09 - Chapter 4

Chapter – 4
Data Analysis
&
Interpretation
92
Data Analysis & Interpretation
A. Data Analysis for Customer Questionnaire
4.1. Detail of Demographic Profile of Respondents
4.1.1. Gender-wise Frequency Distribution
The majority of male respondent 229 total frequencies i.e. 55.2% of total population is greater
than female respondent i.e. 186 total frequencies i.e. 44.8%. it indicate that in this
study male participate better than female.
Table No.4.1. Gender-wise Frequency Distribution

Frequency Percent
Male 229 55.2
Female 186 44.8
Valid
Total 415 100.0
4.1.2. Age Groups-wise Frequency Distribution
Frequency distribute according to age groups of respondent, and found 31-45 years
old respondent participate in this study and they were interested to obtain personal
loan from banks they have total frequency 236 i.e. 56.9%, secondly age groups 16-30
years old respondent also have same condition they have total frequency 93 i.e. 22.4%
of total sample. Rest of age groups respondent are not participating well they have 70
(16.9%) for 46-60 years old, 10 frequency (2.4%) for 61-75% and 6 frequency for 76
years above respondent (1.4%) respectively. Indicate that young populations were
understanding importance of such kind of research and they participate great than
other.
93
Table No. 4.2. Age Groups-wise Frequency Distribution
Frequency Percent Valid Percent
Cumulative Percent
16-30 93 22.4
31-45 236 56.9
46-60 70 16.9
61-75 10 2.4
Valid
76 Above 6 1.4
Total 415 100.0
4.1.3. Residential Area-wise frequency Distribution
According to residential area of bank customer-wise frequency distribution, in below

table clearly saws urban people plat great role in this study than rural respondent, they
have 276 total frequencies i.e. 66.5% for urban population and 139 total frequencies
i.e. 33.5% for rural population. Indicate that urban people more interested to take loan
from banks than rural people.
Table No. 4.3. Residential Area

Frequency Percent Pie Chart
Urban 276 66.5
Rural 139 33.5
Valid
Total 415 100.0
4.1.4. Education Level-wise frequency Distribution
According to Education level the bank customer graduate and higher secondary
passed respondent were participate well in this research they have 139 total frequency
i.e. 33.5% for Graduate and 119 total frequency i.e. 28.7% for higher educated people,
94
the customer of bank have these kind of educate they willing to take loan from bank
and they interested to deal with bank. Other hand illiterate people participate in this
study great they have 91 total frequencies 21.9% (with the help of translator and
writer), rest of educated respondent like high school respondent, post graduate and
others eructation have less frequency i.e. 37 (8.9%) for up to high school, 10
frequency (2.4%) for post graduate and last 19 (4.6%) for other education customer.
Table No.4.4. Education Level

Frequency Percent Pie Chart
Illiterate 91 21.9
Up to High
37 8.9
School
Higher
119 28.7
Secondary
Valid
Graduation 139 33.5
Post
10 2.4
Graduation
Others 19 4.6
Total 415 100.0
4.1.5. Income Monthly-wise frequency Distribution
According to monthly income-wise frequency distribution we found the income

between 0-10000 salaried persons interested in such kind of research study and
participate greatly than other higher income group salaried parson i.e. 240 total
frequency (57.8%) for 0-10000, 66 total frequency (15.9%) for 10001-20000, 37 total
frequency (8.9%) for 20001-30000, 24 total frequency (5.8%) for all other three
category 30001-40000, 40001-50000 and 50001 above. Indicate that low income
group persons interested to take loan from bank.
95
Table No. 4.5. Income Monthly
Cumulative Percent
0-10000 240 57.8
10001-
66 15.9
20000
20001-
24 5.8
30000
30001-
Valid 24 5.8
40000
40001-
24 5.8
50000
50000 &
37 8.9
Above
Total 415 100.0
4.1.6. Occupation-wise frequency Distribution
According occupation of respondent frequency was distributed in three categories i.e. Job,
Business and Other profession and found businessmen have participate greater than
other, they have 199 total frequencies i.e. 48.0%, secondly salaried personal of an
employees who engage in job have 157 total frequency i.e. 37.8% lastly other
profession have less frequency 59 i.e. 14.2%, indicate that businessmen were
participated well in that kind of study and they also interested to take loan than
salaried persons.
96
Table No. 4.6. Occupation
Cumulative Percent
Job 157 37.8
Business 199 48.0
Other 59 14.2
Valid
Total 415 100.0
4.1.7. Religion-wise frequency Distribution
According to religion wise frequency distribution it found Hindu participate well in this study
they have 182 total frequencies i.e. 43.9%, rest of religion are Sikh have 78 of total
frequencies i.e. 18.8%, other religion have 68 total frequencies i.e. 16.4%, Muslim religion
have 51 total frequencies i.e. 12.3% last not the least Christian have 36 total frequency i.e.
8.7%. indicate that India Hindu play greater than other obviously.
Table No.4.7. Religion
Cumulative Percent
Christian 36 8.7
Muslim 51 12.3
Sikh 78 18.8
Valid
Hindu 182 43.9
Other 68 16.4
Total 415 100.0
4.1.8. Marital Status-wise frequency Distribution
Married populations participate greater than unmarried clearly indicate in below table,
here the frequencies of married was 276 i.e. 66.5% of total population and the
frequencies of unmarried are 33.5% of total population. Indicate that married
respondent willing to take loan from banks.
97
Table No. 4.8. Marital Status
Cumulative Percent
Married
276 66.5
Unmarried
Valid 139 33.5
Total 415 100.0
4.2. Consistency & Reliability Measurements

4.2.1. Consistency Measurements
Consistency of all the statements in the questionnaires was verified through item total
correlation. Under this methods of every item with total was measured and the
computed value was compared with standard value. The all statements having item to
total correlation lower than the critical value, were declared as inconsistent and
dropped from the questionnaire.
Table No. 4.9. Item-Total Statistics

