See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/289813428

Service quality attributes affecting customer satisfaction in banking sector of

India

Article in European Journal of Economics, Finance and Administrative Sciences · September 2010

CITATIONS READS

12 4,179

4 authors, including:

Uma Sankar Mishra Snigdha Pattnaik

Central University of Rajasthan SCHOOL OF PHARMACEUTICAL SCIENCES, SOA UNIVERSITY,
53 PUBLICATIONS 399 CITATIONS 51 PUBLICATIONS 680 CITATIONS

SEE PROFILE SEE PROFILE

Ayasa Kanta Mohanty

Siksha O Anusandhan University
14 PUBLICATIONS 71 CITATIONS

SEE PROFILE

All content following this page was uploaded by Uma Sankar Mishra on 06 August 2016.

The user has requested enhancement of the downloaded file.

European Journal of Economics, Finance and Administrative Sciences
ISSN 1450-2275 Issue 24 (2010)
© EuroJournals, Inc. 2010
http://www.eurojournals.com

Service Quality Attributes Affecting Customer Satisfaction in

Banking Sector of India

Uma Sankar Mishra

Siksha ‘O’ Anusandhan University, Bhubaneswar, India, Pin-751030
E-mail: [email protected]

Jyoti Ranjan Das

Siksha ‘O’ Anusandhan University, Bhubaneswar, India, Pin-751030
E-mail: [email protected]

Sanjib Pattnaik
Siksha ‘O’ Anusandhan University, Bhubaneswar, India, Pin-751030
E-mail: [email protected]

Ayasa Kanta Mohanty

Siksha ‘O’ Anusandhan University, Bhubaneswar, India, Pin-751030
E-mail: [email protected]

Abstract
Customer service is the primary end of any bank. A customer always wants something and
expects that the bank should come up to the level to fulfil those needs. Again, the more you
provide, still more the consumer needs. Service quality is about meeting customers’ needs
and requirements, and how well the service level delivered matches customer expectations.
Service quality in banking implies consistently anticipating and satisfying the needs and
expectations of customers. In this context, the present paper proposes two structural
equation models (SEMs), one for public and another for private sector banks in India, to
show the relationship between customer satisfaction on bank services and the attributes of
the perceived service quality. Specifically, structural equation model is formulated to
explore the impact of the relationship between customer satisfaction and service quality
attributes.

Keywords: Service Quality, Customer Satisfaction, Banking Sector

1. Introduction
Service quality is about meeting customer needs satisfactorily by matching to his expectations. Service
quality in banking implies consistently anticipating and satisfying the needs and expectations of
customers (Howcrof 1991). The importance of service quality in Banks has been emphasised in many
studies and perceived quality advantage leads them to higher profit (Raddon 1987; Buzzell & Gale
1987). Parasuraman and Berry (1991) holds the view that high quality service gives credibility to field
sales force. Heskett et al. (1990) observed that the longer a company keeps a customer, the more
money it stands to make. There is enough evidence that demonstrates the strategic benefits of quality in
contributing to profit, market share and returns on investment (Adrian 1995; Bateson 1995; Berry et al
89 European Journal of Economics, Finance and Administrative Sciences - Issue 24 (2010)

