Ms. Devina Upadhyay et al.

2015, Volume 3 Issue 6 International Journal of Science,

ISSN (Online): 2348-4098 Engineering and Technology
ISSN (Print): 2395-4752 An Open Access Journal

Multiple Regression: A Multivariate Technique to

Study Bank Sector
1
Ms. Devina Upadhyay, 2Dr. Chetna Bhavsar

Abstract

A statistical tool that allows you to examine how multiple independent variables are related to
dependent variable. We can predict the value of dependent variable using the equation which shows
relationship between independent and dependent variables. This process is called “Multiple
Regression”. Here in our data set 5 independent variables and 1 dependent variable is there.
Dependent variables: Loyalty

Independent Variables: gender, age, education, income, occupation

To check establish the relationship bet independent and dependent we have used multiple regression
analysis .to reach the result we have used statistical software SPSS and result shows that, in case of SBI
age and income plays important role while in case of ICICI gender plays an important role.

Keywords: Loyalty, independent variables, dependent variables, SPSS

Introduction
Multiple regression is the same scheme as simple
linear regression, except that here we have two or
more than two independent variables. We can
predict dependent variable from several independent
variables, example, imagine that you now wanted to
predict a person's height from the gender of the
person and from the weight. We can use multiple
regression in which gender and weight were the
independent variables and height was the dependent
variable.
Many practical questions involve the relationship
between a dependent or criterion variable of interest
(call it Y) and a set of k independent variables or
potential predictor variables (call them X1, X2,
X3,...,Xk), where the scores on all variables are
measured for N cases. 1For example, you might be
interested in predicting performance on job (Y) using
information on years of experience (X1), performance
in a training program (X2), and performance on an
aptitude test (X3). A multiple regression equation for
predicting Y can be expressed a follows:
1 on customer loyalty. A questionnaire derived from
Research Scholar, Gujarat University, Email:
devinaupadhyay@gmail.com
Y '  A  B1 X 1  B2 X 2  B3 X 3
To apply the equation, each Xj score for an individual
case is multiplied by the corresponding Bj value, the
products are added together, and the constant A is
added to the sum. The result is Y, the predicted Y value
for the case.
Literature survey
Asunción Beerli, Josefa D. Martín, Agustín Quintana,
(2004), have concluded that satisfaction together with
personal switching costs are antecedents leading
directly to customer loyalty, with the former exerting the
greatest influence; and perceived quality is a
consequence of satisfaction on the basis of empirical
research carried out in the retail banking market. At the
same time, the paper shows that the degree of
elaboration in the bank selection process does not have
a moderating influence on the causal relationships
between satisfaction/switching costs and customer
loyalty.
Nelson OlyNdubis i, (2007), purpose of his study was to
examine the impact of relationship marketing strategy
previous studies and the relevant literature was
completed by 220 bank customers in Malaysia. Author

© 2015 Ms. Devina Upadhyay et al. This is an Open Access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the
 
original work is properly credited. Page 20
Devina Upadhyay et al. International Journal of Science, Engineering and Technology, 2015, Volume 3 Issue 6
ISSN (Online): 2348-4098 , ISSN (Print): 2395-4752

has used Multiple regression analysis to assess the Discussion

impact on customer loyalty of four key constructs of
relationship marketing (trust, commitment, Here in this data set 5 independent variables and 1
communication and conflict handling) and the study dependent variable is there.
revealed that four variables have a significant effect and
Dependent variables: Loyalty
predict a good proportion of the variance in customer
loyalty. Moreover, they are significantly related to one Independent Variables: gender, age, education,
another. income, occupation
Yong Hyo Cho (2011), has used regression analysis in his Regression equation,
study to examines whether various measures of public
policies have significant effect in reducing serious crimes Y= a + b1x1+b2x2+b3x3+b4x4+b5x5
and in suppressing the upward trends of crime rates in
50 largest cities of US.