Corrected Cronbach's Status
Item-Total Alpha if
Correlation Item Deleted
It easy to process to obtain personal loans
.525 .927 Required
through banks
Easy to understand procedures of loan .497 .927 Required
Documentation is relevant under the
.584 .926 Required
schemes
Lower interest rate influence you to obtain
.491 .927 Required
personal loan
Beneficiary schemes motivated to obtain
.509 .927 Required
loan
Family needs is greater purpose to taking
.554 .926 Required
loan
98
Employee Empowerment of Banks solve
.461 .928 Required
your problem
Nearby location banks is only reason to
.447 .928 Required
taking loan
Banks brochure offering personal loan .506 .927 Required
Agent are helpful to obtain loan from
.553 .926 Required
beginning
Television or advertising offering to get
.456 .928 Required
personal loan
It is an essential when you need of money .556 .926 Required
File charges and other processing fee low .577 .926 Required
Banks give you relevant advice about
.403 .928 Required
personal loan
Bank employee pay 100 percent attention
.604 .926 Required
towards you
Well dress up employee attract you .586 .926 Required
It remain information of your loan
.545 .927 Required
confidential
Cleanliness of banks make you happy .567 .926 Required
Other supporting staffs helpful in nature .579 .926 Required
Toiletry facility of banks subsidiary reason
.542 .927 Required
for taking loan
Drinking water etc. make you happy .519 .927 Required
They solve your problem internally .418 .928 Required
Banks employees are trustworthy about loan .558 .926 Required
Networking of banks affects you .440 .928 Required
Continues service attract you for taking loan .495 .927 Required
Modernization of banks gives you well
.542 .927 Required
performance
Online banking with your loan schemes .463 .928 Required
Safety of money .535 .927 Required
Wide range of products and services .480 .927 Required
Location of banks .520 .927 Required
Feature of loan .569 .926 Required
Contents of loan .543 .927 Required
Cronbach Alpha had been obtained 0.929 after reliability test through SPSS
which is more than the cut off value (0.7). In above table of consistency measures
mainly two things are considered,
99
1. Corrected Item to Total Correlation values which is acceptable if greater
than.
2. Secondly Cronbach's Alpha if Item Deleted value for each item is evaluated
and if found value greater than the calculated reliability value (0.929) than that
item should be dropped from the questionnaire and not considered for further
study.
So that from above table of consistency of all statements having item to total
correlation lower than the critical value. Obtain results indicate that there is no need to
drop any statement in questionnaire because all statements.
4.2.2. Reliability Test
Cronbach‟s Alpha test had been applied to calculate reliability of all items of Loan
Schemes of bank from the questionnaire. It is considered that the reliability value
more than 0.7 is good and it can be seen that in almost all the reliability values is quite
higher than the standard value.
The Cronbach‟s alpha reliability method was applied to check the reliability of all the
items of the questionnaire. The reliability test measures are as shown below:
Table No. 4.10. Reliability Statistics

Test Variables Cronbach's N of Test Results
Alpha Items
Loan Schemes of Data are Highly
.929 32
Bank Reliability
The reliability coefficient value of Loan Schemes of Bank were highly significant i.e.
0.929 that shows high reliability of the questionnaire, so it can be considered good
enough for the present study, the total number of statement have 32 for this study.
4.3. Normality Analysis
There are two main methods of assessing normality: graphically and numerically. The
table below presents the results from two well-known tests of normality, namely the
Kolmogorov-Smirnov Test and the Shapiro-Wilk Test. The Shapiro-Wilk Test is more
100
appropriate for small sample sizes (< 50 samples), but can also handle sample sizes as
large as 2000. Due to this consideration, Shapiro-Wilk test is used to assessment of
normality as numerical means of assessing normality is done in this study.
Table No. 4.11. Tests of Normality

Kolmogorov- Shapiro-Wilk Test results
a
Smirnov
Statistic df Sig. Statistic df Sig. Data is Normally
Loan Schemes Distributed
.041 415 .096 .988 415 .002
of Bank
a. Lilliefors Significance Correction
Above table shows that data is normally distributed with insignificant level of .096. It
is because if the significance value of the Shapiro-Wilk Test is greater than 0.02, the
data is normal. If it is below 0.05 (P<0.05), the data significantly deviate from a
normal distribution.
4.4. Factor Analysis
Factor analysis is a statistical method used to describe variability among observed,

correlated variables in terms of a potentially lower number of unobserved variables
called factors. Factor analysis originated in psychometrics, and is used in behavioural
sciences, social sciences, marketing, product management, operations research, and
other applied sciences that deal with large quantities of data.
Factor analysis searches for such joint variations in response to unobserved latent
variables. The observed variables are modelled as linear combinations of the potential
factors. The information gained about the interdependencies between observed
variables can be used later to reduce the set of variables in a dataset.
Correlation matrix is the starting point of factor analysis. In this analysis inter
correlations between the studied variables are presented. The dimensionality of this
matrix can be reduced by looking for variables that correlate highly with a group of
other variables, but correlate very badly with variables outside of that group (Field;
2000). Then after these variables with high inter correlations could well measure one
underlying variable, which is called a „factor‟. The obtained factor creates a new
dimension which can be visualized as classification axes along with measurement
101
variables can be plotted (Field; 2000). This projection of the scores of the original
variables on the factor leads to two results: factor scores and factor loadings.
Exploratory factor analysis (EFA) is used to identify complex interrelationships

among items and group items that are part of unified concepts. The researcher makes
no "a priori" assumptions about relationships among factors.
4.4.1. K. M. O. measurement of sampling Adequacy and Bartlett’s Test of

Sphericity for Validity Analysis of Loan Schemes of Bank
Table No. 4.12. KMO and Bartlett's Test
Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .891
Approx. Chi-Square 7046.320
Bartlett's Test of Sphericity df 496
Sig. .000
Test results This data is Adequate for Factor Analysis
A KMO measure of sampling reflected KMO value of 0.891 which indicated that the
size of sample is adequate for the present study. The Bartlett‟s test measuring Chi-
Square value 7046.320 which is significant at 0.000% level of significance, indicating
that null hypothesis is rejected. Hence, it shows that item matrix is not an identity
matrix and the value revealed that the data is normally distributed and suitable enough
for factor analysis.
4.4.2. Description of Factors
Principle component factor analysis with Varimax rotation was applied. The raw
scores of 32 items were subjected to factor analysis to find out the factors that
contribute towards Loan Schemes of Bank. After factor analysis 6 factors were
identified. The rotation converged in 7 iterations.
102
Table No. 4.13. Factor Description
Variable Eigen Variance Statement Loading
Value Value
Service Employee Empowerment of Banks solve your
Quality .653
problem
Wide range of products and services .616
Television or advertising offering to get
.564
personal loan
10.103 31.571 Banks brochure offering personal loan .561
Agent are helpful to obtain loan from beginning .555
Documentation is relevant under the schemes .520
Nearby location banks is only reason to taking
.475
loan
Lower interest rate influence you to obtain
.437
personal loan
Employees Networking of banks affects you .712
Commitment
Continues service attract you for taking loan .617
Other supporting staffs helpful in nature .569
It easy to process to obtain personal loans
.542
through banks
2.192 6.851
They solve your problem internally .533
Banks employees are trustworthy about loan .531
Drinking water etc. make you happy .524
Family needs is greater purpose to taking loan .518
Modernization of banks gives you well
.405
performance
Safety & Banks give you relevant advice about personal
Security .701
loan
It is an essential when you need of money .648
Well dress up employee attract you .601
1.644 5.137
File charges and other processing fee low .525
Easy to understand procedures of loan .484
Safety of money .481
Bank employee pay 100 percent attention
.474
towards you
Ease Process Feature of loan .822
1.478 4.620
It remain information of your loan confidential .821
Online banking with your loan schemes .392
Sanitation Toiletry facility of banks subsidiary reason for
Facility .853
taking loan
1.319 4.121
Location of banks .845
Cleanliness of banks make you happy .493
103
Promotion Contents of loan .886
through Ads
1.065 3.329 Beneficiary schemes motivated to obtain loan
.880
by ads
Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization.
a. Rotation converged in 7 iterations.
i. Factor One: Service Quality