1991, Buzzel et al 1981, etc) and lowering cost and improving productivity (Garvin 1983; Kotler 2003;
etc.). Maximising customer satisfaction through quality customer service has been described as the
‘ultimate weapon’ (Davidow & Uttal 1989).
The banking industry as a service industry, directed towards the customer’s money and its
management. A membership relationship is entailed in this industry due to its continuous nature.
Banking is also high in credence quality, meaning that it can not be evaluated confidently even
immediately after receipt of the goods / services. In addition, an extended period of time may be
required in this industry for a fully informed evaluation (Devlin 2001). Hence, customer satisfaction in
banking is both difficult to measure and ascertain.
Although researchers have studied the concept of service for decades, there is no
conceptualisation of service quality (Cronin & Taylor, 1992; Rust & Oliver, 1994). Different
researchers focused on different aspects of service quality. The most common definition views quality
as the customers’ perception of service excellence (Berry et al. 1988; and Parasuraman et al. 1985).
The underlying meaning of this definition is that customers form the perception of service quality
according to the service performance they experience in the past. Therefore, the service quality is
dependent on customer’s perceptions of the service.
Today, the banking industry has become highly competitive in India. It is not only focusing on
providing wide range of products to create competitive advantages; but also emphasises on the
importance of services, particularly in maintaining service quality (Sousa 1999). Using the service
quality model (SERVQUAL of Zeithaml et al, 1988), a comparative study of customers of private and
public sector banks operating in India is undertaken to identify the differences between the banking
attitudes of the respondents toward the service quality management in banks. Further, the study also
attempts to identify the customers’ expectations of banking services and their respective performances.
In the light of the above research findings, interest in service quality is, thus, unarguably high.
Poor quality places a firm at a competitive disadvantage. If customers perceive quality as
unsatisfactory, they may be quick to take their businesses elsewhere. Thus, it is clear that service
quality offers a way of achieving success among competing services, particularly in the case of firms
that offer nearly identical services, such as banks, where establishing service quality may be the only
way of differentiating oneself. Such differentiation can yield a higher proportion of consumers’ choices
and, hence, mean the difference between financial success and failure.
In the changing banking scenario of 21st century, the banks have to have a strong identity to
provide world-class services. The banks are now have to be of world-class standard, committed to
excellence in customers, shareholders and employees’ satisfaction, and to play a leading role in the
expanding and diversifying financial sector (Balachandran, 2005). There has been a tremendous
change in the way of banking in the last few years. Customers have also rightly demanded world class
quality services from the banks. With multiple choices available, customers are not willing to put up
with anything less than the best. Banks have recognized the need to meet customers’ aspirations. So it
is an urgent drive to move the bank up in the high tech ladder.
The structural equation model proposed in this article investigates the impact of service quality
aspects on bank customer satisfaction both in public and private sectors in India. This article begins
with an introduction to a theoretical framework of structural equation models. Next, the demographic
survey is described and the statistical descriptive analysis of the sample is reported. The latter sections
describe the general structure of the proposed model and present the model results.

2. Structural Equation Models

In the last 25 years, the use of Structural Equation Modeling (SEM) in management research has
mushroomed. The idea and method itself are not new, but the availability of easy-to-use software (e.g.,
AMOS, EQS, LISREL) in particular has made SEM very popular among organization scientists, and
90 European Journal of Economics, Finance and Administrative Sciences - Issue 24 (2010)

numerous applications of SEM can be found in all leading management research literatures (Ketokivi
M, 2006).
Structural equation models describe relationships between variables. They are similar to
combining multiple regression and factor analysis (Bacon L. D et al, 1997). SEMs also offer some
important, additional benefits over these techniques including an effective way to deal with
multicollinearity, and methods for taking into account the unreliability of consumer response data.
SEMs can include two kinds of variables: observed and latent. Observed variables have data,
like the numeric responses to a rating scale item on a questionnaire such as gender or height. Observed
variables in SEMs are usually continuous. Latent variables are not directly observed, but you still want
to know about them. Although various forms of SEMs have been developed, majority of these express
linear relationships between variables. This is also how variables are usually related in regression and
factor analysis models — variables are expressed as weighted linear combinations of other variables.
Variables that depend on other variables are called “dependent.” Variables that do not depend on other
variables in a model are called “independent.” When using SEMs, these variables are also called
“endogenous” and “exogenous,” respectively.
Structural equation models, also called simultaneous equation models, are multivariate (i.e.,
multi-equation) regression models (Fox J, 2002). Unlike the more traditional multivariate linear model,
however, the response variable in one regression equation in an SEM may appear as a predictor in
another equation; indeed, variables in an SEM may influence one-another reciprocally, either directly
or through other variables as intermediaries. These structural equations are meant to represent causal
relationships among the variables in the model.
Wei K K in the year 2009, attempted a research study to construct a measure in service quality
for Malaysian banks. It focuses on the technique used in creating an index for service quality. The
SERVQUAL model was used as the underlying theoretical framework. The quantification of service
quality led to the attempt to construct an index. The index was constructed using Structural Equation
Modeling (SEM) and American Customer Satisfaction Index (ACSI) as the underlying frameworks.
An adapted ACSI was enhanced and improved to accommodate 2 exogenous constructs. The attempt
was successful (Wei K K, 2009). A survey was conducted on 350 respondents where 200 were
completed. Results provided evidence relevant to the literature where service attributes in service
quality could improve customer satisfaction.
Eboli L et al, 2007 in their research article a structural equation model has been proposed to
show the relationship between passenger satisfaction on bus services and the attributes of the services
supplied. Although SEM methodology is well known and widely applied in several fields of research,
presently there are not many practical applications in public transport, and specifically for measuring
customer satisfaction (Eboli L et al, 2007). In this research authors have applied this methodology on
the basis of needs and expectations expressed by customers of a bus service. The proposed model
identifies service quality attributes to improve, with the aim of offering bus services characterized by
higher levels of quality.