The result of regression model is given in following table:

SBI Bank ICICI Bank

Dependent Independent variable- Loyalty Dependent Independent variable- Loyalty

variables variables

Significant B Significant B
value value

Constant 0.000 3.740 Constant 0.000 3.866

Gender 0.110 0.171 Gender 0.050 0.199

Age 0.026 -0.096 Age 0.637 -0.020

Education 0.278 -0.069 Education 0.699 -0.023

Income 0.001 0.139 Income 0.497 -0.023

Occupation 0.959 0.002 Occupation 0.145 -0.054

The values can be substituted in the equation, statistically significant in the model. The other three
variables are individually not significant. One can use
For SBI, this model as it is for decision making.

Loyalty= 3.740 + 0.171 (gender) - 0.096 (age) - 0.069 There is a negative coefficient, that of age, which can
(education) + 0.139 (income) + 0.002 (occupation) be interpreted to mean that if we increase the age of
the customer, the loyalty will decrease.
Statistical significance of the model:
Either forward stepwise regression or backward
Result of ANOVA shows that P value is 0.014. This
stepwise regression can be used to eliminate
indicates that the model is statistically significant at a
insignificant variables from full regression model
confidence level of (1-0.014)*100 i.e. 98.6. The p
containing all the five independent variables.
indicates the significance of the ANOVA.
For ICICI,
The R square value is 0.07. From the result we can
say that only two variables age and income are

Page 21
Devina Upadhyay et al. International Journal of Science, Engineering and Technology, 2015, Volume 3 Issue 6
ISSN (Online): 2348-4098 , ISSN (Print): 2395-4752

Result of ANOVA shows that P value is 0.22. This
indicates that the model is statistically significant at a
confidence level of (1-0.22)*100 i.e.0.78. The p
indicates the significance of the ANOVA.
The R square value is 0.188. From the result we can
say that only one variable gender is statistically
significant in the model. The other four variables are
individually not significant. One can use this model
as it is for decision making.
Either forward stepwise regression or backward
stepwise regression can be used to eliminate
insignificant variables from full regression model
containing all the five independent variables.

Backward stepwise regression:

SBI Bank ICICI Bank

Independent Dependent variable- Loyalty Independent Dependent variable- Loyalty

variables variables

Significant B Significant B
value value

Constant 0.000 3.622 Constant 0.000 3.467

Age 0.040 -0.087 Gender 0.046 0.201

Income 0.001 0.117

For SBI bank result of multiple regression suggest http://ssrn.com/abstract=2259524 or http://dx.doi.or

the following equation: g/10.2139/ssrn.2259524

Loyalty = 3.622 -0.087(Age) + 0.117(Income)
ICICI Bank,
Loyalty = 3.467+0.201(Gender)
Here results suggest that loyalty differs age wise and
income wise in case of SBI bank while in case of ICICI
bank loyalty differs gender wise.
References
Cooper, D. & Schindler, P. (2008), Business Research
Methods, 10th edn, Boston: McGraw- Hill
International Edition.
Hair J.F, Black W.C., Babin B.J., Anderson R.E. and
Tatham R.L. (2006), Multivariate Data Analysis (6th
ed.), Upper Sadle River, New Jersey: Prentice Hall
Asunción Beerli, Josefa D. Martín, Agustín Quintana,
(2004) "A model of customer loyalty in the retail
banking market", European Journal of Marketing,
Vol. 38 Iss: 1/2, pp.253 – 275
Nelson OlyNdubisi, (2007) "Relationship marketing
and customer loyalty", Marketing Intelligence &
Planning, Vol. 25 Iss: 1, pp.98 – 106
Yong Hyo Cho, (2011), A Multiple Regression Model
for the Measurement of the Public Policy Impact on
Big City Crime Policy Sciences, Vol. 3, No. 4 (Dec.,
1972), pp. 435-455
http://www.jstor.org/stable/4531504
Author’s Affiliation
1
Research Scholar, Gujarat University, India

Mishra, Debi P. and Min, Junhong, Analyzing the 2

Associate Professor, Gujarat University, India
Relationship between Dependent and Independent
Variables in Marketing: A Comparison of Multiple Copy for Cite this Article- Ms. Devina Upadhyay and Dr.
Regression with Path Analysis (September 2010). Chetna Bhavsar, ‘Multiple Regression: A Multivariate
Technique to Study Bank Sector’, International Journal of
Available at SSRN: Science, Engineering and Technology, Volume 3 Issue 6: 2015,
pp. 20-22.

Page 22