The total Eigen value of this factor is 10.103 and this factor has emerged as the most
important determinant of Loan Schemes of Banks and explains 31.571 % of variance
of Loan Schemes of Banks. The important statement constituting this factor include;
Employee Empowerment of Banks solve your problem (.653), Wide range of products
and services (.616), Television or advertising offering to get personal loan (.564),
Banks brochure offering personal loan (.561), Agent are helpful to obtain loan from
beginning (.555), Documentation is relevant under the schemes (.520), Nearby
location banks is only reason to taking loan (.475) and Lower interest rate influence
you to obtain personal loan (.437).
ii. Factor Two: Employees Commitment

important determinant of Loan Schemes of Banks and explains 6.851 % of variance of
Loan Schemes of Banks. The important statement constituting this factor include;
Networking of banks affects you (.712), Continues service attract you for taking loan
(.617), Other supporting staffs helpful in nature (.569), It easy to process to obtain
personal loans through banks (.542), They solve your problem internally (.533),
Banks employees are trustworthy about loan (.531), Drinking water etc. make you
happy (.524), Family needs is greater purpose to taking loan (.518) and Modernization
of banks gives you well performance (.405).
iii. Factor Three: Safety and Security

Banks give you relevant advice about personal loan (.701), It is an essential when you
need of money (.648), Well dress up employee attract you (.601), File charges and
104
other processing fee low (.525), Easy to understand procedures of loan (.484), Safety
of money (.481) and Bank employee pay 100 percent attention towards you (.474).
iv. Factor Four: Ease Process

Feature of loan (.822), Cleanliness of banks make you happy (.821) and Online
banking with your loan schemes (.392).
v. Factor Five: Sanitation Facility

Toiletry facility of banks subsidiary reason for taking loan (.853), Contents of loan
(.845) and Beneficiary schemes motivated to obtain loan (.493).
vi. Factor Six: Promotion through Advertisements (Ads)

Loan Schemes of Banks. The important statements constituting these factors include;
Location of banks (.886) and it remain information of your loan confidential (.880).
105
1
Service
Quality
2
6
Employee
Promotion 10.103 Commitment
through Ads
1.065 2.192
Loan
Schemes of
Banks
1.319 1.644
5 3
Sanitation Safety and
Facility 1.478 Security
4
Ease
Process
Fig. 02. Factor Model of Loan Schemes of Bank
4.4.3. Scree Plot

Both the Eigen value rule and the scree plot are accurate if the sample is >250 and
the commonalities are greater than 0.6. Scree plot test is another popular test for the
number of factors is Catell‟s scree criterion (Pallant, 2005). This criterion involves
plotting the Eigen values of the factors, and looking for the point at which the plot
begins to level off from vertical to horizontal. In this case, similar result is suggested
by the scree plot, which would suggest the retention of seven significant factors
(Hair et al., 2003). The results of Catell‟s scree plot are shown in Figure below.
106
Fig. 03: Scree Plot of Loan Schemes of Bank
4.4.4. Component Score Covariance Matrix
The result of the analysis was a rotated component matrix consisting of six
components that account for 55.63%of the variance. The breaking point of the
Scree-plot, however, was at 6 or 7 factors; extraction of only 6 factors would
account, but enhances the overview of the rotated component matrix
considerably.
Therefore, second time factor analysis was carried out, choosing for the
extraction of only 6 factors with factor extraction via Eigen values >1 and then
via the extraction of 6 factors, as factors would be correlated. This has done to
check the correlation among the various factors obtained from factor analysis.
Through this analysis association of each factor can be checked and finally
107
regression analysis can be applied among six factors so that a Loan Schemes of
Banks model can be developed.
Table No. 4.14. Component Transformation Matrix

Component Service Employees Safety & Ease Sanitation Promotion
Quality Commitment Security Process Facility through Ads
Service Quality .515 .480 .463 .322 .326 .282

Employees
.350 .463 .013 -.565 -.226 -.541
Commitment
Safety & Security -.715 .476 .149 -.212 .441 -.015
Ease Process -.134 .309 .018 -.210 -.635 .663
Sanitation Facility .154 .348 -.874 .122 .232 .153
Promotion through
Ads .242 -.334 .000 -.688 .437 .406

In the view, the rotation with six extracted factors gives the best possibility to
interpret the factor solution
4.5. Structure Equation Modeling
SEM can best be defined as a class of methodology that seeks to represent hypotheses
about the means, variances, and covariance of observed data in terms of smaller
number of structural parameters defined by a hypothesized underlying model (Kaplan,
2000)1. In other words, SEM is a comprehensive statistical approach to testing
hypotheses about relations among observed and latent variable (Hoyle, 1995) and one
of the more popular statistical methodologies available to quantitative researchers2.
SEM begins with the specification of a model to be estimated. A model is a statistic