3. Study Design and Methodology

The data for the study were collected through a structured questionnaire from 387 customers selected
on stratified random basis. Initially 600 samples were planned covering 300 customers of both private
and public sector banks. Because of the small number of branches of private banks and their urban
concentration, unwillingness of the customers to provide data, time and budgetary constraints restricted
the sample size to 387, out of which 242 from public banks and 145 from private banks. A
questionnaire for customer survey is designed keeping the broad parameters in mind, which was pre-
tested before finalisation. The questionnaire containing all the 22 numbers of statements of
SERVQUAL instrument developed by Parsuraman et al for customer survey is administered keeping
the broad parameters in mind. The data regarding perceptions and expectations of customers were
91 European Journal of Economics, Finance and Administrative Sciences - Issue 24 (2010)

collected in a 7 – point interval scales, where 1 stands for strongly agree and 7 stands for strongly
disagree. All the data were collected from bank customers through personal contact approach.

4. Sample Profile
The demographic backgrounds of the sample respondents in six parameters are presented in Table 1 to
understand the customer profiles i.e., age, education, gender, occupation, and income. Table describes
the customer profile and the type of banks they have chosen for their transactions. It is evident from the
table that 58.82 per cent of the young respondents transacts with public sector banks, while, 41.18 per
cent with private banks. All the respondents in different age groups have shown similar behaviour,
except 40-60 years age group; where the dominance of public sector banks (78.79 %) is visible. This
may be due to the fact that they have an account with public sector banks before the entry of private
banks into their location in mid-1990s. The χ2 value is significant at 1 per cent level of significance
indicating age-groups of the respondents and their choices of the bank are dependent on each other.

Table 1: Customer Profile by Type of Banks

Public Private Total Chi

Parameters
f % f % f % Square
Below 30 years 110 58.82 77 41.18 187 100.00
30-40 years 61 59.80 41 40.20 102 100.00
Age 12.81*
40-60 years 52 78.79 14 21.21 66 100.00
60 years and above 19 59.38 13 40.63 32 100.00
Up to HSC 31 81.58 7 18.42 38 100.00
Graduate 67 62.62 40 37.38 107 100.00
Education 13.08*
Post Graduate 66 54.10 56 45.90 122 100.00
Professional 78 65.00 42 35.00 120 100.00
Male 191 64.31 106 35.69 297 100.00
Gender 2.07
Female 51 56.67 39 43.33 90 100.00
Salaried 134 59.56 91 40.44 225 100.00
Self-employed 6 75.00 2 25.00 8 100.00
Professional 12 48.00 13 52.00 25 100.00
Occupation Business 21 87.50 3 12.50 24 100.00 23.33*
Student 42 77.78 12 22.22 54 100.00
Retired 17 56.67 13 43.33 30 100.00
Housewife 10 47.62 11 52.38 21 100.00
Less than Rs.1 lakh 97 66.90 48 33.10 145 100.00
Rs.1 – 2 lakh 92 54.12 78 45.88 170 100.00
Income Rs.2 –3 lakh 31 73.81 11 26.19 42 100.00 15.27*
Rs.3-4 lakh 16 80.00 4 20.00 20 100.00
Rs.4 lakh and above 6 60.00 4 40.00 10 100.00
* 1% level of
All Samples 242 62.53 145 37.46 387 100.00
significance

It is observed from the same table that less educated customers (81.58 %) have a choice for
public banks and the post-graduates more inclined to private banks (45.90%). This trend is observed
may be due to the large presence of public banks with wide ranges products more suitable for lower
income groups, while private banks are offering value-added services for special group of customers
(mass-banking vs. class-banking approaches). The χ2 - value is significant at 1 per cent level of
significance indicating the dependence between type of education and choice of a bank. More or less
the behaviour in gender is similar, and the χ2 value for gender variation is not significant.
Taking into account the occupation of the customers, salaried persons dominate the sample.
The professionals and house-wives have shown more preference for private banks compared to the
other groups (52% and 52.38% respectively); while almost all the businessmen have chosen to transact
92 European Journal of Economics, Finance and Administrative Sciences - Issue 24 (2010)

with public sector banks (87.50 %). Students and self-employed persons have a clear preference for
public sector banks. The χ2 value is significant at 1 per cent level of significance, implying that
customers with occupational variations have a preference for a bank. Similarly, for different income
groups public sector banks are more preferred. The χ2 value is significant at 1 per cent level, indicating
the influence of levels of income for choice of a bank.

