ...
The technology-based service encounter has received significant attention with the advance of technology, especially artificial intelligence robots, in the healthcare industry. In the technology-based service encounter, the technology as a service provider plays a critical role in succeeding the service encounter. The purpose of this study is to develop a framework for the success factors of a technology-based service encounter in the healthcare industry. Roles of subjects (human and technologies) and success factors in the encounter are reviewed and proposed. A systematic literature review supports the proposed framework, which explains success factors for technology-based service encounters in the healthcare industry.
This study seeks to explain why telecom customers would continue to use mobile value-added services (MVAS), including information, communication, entertainment, and transaction services. We developed a model based on the S-O-R paradigm and existing research on MVAS and mobile service quality, in which we hypothesize that customers' intention to continue using MVAS as a behavioural response (R) is a direct result of their satisfaction with the services and their perception of its value as affective and cognitive states (O), respectively. The developed model also recognizes the importance of mobile service quality (S) in terms of characteristics and features in developing customer satisfaction and perceived value. In this investigation, mobile service quality aspects: customer service, service content, and mobile network quality are utilized as stimuli. The data was collected from 371 respondents utilizing an online survey instrument. PLS-SEM with SmartPLS3 software was used for data analysis. While both customer satisfaction and perceived value have direct positive influences on continuance usage intention, the customer satisfaction effect is more significant, according to the structural findings. Moreover, customer service quality, service content quality, and mobile network quality were all discovered to be direct predictors of perceived value. In contrast, mobile network quality was not revealed to be a significant predictor of customer satisfaction. This paper is one of the few investigating the major factors contributing to the continued use of MVAS in the Arab world. This work adds to the growing knowledge on post-adoption in mobile services and business. This investigation provides crucial suggestions for decision-makers in the mobile telecommunications industry from a practical standpoint.
Nowadays, retailers are interested in how customer preferences regarding service quality are changing due to the adoption of different devices for shopping purposes in both the desktop and mobile contexts. To answer this question, this paper first replicates, in the mobile commerce (m-commerce) context, the results from Blut et al. (2015), who conducted a meta-analytic review of electronic service quality. Replication results question the robustness and generalizability of the conceptualization in the mobile service quality context. Thereby, practitioners and academics are encouraged to adapt a customer-centric approach in organizing marketing practices. The replication extends the conceptualization of electronic service quality by considering a unique dimension named ubiquity of services, defined as the retailer's ability to provide offers based on location and time. To reveal psychological mechanisms explaining the results of the replication study, a follow-up study draws on these contextual factors. In this context, this study uses a quasi-experimental approach by utilizing propensity score matching to account for self-selection effects to examine differences between desktop and mobile device users. As a result, this research contributes to the literature by identifying contextual boundary conditions regarding the shopping trip intentions and risk perceptions of mobile device users and desktop device users. Based on the results, major implications for retailers and further research are given.
There is little research that addresses methodological issues in the context of e-service quality. This study content analyzed e-service quality articles published between 2003 and 2013 to assess methodological issues in this field. Sampling methods (probability vs. nonprobability), use of undergraduates (UGs) in research, service context, geographic location, demographic information, research strategy, and justification for the use of UGs were identified as key variables. In addition, the authors argued for the need to conduct replications in research on e-service quality. A total of 72 research articles on e-service and e-service quality were content analyzed. All articles focused primarily on e-service quality in the context of B2C e-retailing. On the basis of the findings the present study provides several suggestions that should receive close attention from e-service quality researchers.
Full-text available
In a recent article on conducting international marketing research in the twenty-first century (Craig & Douglas 2001), the application of new (electronic) technology for data collection was encouraged. Email and web-based data collection methods are attractive to researchers in international marketing because of low costs and fast response rates. Yet the conventional wisdom is that, as some people still do not have access to email and the Internet, such datacollection techniques may often result in a sample of respondents that is not representative of the desired population. In this article we evaluate multimode strategies of data collection that include web-based, email and postal methods as a means for the international marketing researcher to obtain survey data from a representative sample. An example is given of a multimode strategy applied to the collection of survey data from a sample of respondents across 100 countries.
Full-text available
Despite the critical need to know how consumers' perceptions of Web sites influence their behavior, and especially their intention to revisit or purchase, there is no extant general measure for evaluating Web sites and no consensus on what such an instrument should measure. The authors used the Theory of Reasoned Action and the Technology Acceptance Model to develop the WebQual instrument for consumer evaluation of Web sites. They refined it through a literature review and interviews with Web designers and users, and tested it using four samples of Web consumers. WebQual includes 12 dimensions (informational fit-to-task, tailored information, trust, response time, ease of understanding, intuitive operations, visual appeal, innovativeness, emotional appeal, consistent image, on-line completeness, relative advantage) and shows strong measurement validity. It is a highly validated instrument that can provide both wide- and fine-grained measurements of organizational Web sites.