statement about the relations among variables. The specification is the exercise of
formally stating a model and it is central to the SEM approach. Indeed, no analysis
can take place until the researcher has specified a model of the relations among the
variables to be analyzed (Hoyle, 1995). In SEM model specification involves
formulating a statement about a set of parameters. Parameters typically are specified
as either fixed or free. Fixed parameters are not estimated from the data and their
value that the investigator believes to be nonzero (Hoyle, 1995). There are two
1
Kaplan D 2000 Structural Equation Modeling: Foundations and Extensions. Sage, Newbury Park, CA
2
Hoyle, R.H. (1995) (ed.). Structural equation modeling: concept, issue and applications. Newbury
Park CA
108
components of the general structural equation model: the measurement model is that
components of the general model in which latent variables are prescribed. The
structural model is that component of the general model that prescribes relations
between latent variables and observed variables that are not indicators of latent
variables (Hoyle, 1995). A fundamental consideration when specifying models in
SEM is identification. Identification concerns whether a single unique value for each
and every free parameter can be obtained from the observed data3 (Hoyle, 1995).
Once a model has been specified, the next task is to obtain estimates of the free
parameters from a set of observed data. Although single-stage least square methods
such as those used in standard ANOVA or multiple regressions can be used to derive
parameter estimates, iterative methods such as maximum likelihood or generalized
least square are preferred.
The index of fit provides a perspective on the fit of structural equation models. The
basic idea behind indices is that the fit of the model is compared to the fit of some
baseline model that usually specifies complete independence among the observed
variables. Some of these indices include normed fit index (NFI), Tucker-Lews Index
(TLI), and comparative fit Index (CFI). According to Kaplan, (2000), these indices
are typically scaled to lie between zero and one, with one representing perfect fit
relative to the baseline model. He noted that 0.95 is indicative of good fit relative to
the baseline model. In other words, the value of 0.95 is considered evidence that the
target model fit is a god fit to the data relative to the baseline model4.
4.5.1. Goodness of Indices
Goodness of fit indices (or fit indices) indices the goodness of fit between the
hypothesized model and the observed data, in this sub-chapter, the most commonly
used goodness of fit indices, and then cut off (suggested) value of those indices are
presented.
4.5.2. Chi-Square
Chi-square (χ2) is a traditional measure of overall model fit (Howell, 1997), Chi-
square test the validity of the specifications of factor loading, factor covariance, and
3
Ibid
4
109
error variances for the studied model5 (Byrne, 2010, P 79). The Chi-square statistics is
associated with a probability. Low probability indicates a poor fit of the model
(Byrne, 2001, p 80), for a good model fit, the probability should be no significant, that
is greater than .05 (Hatcher, 1994, p. 339). There is also a guideline for the ratio of
chi-square and degree of freedom (df)6. According to hatcher (1994, p 339), the chi-
square df ratio should be at least 2. The use of chi-square has major drawback; for
example, with larger sample sizes the chi-square can reject a valid7 (Bentler and
Bonnet, 1980; Kline, 1998).
SEM was used in our data analysis for various reasons. First, there was the need to
conduct a series of confirmatory factor analysis (CFA) for Organizational Level
variables data and respondents‟ behavioral data ensure model fit.
4.5.3. Confirmatory Factor Analysis of Corporate Social Responsibility

Effectiveness
Confirmatory factor analysis (CFA) was used to help the selection process of the scale
items. The use of CFA requires knowledge of the underlying latent variable structure
(Byrne, 2001, p6). A model or relationship of the observed variables must be
specified before the factor analysis, whose task then is to test goodness of fit between
the model and observed data (Byrne, 2001, p.6; Hatcher, 1994, p.289). in essence,
CFA is a way to test how measured variables represent a smaller number of constructs
(Hair et al., 2010 p.693). The analysis helps to identify the factor loading of individual
items. Cross loadings can be studied with help of CFA. The analysis also helps to
define the optional number of items.
Model fits was evaluated by using the normed chi-square, Tucker-Lewis Index (TLI),
Comparative Fit Index (CFI) and Root-Mean-Square-Error of Approximation
(RMSEA). The first indicated a poor fit of the model to the data. There model
improvement techniques were applied (Anderson & Gerbing, 1998), and factor
loading and significance levels were examined.
5
Byrne, B. M. (2010). Structural equation modeling with AMOS: Basic concepts, applications, and
programming (2nd ed.). Mahwah, NJ: Erlbaum
6
Hatcher, L. (1994). A step-by-step approach to using the SAS® System for factor analysis and
structural equation modeling Cary, NC: SAS Institute Inc.
7
Bentler, P. M. (2005). EQS 6 structural equations program manual. Encino, CA: Multivariate
Software.
110
A χ2 statistic is listed in the column between the tools and the path diagram. The value
of χ2was 109.43 and degree of freedom (df) was found 78 at significant probability
level of 0.00, such insignificant results was possible due to the large sample size. The
value of other estimates were found as below mentioned
111
Fig. No. 04: CFA (Confirmatory Factors Analysis)
112
Table No. 4.15. Confirmatory Factor Analysis
Model Estimate Obtain Value Standard Value

Chi-Square 109.43
Df 78
P-Value 0.00000 >.05
RMSEA 0.297 <.08
NFI 0.903 >.90
IFI 0.932 >.90
CFI 0.899 >.90
RMR 0.375 <.08
GFI 0.897 >.90
AGFI 0.941 >.90
TLI 0.924 >.90
PNFI 0.521 >.50
PGFI 0.496 >.50
Here values which indicate satisfactory adjustment for a model Loan Schemes of
Banks (CMIN/DF), values 1.0 or 3.0 or, at most up to 5.0; for CFI values similar than
.90 and for RMSEA, value lower than .06 or up to .08 from above table, it was
evident that‟s of the value of Normed Fit Index (NFI) was found .90, comparative fit
index (CFI) was found .899, IFI was found .932, goodness of fit index (GFI) was
found .897, Adjusted goodness of the fit index (AGFI) was found .941, Tucker-Lewis
index (TLI) was found .924, root mean square of residual and root means squares
residual (RMR) was found .375 and the value for root mean square error of
approximation (RMSEA) was found .297.
All value were under the satisfactory limit and shown a good model fit while the
value of parsimonious Normed fit index (PNFI) and Parsimonious goodness of fit
index (PCFI) were found high than the cutoff value which was possible because of
large sample size.
113
4.6. Correlation between Factors of Loan Schemes of Bank
The correlation was calculated by taking the total of obtain factors of Loan Schemes
of Banks by using SPSS software. The main purpose of correlation analysis is to find
whether all factors have positive correlation one and another.
Table No. 4.16: Correlation between Obtain Factors of Bank
Service Employee Safety Ease Sanitation Promotion