5. Model Development
Structural Equation Modeling approach allows the modeling of a phenomenon by considering both the
unobserved “latent” constructs and the observed indicators that describe the phenomenon (Eboli L et
al, 2007). SEMs are made up of two components: the first describes the relationship between
endogenous and exogenous latent variables, and permits the evaluation of both direction and strength
of the causal effects among these variables (latent variable model); the second component describes the
relationship between latent and observed variables (measurement model). The structural equation
system is generally estimated by using the maximum likelihood method (ML). In other cases, the
structural equation model parameters can be estimated by using other estimation methods, such as
unweighted least squares (ULS), weighted least squares (WLS), generalized least squares (GLS), and
so on.
In line with the objectives of the study, the main areas of questioning and analysis concerned
expectations and perceptions of service quality (ESQ and PSQ) and its dimensions: tangibility,
reliability, responsiveness, assurance and empathy. Mean differences between customer expectations
and perceptions of service quality of banks were calculated separately, showing the degree of customer
satisfaction represented by the gap, (PSQ-ESQ) or simply (P-E).
In the proposed structural equation models, the observed variables are the 22 service quality
attributes perceived by the bank customers (PSQ) and one customer satisfaction indicator (P-E). The
latent variables are the unobserved service quality aspects that can be explained by the observed
variables. The latent variables were defined by means of an exploratory factor analysis (EFA)
implemented in the form of principal component analysis. EFA was conducted by using a correlation
matrix. To determine the number of components, only the eigen values greater than or equal to 1 were
considered. An orthogonal rotated solution (Varimax) was adopted. In addition, the KMO test (Kaiser,
Mayer and Olkin) were implemented. Two separate structural equation models are proposed over here,
one for Public Bank and another for Private Bank.
By means of the EFA, four latent variables for public sector banks and six latent variables for
private sector banks were identified. In public sector banks, the first variable represents
Responsiveness and Empathy. The second variable relates Reliability and Tangibility. Similarly, third
and fourth variable represent Assurance and Reliability respectively. In private sector banks, the first,
second, third, fourth, fifth and sixth latent variables are Reliability / Assurance, Responsiveness /
Empathy, Tangibility, Empathy, Reliability / Responsiveness and Assurance / Empathy respectively.
The measurement model relates each extracted latent variable to the observed variable that
measures customer satisfaction. Specifically, we supposed that the exogenous latent variables are
measured by the 22 perceived service quality attributes and the observed variable “customer
satisfaction” is measured by the gap indicators of perceptions and expectations. Besides, all the
exogenous latent variables are correlated among them.

6. Model Description
After identifying all the unobserved latent variables and observed variables, path diagrams were drawn
and analyzed separately both for public and private banking sector by the help of Amos software
package, which are shown in figure 1 and 2. Figure 1 is the SEM of public sector bank customers and
here F1, F2, F3 and F4 are four extracted latent variables. The observed variable named by (P – E) is
93 European Journal of Economics, Finance and Administrative Sciences - Issue 24 (2010)

the indicator of degree of Customer Satisfaction and this variable depends on four perceived service
quality constructs F1, F2, F3, and F4. Similarly, Figure 2 is the SEM of private sector bank customers
containing same types of relationships among the variables having six latent variables F1, F2, F3, F4,
F5 and F6.
Figure 1: Structural Equation Model of Public Sector Bank Customers

2.04
1.13 1.58 2.45 1.80 2.26
1.86 1.39 e4 e6 e7
e3 1 e5 1 1
e8 1.97
e1 e2 1 1 1
e9
1 PSQ19 PSQ18 PSQ13 PSQ12 1
1 PSQ20 PSQ11
.75 1.37
1.20 1.02 PSQ10
PSQ21 1.36 1.03
1.22 1.36
1.00
PSQ22 e23
F1 .52
1.77 .42 1
1 1.23
e16 PSQ1
1.59 P_E
1
e15 PSQ2 1.00 .44
1.37 .88
1
e14 PSQ3 .74 1.71
1.36 .28
1 .55
e13 PSQ4 F2
1.07 .98
1
e12 PSQ5 .96 .16 .03
1.12 1.00
1 .82
e111.08 PSQ8 1.37
.14 .16
1
e10 PSQ9
2.96
1 .80
e19 PSQ14 .64
2.48
1 .90 F3
e18 PSQ15 .67
F4
1.07 1.00 1.00
1 1.78 1.58
e17 PSQ16 .99
1.90 PSQ6 PSQ7
1 1 1
e24 PSQ17 .94 2.11
e21 e20

The arrows show dependencies in the models. A double-curved arrow, such as that among all
the extracted latent variables, indicates that two variables co-vary. A single- headed arrow, such as
between latent variables and (P-E) variable, indicates that the variable receiving the arrow head, (P-E)
or Customer Satisfaction, depends on the variable from which the arrow originates, i.e constructs of
perceived service quality. Rectangular and circular symbols are for observed and unobserved variables
respectively. For each latent variable (F1, F2, etc.) different types of observed indicator variables are
attached to it depending on their respective factor loadings in two types of banking sector. All other
variables represented by e1, e2, etc. are the residual variables. Numeric figures written near single
headed arrow marks from F1, F2, etc. to (P-E) represent the regression weights, and figures written
near each of the double headed arrow marks represent co-variances. Factor variances and residual error
variances figures are also written near each explored latent variables and residual variables
respectively.
94 European Journal of Economics, Finance and Administrative Sciences - Issue 24 (2010)
Figure 2: Structural Equation Model of Private Sector Bank Customers