Full-text available
Using the means-end framework as a theoretical foundation, this article conceptualizes, constructs, refines, and tests a multiple-item scale (E-S-QUAL) for measuring the service quality delivered by Web sites on which customers shop online. Two stages of empirical data collection revealed that two different scales were necessary for capturing electronic service quality. The basic E-S-QUAL scale developed in the research is a 22-item scale of four dimensions: efficiency, fulfillment, system availability, and privacy. The second scale, E-RecS-QUAL, is salient only to customers who had nonroutine encounters with the sites and contains 11 items in three dimensions: responsiveness, compensation, and contact. Both scales demonstrate good psychometric properties based on findings from a variety of reliability and validity tests and build on the research already conducted on the topic. Directions for further research on electronic service quality are offered. Managerial implications stemming from the empirical findings about E-S-QUAL are also discussed.
Full-text available
The authors argue that perceptions of service quality vary across cultural groups, as defined by each culture’s position on Hofstede’s dimensions. They explicitly map the relationship between service quality perceptions and cultural dimension positions and draw the implications for international service market segmentation. They also test the hypotheses constituting their theoretical analysis. They show that the importance of SERVQUAL dimensions is correlated with Hofstede’s cultural dimensions. They also used the correlation coefficients to compute a Cultural Service Quality Index that could be used to segment international service markets and allocate resources across segments.
Full-text available
Focuses on the issues associated with Internet banking service quality. Customer anecdotes of critical incidents in Internet banking were content-analyzed. Identified a total of 17 dimensions of Internet banking service quality, which can be classified into three broad categories – customer service quality, banking service product quality, and online systems quality. The derived dimensions include: for customer service quality, ten dimensions such as reliability, responsiveness, competence, courtesy, credibility, access, communication, understanding the customer, collaboration, and continuous improvement; for online systems quality, six dimensions such as content, accuracy, ease of use, timeliness, aesthetics, and security; and for banking service product quality, one dimension of product variety/diverse features. Also revealed that, in terms of frequency of references to the 17 dimensions, no substantial differences exist between Internet-only banks and traditional banks offering Internet banking service. The most frequently mentioned dimensions, as the main sources of satisfaction or dissatisfaction, were reliability, responsiveness, access, and accuracy. Some suggestions and recommendations were provided to improve the Internet banking service quality and, in turn, customer satisfaction.
Cross-validation is an important and often neglected step in the scientific process. Measurement models can vary across samples and must be tested and retested before they are accepted as valid. In a review of user satisfaction instruments, Klenke concludes that there is an appalling lack of effort to cross-validate MIS instruments and calls for efforts to retest the End-User Computing Satisfaction (EUCS) instrument using new data. Using different sampling methods and a new sample of 359 respondents, this study replicates an earlier confirmatory factor analysis of the EUCS instrument. This replication suggests that the EUCS instrument is robust (i.e., not affected by sampling methods) and can be used with confidence to evaluate information systems.
This paper reviews prior applications of structural equation modeling in four major marketing journals (the Journal of Marketing, Journal of Marketing Research, International Journal of Research in Marketing, and the Journal of Consumer Research) between 1977 and 1994. After documenting and characterizing the number of applications over time, we discuss important methodological issues related to structural equation modeling and assess the quality of previous applications in terms of three aspects: issues related to the initial specification of theoretical models of interest; issues related to data screening prior to model estimation and testing; and issues related to the estimation and testing of theoretical models on empirical data. On the basis of our findings, we identify problem areas and suggest avenues for improvement.
Through qualitative and empirical research, the authors find that the service quality construct conforms to the structure of a third-order factor model that ties service quality perceptions to distinct and actionable dimensions: outcome, interaction, and environmental quality. In turn, each has three subdimensions that define the basis of service quality perceptions. The authors further suggest that for each of these subdimensions to contribute to improved service quality perceptions, the quality received by consumers must be perceived to be reliable, responsive, and empathetic. The authors test and support this conceptualization across four service industries. They consider the research and managerial implications of the study and its limitations.
The purpose of this article is to illustrate the steps involved in testing for multigroup invariance using Amos Graphics. Based on analysis of covariance (ANCOV) structures, 2 applications are demonstrated, each of which represents a different set of circumstances. Application 1 focuses on the equivalence of a measuring instrument and tests for its invariance across 3 teacher panels, given baseline models that are identical across groups. Application 2 centers on the equivalence of a postulated theoretical structure across adolescent boys and girls in light of baseline models that are differentially specified across groups. Taken together, these illustrated examples should be of substantial assistance to researchers interested in testing for multigroup invariance using the Amos program.