Quality Commitment Security Process Facility Advertising
Service Pearson 1 .636** .623** .501** .480** .390**
Quality Correlation
Sig. (2- .000 .000 .000 .000 .000
tailed)
N 415 415 415 415 415 415
** ** ** **
Employee Pearson .636 1 .653 .426 .496 .369**
Commitment Correlation
Sig. (2- .000 .000 .000 .000 .000
tailed)
N 415 415 415 415 415 415
** ** ** **
Safety Pearson .623 .653 1 .527 .509 .445**
Security Correlation
Sig. (2- .000 .000 .000 .000 .000
tailed)
N 415 415 415 415 415 415
** ** ** **
Ease Process Pearson .501 .426 .527 1 .479 .485**
Correlation
Sig. (2- .000 .000 .000 .000 .000
tailed)
N 415 415 415 415 415 415
** ** ** **
Sanitation Pearson .480 .496 .509 .479 1 .360**
Facility Correlation
Sig. (2- .000 .000 .000 .000 .000
tailed)
N 415 415 415 415 415 415
** ** ** ** **
Promotion Pearson .390 .369 .445 .485 .360 1
Advertising Correlation
Sig. (2- .000 .000 .000 .000 .000
tailed)
N 415 415 415 415 415 415
**. Correlation is significant at the 0.01 level (2-tailed).
Hence the correlation clearly indicates that there are Service Quality, Employee
Commitment, Safety Security, Ease Process, Sanitation Facility and Promotion Advertising
have positive correlation between each other.
114
4.7. t-Test between Perception of Demographic Variables and Loan
Schemes of Banks by Customer
4.7.1. Independent Samples Test between Perception Gender and Loan Schemes
of Banks by Customer
T-test was applied to find out difference between perception of male and female
gender for Loan Schemes of Banks.
H01. There is no relationship between male and female bank customer on Loan
Schemes of Banks.
Table No. 4.17. Group Statistics

Gender N Mean Std. Deviation Std. Error Mean
Male 229 115.1878 20.49186 1.35414
Loan Schemes of Bank
Female 186 113.6505 20.25463 1.48514
Group statistics table explore the mean value of male individuals is 115.1878, for std
deviation is 20.49186 and for std error mean is 1.35414 whereas for the mean value of
Female individuals is 113.6505, for std deviation is 20.25463 and for std error mean is
1.48514.
Table No. 4.18. Independent Samples Test
Levene's t-test for Equality of Means
Test for
Equality
of
Variances
F Sig. t df Sig. Mean Std. Error 95% Confidence
(2- Differenc Differenc Interval of the
tailed e e Difference
) Lower Upper
Equal
variances .066 .798 .764 413 .445 1.53724 2.01224 -2.41828 5.49275
Loan Schemes assumed
of Bank Equal
variances not .765 397.530 .445 1.53724 2.00981 -2.41395 5.48842
assumed
Above table of Independent Samples Test shows Leven‟s test for equality of variances
to check whether equal variances were assumed or equal variance were not assumed.
115
Value of F is .066 (<.5) with significant level of .798 which means F value is not
significant hence Equal variances assumed. Value of T is .798 which is insignificant
at .445 levels. It means no significant difference between perception of Male and
Female individuals on Loan Schemes of Bank. Hence the null hypothesis was
accepted by the results.
4.7.2. Independent Samples Test between Perception Marital Status and Loan
T-test was applied to find out difference between perception of married and unmarried
for Loan Schemes of Banks.
H02. There is no relationship between married and unmarried bank customer on Loan
Schemes of Banks.
Marital N Mean Std. Std. Error
Status Deviation Mean
Loan Schemes of Married 276 115.4275 20.93432 1.26010
Bank Unmarried 139 112.6547 19.15827 1.62498
Group statistics table explore the mean value of married individuals is 115.4275, for
std deviation is 20.93432 and for std error mean is 1.26010 whereas for the mean
value of unmarried individuals is 112.6547, for std deviation is 19.15827and for std
error mean is 1.62498.
Test for
Equality of
Variances
F Sig. t df Sig. Mean Std. 95% Confidence
(2- Differen Error Interval of the
taile ce Differe Difference
d) nce Lower Upper
Equal
variances 1.719 .191 1.310 413 .191 2.77286 2.11738 -1.38933 6.93505
of Bank Equal
variances not 1.348 299.517 .179 2.77286 2.05631 -1.27379 6.81951
assumed
116
Value of F is .1.719 significant level of .191 (P>0.05) with which means F value is not
significant hence Equal variances assumed. Value of T is 1.310 which is insignificant at .191
levels. It means no significant difference between married and unmarried individuals on Loan
Schemes of Bank. Hence the null hypothesis was accepted by the results.
4.7.3. Independent Samples Test between Perception Residential Area of People

and Loan Schemes of Banks by Customer
T-test was applied to find out difference between perception of Rural and Urban bank
customers for Loan Schemes of Banks.
H03. There is no relationship between Rural and Urban bank customers on Loan
Schemes of Banks.

Residential N Mean Std. Std. Error
Area Deviation Mean
Loan Schemes of Urban 276 114.0109 19.74832 1.18871
Bank Rural 139 115.4676 21.60827 1.83279
Group statistics table explore the mean value of Urban individuals is 114.0109, for std
deviation is 19.74832and for std error mean is 1.18871whereas for the mean value of
Rural individuals are 115.4676, for std deviation is 21.60827 and for std error mean is
1.83279.
117

Test for
Equality of
Variances
F Sig. t df Sig. Mean Std. 95% Confidence
(2- Differen Error Interval of the
tailed ce Differe Difference
) nce Lower Upper
Equal
variances 2.636 .105 -.687 413 .492 -1.45676 2.12056 -5.62520 2.71169
of Bank Equal
variances not -.667 255.805 .505 -1.45676 2.18452 -5.75870 2.84519
assumed
Value of F is 2.636 significant level of .105 (P>0.05) with which means F value is not
significant hence Equal variances assumed. Value of T is -.687 which is insignificant at .492
levels. It means no significant difference between Urban and Rural individuals on Loan
Schemes of Bank. Hence the null hypothesis was accepted by the results.
4.7.4. Independent Samples Test between Perception of Family and Loan

T-test was applied to find out difference between perception of Single Family and
Joint family bank customers for Loan Schemes of Banks.
H04. There is no relationship between Single Family and Joint family bank customers
on Loan Schemes of Banks.
Family N Mean Std. Std. Error
Types Deviation Mean
Loan Schemes of Single 311 114.7170 20.26765 1.14927
Bank Joint 104 113.8462 20.78076 2.03772
118
Group statistics table explore the mean value of Single family type of individuals is
114.7170, for std deviation is 20.26765 and for std error mean is 1.14927 whereas for
the mean value of joint type of individuals is 113.8462, for std deviation is 20.78076
and for std error mean is 2.03772.
Test for
Equality of
Variances
(2- Differen Differenc Interval of the
tailed ce e Difference
) Lower Upper
Equal
variances .021 .884 .377 413 .706 .87089 2.31041 -3.67075 5.41253
of Bank Equal
variances not .372 173.129 .710 .87089 2.33947 -3.74667 5.48845
assumed
Value of F is .021 significant level of .884 (P>0.05) with which means F value is not
significant hence Equal variances assumed. Value of T is .377 which is insignificant at .706
levels. It means no significant difference between Single and Joint family types of individuals
on Loan Schemes of Bank. Hence the null hypothesis was accepted by the results.
4.7.5. Independent Samples Test between Perception of Head of Family and

Loan Schemes of Banks by Customer
T-test was applied to find out difference between perception of Self head and Family
Member Head bank customers for Loan Schemes of Banks.
H05. There is no relationship between Single Self head and Family Member Head
bank customers on Loan Schemes of Banks.
119
Table No.4.25. Group Statistics
Head of N Mean Std. Std. Error
Family Deviation Mean
Self 311 114.7170 20.26765 1.14927
Loan Schemes of
Family
Bank 104 113.8462 20.78076 2.03772
Member
Group statistics table explore the mean value of self family head of individuals is
114.7170, for std deviation is 20.26765 and for std error mean is 1.14927 whereas for
the mean value of Other Family head of individuals is 113.8462, for std deviation is
20.78076 and for std error mean is 2.03772.