2.25 2.12 1.41 1.86

1.36 .95
e2 1 e3 1 1 e4 e5
e1 1 1 e6
PSQ15 PSQ14 PSQ9 1
PSQ16 1.02 PSQ8 PSQ6 .81
1.04 .98 .87 1
.79 e7
1.85 PSQ5
1 1.00 F1 1.47 1.15
e12 PSQ10
1 1.17
e11 PSQ12 1.28 -.23
2.47 1 .83 .51
e10 PSQ13
1.74 1 1.44 F2 .64
.39
e9 .40
PSQ20 e23
1.00 1.19 .25 1
.37
1
e8 PSQ21 P_E
-.66 .49
1.78 -.11
1 -.12
e16 PSQ1 .11
1.81 -.43
1.64 1 .46
e15 PSQ2 2.01
.90 -.58
.56 1.55 F3 -.12
1
e14 PSQ3
1.37 1.00
1
e13 PSQ4 .04
1.92 .27 .76

1 -.48
e18 PSQ18
1.18 .94 F4
1 2.11
e17 PSQ19 1.00
.85 .68 1
1 PSQ17 e21
e20 PSQ7 1.31 1.00 2.10
3.11 -.67 F5 -.87
F6
1
e19 PSQ11 1.00 -.76 1
3.34 -.41 PSQ22 e22
-.78 2.42

Here it is seen that from two models that, for each factor, one factor loading is fixed to 1.This is
needed to give the latent factor an interpretable scale (Hox J J and Bechger T M). If we do not fix one
factor loading to 1, the scale of the latent factor is undetermined. For each latent factor, we can
estimate the loadings given a fixed variance for the latent factor, which standardizes the scale of the
factor to a Z-score, or we can estimate the factor variance given at least one fixed loading. Since the
loadings are a function of the variance of the latent factor, and the variance of the latent factor is a
function of the loadings, we cannot simultaneously estimate unique values for all of these. Thus, one
solution here is to fix the variance of all factors to 1, and estimate all factor loadings.

7. Model Results and Test of Goodness of Fit

The model was calibrated by using the AMOS software package. Model results are shown in Tables 2
to 5. Specifically, the parameters estimated (regression weights), the standard error (S.E.), the critical
ratio (C.R.), and the level of statistical significance (P) of each variable are reported in Tables 2 and 3
separately both for public banks and private banks respectively. The details of computing degrees of
95 European Journal of Economics, Finance and Administrative Sciences - Issue 24 (2010)

freedom both for public and private banks are shown in Table 4. Some tests on the goodness of fit are
given in Table 5.

Table 2: Regression Weights (Public Sector Banks)

Unstandardized Estimate Standardized Estimate

Estimate S.E. C.R. P-Label Estimate
PSQ21 <-- F1 1.215 0.126 9.679 0.000 PSQ21 <-- F1 0.753
PSQ20 <-- F1 1.360 0.132 10.271 0.000 PSQ20 <-- F1 0.817
PSQ19 <-- F1 1.202 0.127 9.430 0.000 PSQ19 <-- F1 0.727
PSQ18 <-- F1 1.019 0.129 7.918 0.000 PSQ18 <-- F1 0.585
PSQ13 <-- F1 0.751 0.108 6.963 0.000 PSQ13 <-- F1 0.504
PSQ11 <-- F1 1.032 0.127 8.149 0.000 PSQ11 <-- F1 0.606
PSQ10 <-- F1 1.359 0.143 9.471 0.000 PSQ10 <-- F1 0.731
PSQ9 <-- F2 1.000 PSQ9 <-- F2 0.783
PSQ8 <-- F2 0.962 0.077 12.505 0.000 PSQ8 <-- F2 0.766
PSQ5 <-- F2 0.983 0.077 12.764 0.000 PSQ5 <-- F2 0.779
PSQ4 <-- F2 0.546 0.068 8.057 0.000 PSQ4 <-- F2 0.521
PSQ3 <-- F2 0.740 0.073 10.088 0.000 PSQ3 <-- F2 0.637
PSQ2 <-- F2 0.880 0.082 10.757 0.000 PSQ2 <-- F2 0.674
PSQ1 <-- F2 0.999 0.089 11.245 0.000 PSQ1 <-- F2 0.700
P_E <-- F1 0.420 0.063 6.653 0.000 P_E <-- F1 0.433
P_E <-- F4 0.033 0.154 0.212 0.002 P_E <-- F4 0.024
P_E <-- F3 0.142 0.086 1.643 0.001 P_E <-- F3 0.176
P_E <-- F2 0.276 0.124 2.223 0.016 P_E <-- F2 0.335
PSQ14 <-- F3 0.800 0.105 7.616 0.000 PSQ14 <-- F3 0.527
PSQ15 <-- F3 0.896 0.102 8.768 0.000 PSQ15 <-- F3 0.604
PSQ16 <-- F3 1.000 PSQ16 <-- F3 0.790
PSQ17 <-- F3 0.992 0.099 10.055 0.000 PSQ17 <-- F3 0.693
PSQ7 <-- F4 1.000 PSQ7 <-- F4 0.483
PSQ6 <-- F4 1.580 0.270 5.856 0.000 PSQ6 <-- F4 0.794
PSQ12 <-- F1 1.368 0.142 9.644 0.000 PSQ12 <-- F1 0.749
PSQ22 <-- F1 1.000 PSQ22 <-- F1 0.631