Test for
Equality of
Variances
(2- Differe Differenc Interval of the
tailed nce e Difference
) Lower Upper
Equal
variances .032 .879 .369 413 .698 .86989 2.31041 -3.67075 5.41253
of Bank Equal
variances not .372 173.129 .710 .86989 2.33947 -3.74667 5.48845
assumed
levels. It means no significant difference between self head of family and other member
family head types of individuals on Loan Schemes of Bank. Hence the null hypothesis
was accepted by the results.
120
4.8. ANOVA Test of Equality of Error Variance
H06. There is no relationship between Age groups of bank customers on Loan

Schemes of Banks.
H07. There is no relationship between Occupations of bank customers on Loan

Schemes of Banks.
H08. There is no relationship between Monthly Income of bank customers on Loan

Schemes of Banks..
H09. There is no relationship between Education levels of bank customers on Loan

Schemes of Banks.
Table No. 4.27. Levene's Test of Equality of Error Variancesa
Dependent Variable: Loan Schemes of Bank

F df1 df2 Sig.
1.349 96 318 .029
Tests the null hypothesis that the error variance of the dependent variable is equal
across groups.
a. Design: Intercept + Age + Education + Monthly Income + Occupation + Age *
Education + Age * Monthly Income + Age * Occupation + Education * Monthly
Income + Education * Occupation + Monthly Income * Occupation + Age *
Education * Monthly Income + Age * Education * Occupation + Age * Monthly
Income * Occupation + Education * Monthly Income * Occupation + Age *
Education * Monthly Income * Occupation
Levene's Test of Equality of Error Variances is calculated to evaluate homogeneity of

groups on error variances. The value of F found insignificant at .150 levels of
significance which is unacceptable. The result has no importance for selecting Post
hoc test as the appropriate tests applicable on groups having different variances on the
variables of test will be selected while selecting options.
121
4.8.1. Tests of Between-Subjects Effects of different groups
The model used for ANOVA has bad fit as indicated by adjusted R square value of
0.219. The model fit was tested using F test and the value of F was found to be .927
which is insignificant at .666% level of significance, indicating that 21.9 % errors in
the dependent variable are explained and overall model fit was low.
Table No. 4.28. Tests of Between-Subjects Effects

Source Type III Sum of df Mean F Sig.
Squares Square
a
Corrected Model 37575.266 96 391.409 .927 .666
Intercept 995025.548 1 995025.548 2355.976 .000
Age 1864.615 4 466.154 1.104 .355
Education 1000.674 5 200.135 .474 .796
Monthly Income 1227.482 5 245.496 .581 .714
Occupation 1775.353 2 887.676 2.102 .124
Age * Education 2303.454 7 329.065 .779 .605
Age * Monthly Income 1142.937 5 228.587 .541 .745
Age * Occupation 8434.219 7 1204.888 2.853 .007
Education * Monthly Income 2263.578 5 452.716 1.072 .376
Education * Occupation 2601.532 8 325.192 .770 .630
Monthly Income * Occupation 4045.312 8 505.664 1.197 .300
Age * Education * Monthly Income 67.539 3 22.513 .053 .984
Age * Education * Occupation 5341.327 9 593.481 1.405 .185
Age * Monthly Income * Occupation 1234.312 5 246.862 .585 .712
Education * Monthly Income *
2638.077 7 376.868 .892 .513
Occupation
Age * Education * Monthly Income *
421.276 4 105.319 .249 .910
Occupation
Error 134304.484 318 422.341
Total 5612519.000 415
Corrected Total 171879.749 414
a. R Squared = .219 (Adjusted R Squared = -.017)
4.8.2. Post Hoc
In statistics, the Bonferroni correction is a method used to counteract the problem of

multiple comparisons. It is considered the simplest and most conservative method to
control the family wise error rate.
122
It is named after Italian mathematician Carlo Emilio Bonferroni for the use of
Bonferroni inequalities, but modern usage is credited to Olive Jean Dunn, who first
used it in a pair of articles written in 1959 and 1961
Table No. 4.29. Multiple Comparisons

Bonferroni
(I) Age (J) Age Mean Difference (I- Std. Sig. 95% Confidence Interval
Groups Groups J) Error Lower Upper
Bound Bound
31-45 -1.3299 2.51612 1.000 -8.4424 5.7826
46-60 -3.1315 3.25188 1.000 -12.3238 6.0608
16-30
61-75 1.8828 6.83926 1.000 -17.4501 21.2157
76 Above -5.3172 8.65630 1.000 -29.7865 19.1520
16-30 1.3299 2.51612 1.000 -5.7826 8.4424
46-60 -1.8016 2.79697 1.000 -9.7079 6.1048
31-45
61-75 3.2127 6.63504 1.000 -15.5429 21.9683
76 Above -3.9873 8.49587 1.000 -28.0030 20.0285
16-30 3.1315 3.25188 1.000 -6.0608 12.3238
31-45 1.8016 2.79697 1.000 -6.1048 9.7079
46-60
61-75 5.0143 6.94749 1.000 -14.6246 24.6531
76 Above -2.1857 8.74206 1.000 -26.8974 22.5260
16-30 -1.8828 6.83926 1.000 -21.2157 17.4501
31-45 -3.2127 6.63504 1.000 -21.9683 15.5429
61-75
46-60 -5.0143 6.94749 1.000 -24.6531 14.6246
76 Above -7.2000 10.61246 1.000 -37.1988 22.7988
16-30 5.3172 8.65630 1.000 -19.1520 29.7865
31-45 3.9873 8.49587 1.000 -20.0285 28.0030
76 Above
46-60 2.1857 8.74206 1.000 -22.5260 26.8974
61-75 7.2000 10.61246 1.000 -22.7988 37.1988
Based on observed means.
The error term is Mean Square(Error) = 422.341.
Bonferroni correction is the basic type of simultaneous inference that aims to control
the family wise error rate. A significant statistical research was done in the field from
early 1960s until late '90s, and many improvements were offered. Most notably are
the Holm–Bonferroni method, which offers a uniformly more powerful test procedure
(i.e., more powerful regardless of the values of the unobservable parameters), and the
Hochberg (1988) method, guaranteed to be no less powerful and is in many cases
123
more powerful when the tests are independent (and also under some forms of positive
dependence).
The Bonferroni correction can be somewhat conservative if there are a large number
of tests and/or the test statistics are positively correlated. Bonferroni correction
controls the probability of false positives only. The correction ordinarily comes at the
cost of increasing the probability of producing false negatives, and consequently
reducing statistical power. When testing a large number of hypotheses, this can result
in large critical values.
In addition, in certain situations where one wants to retain, not reject, the null
hypothesis, and then Bonferroni correction is non-conservative. Hence the null
hypothesis H06 was accepted by the test.

Bonferroni
(I) Occupation (J) Occupation Mean Std. Error Sig. 95% Confidence Interval
Difference (I-J) Lower Bound Upper Bound
Business .1857 2.19372 1.000 -5.0940 5.4653
Job
Other .2111 3.13822 1.000 -7.3417 7.7638
Job -.1857 2.19372 1.000 -5.4653 5.0940
Business
Other .0254 3.04641 1.000 -7.3065 7.3572
Job -.2111 3.13822 1.000 -7.7638 7.3417
Other
Business -.0254 3.04641 1.000 -7.3572 7.3065
Above table in Bonferroni showing that there is insignificant difference between all
factors. Hence the null hypothesis H07 was accepted by the test.
124
Bonferroni
(I) Income (J) Income Mean Difference Std. Sig. 95% Confidence Interval
Monthly Monthly (I-J) Error Lower Upper
Bound Bound
10001-20000 -.4201 2.37886 1.000 -7.4557 6.6155
20001-30000 .2164 4.55215 1.000 -13.2468 13.6796
0-10000 30001-40000 .9809 4.46733 1.000 -12.2314 14.1932
40001-50000 -3.8108 4.46733 1.000 -17.0231 9.4016
50000 & Above 3.8886 3.71134 1.000 -7.0879 14.8650
0-10000 .4201 2.37886 1.000 -6.6155 7.4557
20001-30000 .6365 4.65427 1.000 -13.1286 14.4017
10001-20000 30001-40000 1.4010 4.57133 1.000 -12.1189 14.9210
40001-50000 -3.3906 4.57133 1.000 -16.9105 10.1293
50000 & Above 4.3087 3.83590 1.000 -7.0362 15.6536
0-10000 -.2164 4.55215 1.000 -13.6796 13.2468
10001-20000 -.6365 4.65427 1.000 -14.4017 13.1286
20001-30000 30001-40000 .7645 5.99668 1.000 -16.9710 18.4999
40001-50000 -4.0272 5.99668 1.000 -21.7626 13.7083
50000 & Above 3.6722 5.45686 1.000 -12.4667 19.8110
0-10000 -.9809 4.46733 1.000 -14.1932 12.2314
10001-20000 -1.4010 4.57133 1.000 -14.9210 12.1189
30001-40000 20001-30000 -.7645 5.99668 1.000 -18.4999 16.9710
40001-50000 -4.7917 5.93255 1.000 -22.3374 12.7541
50000 & Above 2.9077 5.38629 1.000 -13.0225 18.8379
0-10000 3.8108 4.46733 1.000 -9.4016 17.0231
10001-20000 3.3906 4.57133 1.000 -10.1293 16.9105
40001-50000 20001-30000 4.0272 5.99668 1.000 -13.7083 21.7626
30001-40000 4.7917 5.93255 1.000 -12.7541 22.3374
50000 & Above 7.6993 5.38629 1.000 -8.2309 23.6295
0-10000 -3.8886 3.71134 1.000 -14.8650 7.0879
10001-20000 -4.3087 3.83590 1.000 -15.6536 7.0362
50000 & Above 20001-30000 -3.6722 5.45686 1.000 -19.8110 12.4667
30001-40000 -2.9077 5.38629 1.000 -18.8379 13.0225
40001-50000 -7.6993 5.38629 1.000 -23.6295 8.2309
factors. Hence the null hypothesis H08was accepted by the test.
125
Bonferroni
(I) Education (J) Education Mean Difference Std. Sig. 95% Confidence Interval
Level Level (I-J) Error Lower Upper
Bound Bound
Up to High
1.6391 4.00696 1.000 -10.2116 13.4899
School
Higher
-.9838 2.86185 1.000 -9.4479 7.4802
Illiterate Secondary
Graduation -1.5621 2.77120 1.000 -9.7580 6.6338
Post Graduation 3.9473 6.84655 1.000 -16.3017 24.1962
Others -4.0949 5.18359 1.000 -19.4255 11.2358
Illiterate -1.6391 4.00696 1.000 -13.4899 10.2116
Higher
-2.6230 3.86830 1.000 -14.0636 8.8177
Up to High Secondary
School Graduation -3.2012 3.80172 1.000 -14.4450 8.0425
Post Graduation 2.3081 7.32453 1.000 -19.3545 23.9707
Others -5.7340 5.80027 1.000 -22.8885 11.4205
Illiterate .9838 2.86185 1.000 -7.4802 9.4479
Up to High
2.6230 3.86830 1.000 -8.8177 14.0636
Higher School
Secondary Graduation -.5783 2.56661 1.000 -8.1691 7.0126
Post Graduation 4.9311 6.76633 1.000 -15.0806 24.9428
Others -3.1110 5.07716 1.000 -18.1269 11.9049
Illiterate 1.5621 2.77120 1.000 -6.6338 9.7580
Up to High
3.2012 3.80172 1.000 -8.0425 14.4450
School
Graduation Higher
.5783 2.56661 1.000 -7.0126 8.1691
Secondary
Post Graduation 5.5094 6.72849 1.000 -14.3904 25.4091
Others -2.5328 5.02662 1.000 -17.3992 12.3337
Illiterate -3.9473 6.84655 1.000 -24.1962 16.3017
Up to High
-2.3081 7.32453 1.000 -23.9707 19.3545
School
Post Graduation Higher
-4.9311 6.76633 1.000 -24.9428 15.0806
Secondary
Graduation -5.5094 6.72849 1.000 -25.4091 14.3904
Others -8.0421 8.02886 1.000 -31.7878 15.7036
Illiterate 4.0949 5.18359 1.000 -11.2358 19.4255
Others Up to High
5.7340 5.80027 1.000 -11.4205 22.8885
School
126
Higher
3.1110 5.07716 1.000 -11.9049 18.1269
Secondary
Graduation 2.5328 5.02662 1.000 -12.3337 17.3992
Post Graduation 8.0421 8.02886 1.000 -15.7036 31.7878
4.9. Comparison between HDFC and SBI Facilities towards Loan