Each unstandardized regression coefficient represents the amount of change in the dependent
variable for each one unit change in the variable predicting it. The probability value associated with the
null hypothesis that the coefficient values are equal to zero, is displayed under P-Level column. Both
for public and private banks, all of the regression coefficients in these two models are almost
significantly different from zero at 1% level of significance. Standardized estimates allow evaluating
the relative contributions of each predictor variable to each outcome variable and these are given in
separate columns in both of the tables 2 and 3.
96 European Journal of Economics, Finance and Administrative Sciences - Issue 24 (2010)
Table 3: Regression Weights (Private Sector Banks)

Unstandardized Estimate Standardized Estimate

Estimate S.E. C.R. P-Label Estimate
PSQ21 <-- F2 1.000 PSQ21 <-- F2 0.632
PSQ20 <-- F2 1.436 0.169 8.524 0.000 PSQ20 <-- F2 0.932
PSQ12 <-- F2 1.277 0.187 6.837 0.000 PSQ12 <-- F2 0.663
PSQ10 <-- F2 1.166 0.167 6.997 0.000 PSQ10 <-- F2 0.682
PSQ4 <-- F3 1.000 PSQ4 <-- F3 0.438
PSQ3 <-- F3 1.552 0.329 4.723 0.000 PSQ3 <-- F3 0.667
PSQ2 <-- F3 2.007 0.394 5.094 0.000 PSQ2 <-- F3 0.876
PSQ1 <-- F3 1.808 0.378 4.785 0.000 PSQ1 <-- F3 0.689
PSQ19 <-- F4 1.000 PSQ19 <-- F4 0.843
PSQ18 <-- F4 0.939 0.132 7.105 0.000 PSQ18 <-- F4 0.782
PSQ11 <-- F5 1.000 PSQ11 <-- F5 0.580
PSQ7 <-- F5 -0.670 0.195 -3.435 0.001 PSQ7 <-- F5 -0.40
PSQ17 <-- F6 1.000 PSQ17 <-- F6 0.620
PSQ22 <-- F6 -0.758 0.177 -4.293 0.000 PSQ22 <-- F6 -0.487
P_E <-- F2 0.248 0.117 2.116 0.014 P_E <-- F2 0.289
P_E <-- F3 0.491 0.199 2.462 0.014 P_E <-- F3 0.354
P_E <-- F4 -0.118 0.103 -1.143 0.003 P_E <-- F4 -0.183
PSQ5 <-- F1 1.148 0.121 9.465 0.000 PSQ5 <-- F1 0.839
PSQ14 <-- F1 0.985 0.137 7.200 0.000 PSQ14 <-- F1 0.634
PSQ9 <-- F1 1.022 0.125 8.193 0.000 PSQ9 <-- F1 0.722
PSQ15 <-- F1 1.045 0.143 7.325 0.000 PSQ15 <-- F1 0.645
PSQ8 <-- F1 0.868 0.125 6.944 0.000 PSQ8 <-- F1 0.611
PSQ16 <-- F1 1.000 PSQ16 <-- F1 0.720
P_E <-- F1 0.401 0.117 3.409 0.001 P_E <-- F1 0.519
PSQ6 <-- F1 0.788 0.099 7.954 0.000 PSQ6 <-- F1 0.700
PSQ13 <-- F2 0.828 0.138 5.980 0.000 PSQ13 <-- F2 0.564
P_E <-- F5 0.115 0.067 1.707 0.018 P_E <-- F5 0.145
P_E <-- F6 -0.434 0.234 -1.853 0.014 P_E <-- F6 -0.531

From table 2, it is observed that the degree of customer satisfaction is best explained by first
factor extracted (Responsiveness / Empathy) in public sector banks, whose coefficient has a value of
0.433. But in case of private banks (table 3), the degree of customer satisfaction is best explained by
sixth factor (Assurance / Empathy) with the coefficient value of (-0.531).
The chi-square test of overall model fit is labeled by Discrepancy in this output (Table - 4). Its
value is 1028 for public banks and 802 for private banks with 221 and 210 degrees of freedom
respectively, returning a probability value of 0.000 that a chi-square value this large or larger would be
obtained by chance if the null hypotheses that the models fit the data are true.

Table 4: Computation of Degrees of Freedom

Public Bank Private Bank

Number of distinct sample moments = 276 Number of distinct sample moments = 276
Number of distinct parameters to be estimated = 55 Number of distinct parameters to be estimated = 66
Degrees of freedom = 276 - 55 = 221 Degrees of freedom = 276 - 66 = 210
Discrepancy = 1028 (p = 0.000) Discrepancy = 802 (p = 0.000)

The degrees of freedom represent the level of over identification of the model. In each of the
models, there 23 observed variables, so there are [23(23+1)]/2 =276 available degrees of freedom,
represented by distinct sample moments in table 7.3. In public sector banks there are 55 numbers of
parameters estimated which includes residual variance, factor variance, path coefficients and factor
loadings. In private banks, this number of parameters estimated is 66.
97 European Journal of Economics, Finance and Administrative Sciences - Issue 24 (2010)

Since the probability value of the chi-square test is smaller than the 0.05 level used by
convention, we would reject the null hypothesis that the models fit the data. Therefore in tests of
absolute fit, the models do not fit the data, but it may fit in tests of relative fit.

Table 5: Table 5 Goodness of Fit Indexes

Values
Indexes
Public Sector Private Sector
Chi-square (Discrepancy) 1028 802
Goodness of fit index (GFI) 0.948 0.912
Comparative fit index (CFI) 0.897 0.881
Adjusted goodness of fit index (AGFI) 0.915 0.891
Root mean square residual (RMR) 0.18 0.19
Root mean square error of approximation (RMSEA) 0.073 0.082

Table 5 represents some indicators of test of relative fit of both of the models. From the above
table it is observed that the tests on the goodness of fit are somehow satisfactory. The goodness of fit
index (GFI) is at 0.948, the adjusted goodness of fit index (AGFI) is 0.915, and the comparative fit
index (CFI) is 0.897 for public sector banks and these values are at 0.912, 0.891 and 0.881 respectively
in case of private sector banks. The root mean square residual (RMR) index has a value of 0.18 and
0.19 and the root mean square error of approximation (RMSEA) has a value of 0.073 and 0.082 for
public and private banks respectively. The values of these indexes are comparatively low and therefore
are quite good.

8. Conclusion
Delivering superior service quality appears to be a prerequisite for success of any service firms. As
electronic banking becomes more prevalent, now-a-days customers are evaluating banks based more on
their “high-touch” factors than on their “high-tech” factors in most of the developing economy like
India. The operationalization of customer satisfaction in banking sector is somewhat hazy, and it
should be operationalized along the same dimensions that constitute service quality. In this context,
two proposed structural equation models (SEMs) show the relationship between customer satisfaction
on bank services and the attributes of the perceived service quality. The proposed models identify
service quality attributes to improve, with the aim of offering bank services characterized by higher
levels of quality.

References
1] Adrian, Pyne (1995), The Essence of Service Marketing, Prentice-hall of India, New Delhi,
pp.224-226.
2] Angur, M.G.; Natarajan, R; and Jahera, J.S. (1999); “Service Quality in the Banking Industry:
An assessment in a Developing Economy”, International Journal of Bank Marketing, Vol.17,
No.3, pp.116-123.
3] Bacon L D (1997); “Using Amos for Structural Equation Modeling in Market Research”; SPSS
White Paper; SPSS Inc.
4] Balachandran, M. (2005); “Strategic Model for Re-positioning of PSBs), IBA Bulletin, Vol.
XXVII, No.8 (August), pp.5-8.
5] Buzzell, R D and Gale, B T (1987). The PIMS Principles, The Free Press, New York.
6] Cronin, J and Taylor, S A (1992). “Measuring Service Quality: A Re-examination and
Extension,” Journal of Marketing, 56(July), 55-67.
98 European Journal of Economics, Finance and Administrative Sciences - Issue 24 (2010)

7] Eboli Laura and Mazzulla,Gabriella (2007) ; “Service Quality Attributes affecting Customer
Satisfaction for Bus Transit”; Journal of Public Transportation; Vol. 10; No. 3; pp 21-34
8] Fox John (2002); “Structural Equation Models”; Appendix to An R and S-PLUS Companion to
Applied Regression; pp 1-20
9] Garvin, D A (1983). “Quality on the Line,” Harvard Business Review, 61(Sept.-October), 65-
73.
10] Howcroft, J.B.(1991), “Customer Satisfaction in Retail Banking”, Service Industry Journal,
(Jan);pp.11-17.
11] Hox J J and Bechger T M; “An Introduction to Structural Equation Modeling”; Family Science
Review; Vol. 11; pp 354-373
12] Hsu Sheng-Hsun, Chen Wun-Hwa and Hsieh Ming-Jyh (2006); “Robustness Testing of PLS,
LISREL, EQS and ANN-based SEM for measuring Customer Satisfaction”; Total Quality
Management; Vol. 17; No. 3; pp 355-371
13] Iacobucci Dawn, Ostrom Amy and Grayson Kent (1995); “Distinguishing Service Quality and
Customer Satisfaction: The Voice of the Consumer”; Journal of Consumer Psychology; Vol. 4;
No. 3; pp 277-303
14] Ketokivi Mikko (2006); “Structural Equation Modeling in Management Research”; Class
Meets, HEC School of Management, Paris; pp 1-3
15] Kotler, P. (2003); Marketing Management, Prentice Hall of India, New Delhi.
16] Lee Yonggoo and Kim Sahmyeong (2002); “Measurement of the Customer Satisfaction Index
of the Bank”; Korea
17] Ndubisi Nelson Oly (2006); “A Structural Equation Modeling of the antecedents of
Relationship Quality in the Malaysia Banking Sector”; Journal of Financial Services
Marketing; Vol. 11; No. 2; pp 131-141
18] Parasuraman, A, Berry, L L and Zeithami, V A (1990); “Guidelines for Conducting Service
Quality Decrease,” Marketing Research, 2(4), 34-44.
19] Parasuraman, A, Berry, L L and Zeithaml, V A (1991). “Refinement and Reassessment of the
SERVQUAL Scale,” Journal of Retailing, 67(4), 420-50.
20] Parasuraman, A, Zeithaml, V A and Berry, L L (1985). “A Conceptual Model of Service
Quality and Its Implications for Future Research,” Journal of Marketing, 49 (Fall), 41-50.
21] Parasuraman, A, Zeithaml, V A and Berry, L L (1988). “SERVQUAL: A Multiple Item Scale
for Measuring Consumer Perceptions of Service Quality,” Journal of Retailing, 64(1), 12-40.
22] Parasuraman, A, Zeithaml, V A and Berry, L L (1994). “Reassessment of Expectations as a
Comparison Standard in Measuring Service Quality: Implications for Further Research,”
Journal of Marketing, 58(January), 111-24.
23] Radon, G.H. (1987); “Quality Service – A Low Cost Profit Strategy”, Bank Marketing, Vol.19,
No.9; pp.10-12.
24] Rust, R T and Oliver, R L (1994). Service Quality — New Directions in Theory and Practice,
Sage Publications, New York.
25] Saurina Carmme and Coenders Germa (2002); “Predicting Overall Service Quality: A
Structural Equation Modeling Approach”; Developments in Social science Methodology; pp
217 – 238
26] Sousa, Cristina (1999); “Customer Expectations and Perceptions of Service Quality in Retail
Banking: A Comparative Study of Macau and Guangzhou Banking Consumers”, available in
www.sab.muhio.edu. Pp.1 – 7.
27] Wei Khong Kok (2009); “Service Quality Index: A Study on Malaysian Bank”; Contemporary
Management Research; Vol. 5; No. 2; June, pp 109-124
28] Zeithaml, V A and Bitner, Mary Jo (2003). Services Marketing: Integrating Customer Focus
Across the Firms, 3rd Edition, Tata-MçGraw Hill, New Delhi.
 99 European Journal of Economics, Finance and Administrative Sciences - Issue 24 (2010)

29] Zeithaml, V A and Parasuraman, A (1996). “The Behavioural Consequences of Service

Quality,” Journal of Marketing, 60(April), 31-46.
30] Zeithaml, V A, Parasuraman, A. and Berry, L L (1990). Delivering Service Quality: Balancing
Customer Perceptions and Expectations, The Free Press, New York.
31] Zeithaml, V A, Parasuraman, A. and Berry, L L (1993), “The Nature and Determinants of
Consuner Expectations of Services”, Journal of Academy of Marketing Science, Vol.21, No.1,
pp.1-12

View publication stats