Schemes
4.9.1. Frequency Distribution for Better Facilities
The main object of frequency analysis is to examine which banks gives you better
facility towards personal loan.
.
Table No. 4.33. Which bank gives you better facility
Cumulative Percent
HDFC
Bank 308 74.2
State
Bank
Valid of 107 25.8
India
Total
415 100.0
Through the frequency found amazing results that private banks provide us better loan
facility than public bank. According to Audience vote maximum people are agree that
HDFC bank give us better facility than SBI. The total value of HDFC Banks
127
have74.2% with 308 frequency and SBI have 25.8% with 107 frequency tested
through frequency test against the “Which bank gives you better facility”
4.9.2. Frequency Distribution for Repayment Facility
The main object of frequency analysis is to examine which bank gives you flexible
repayment facility towards personal loan.
Table No. 4.34. Provides flexible repayment facility

Cumulative Percent
State Bank of
India
281 67.7
Valid HDFC Bank
134 32.3
Total 415 100.0
Here clearly indicate through frequency analysis that State Bank India (SBI) provides
us better repayment facility than HDFC banks. The total frequency of SBI is 281
respondents with 67.7% and HDFC Banks have 134 total frequencies with 32.3%.
4.9.3. Frequency Distribution for Penalty Charge
The main object of frequency analysis is to examine which bank gives penalty charge
towards personal loan.
128
Table No. 4.35. Minimum Penalty Charges on loan
Cumulative Percent
HDFC
Bank
205 49.4
State
Valid
Bank of
India 210 50.6
Total 415 100.0
Near about equal penalty charge by the both banks according to audience poll. Here
clearly shown values in above table of frequency analysis that State Bank India (SBI)
provides us similar repayment facility than HDFC banks. The total frequency of SBI
is 205 respondents with 49.4% and HDFC Banks have 210 total frequencies with
50.6%.
4.9.4. Frequency Distribution for Less Paper Work
The main object of frequency analysis is to examine which bank gives less paper
work towards personal loan.
129
Table No. 4.36. Minimum paper work

Cumulative Percent
HDFC
Bank
346 83.4
State
Valid
Bank of
India 69 16.6
Total 415 100.0
The values clearly shown in above table of frequency analysis that State Bank India
(SBI) provides us poor quality of paper work than HDFC banks. The total frequency
of SBI is 69 respondents with 16.6% and HDFC Banks have 346 total frequencies
with 83.4%. Indicating that SBI have lengthy paper work regarding to personal loan
schemes tested through frequency.
130
4.9.5. Frequency Distribution for Low File Charge
The main object of frequency analysis is to examine which bank gives lower file
charge towards personal loan.
Table No. 4.37. Lower File Charge

Cumulative Percent
State
Bank of
India 233 56.1
Valid
HDFC
Bank
182 43.9
Total 415 100.0
(SBI) provides us lowest file charge on personal loan than HDFC banks. The total
frequency of SBI is 233 respondents with 56.1% and HDFC Banks have 182 total
frequencies with 43.9%. Indicating that SBI have nothing or minimum file charge
regarding to personal loan schemes tested through frequency.
4.9.6. Frequency Distribution for Better Coordination with Staff
The main aim of this frequency analysis is to examine which bank gives better
coordination between customer and employees towards personal loan.
(SBI) provides little bit more coordination between customer and staff towards
personal loan than HDFC banks. The total frequency of SBI is 216 respondents with
52.0% and HDFC Banks have 199 total frequencies with 48.0%.
131
Table No. 4.38. Better coordination
Cumulative Percent
HDFC
Bank
216 52.0
State
Valid
Bank of
India 199 48.0
Total 415 100.0
Above values of entire table indicating that the customers of SBI and HDFC have
good coordination with their staff.
4.9.7. Frequency Distribution for Bank Reputation
The main aim of this frequency analysis is to examine which bank gives good
reputation of bank towards personal loan.
Table No. 4.39. Banks enjoy better reputation

Cumulative Percent
HDFC Bank
349 84.1
State Bank of India

Valid 66 15.9
Total 415 100.0
132
(SBI) provides poor reputation of among customer towards personal loan than HDFC
banks. The total frequency of SBI is 66 respondents with 15.9% and HDFC Banks
have 349 total frequencies with 84.1%.
Above values of entire table indicating that the customers of feel proud to HDFC have
good reputation than SBI.
4.10. Regression Analysis
4.10.1. Linear Regression between Service Quality and Loan Schemes of Banks
The regression was calculated by taking the total of Service Quality and Loan
Schemes of Banks by using SPSS software. In this the Service Quality was
independent variable and Loan Schemes of Banks was the dependent variable.
Therefore, regression is calculated by taking dependent and independent variable.
Null Hypothesis (H010): It stated that there is no significant impact of Service Quality
Alternative Hypothesis (H10): It stated that there is no significant impact of Service

Quality on Loan Schemes of Banks.
Table No. 4.40. Table No. Model Summaryb
Model R R Square Adjusted R Square Std. Error of the Estimate Durbin-Watson
1 .839a .704 .703 10.855 2.027
a. Predictors: (Constant), Service Quality

b. Dependent Variable: Loan Schemes of Banks
The obtain value of R Square is 0.704 indicating that 70.4% relationship between
Service Quality and Loan Schemes of Banks. Durbin Watson value i.e. 2.027 in under
standard value 1-3 indicating that there is no auto correlation.
133
Table No. 4.41. ANOVAb
Model Sum of Squares df Mean Square F Sig.
1 Regression 115660.023 1 115660.023 981.501 .000a
Residual 48667.890 413 117.840
Total 164327.913 414
a. Predictors: (Constant), Service Quality

The value of F i.e. 981.501 indicating good model fit at the significance level 00.00%.
Table No.4.42. Coefficientsa
Unstandardized Standardized
Coefficients Coefficients
Model B Std. Error Beta t Sig.
1 (Constant) 25.312 2.793 9.062 .000
Service Quality 2.966 .095 .839 31.329 .000
a. Dependent Variable: Loan Schemes of Banks

The significance of beta is tested using “t” test and value for model i.e. 31.329 which
was significance level at 0.000% indicating a strong positive relationship between
Service Quality and Loan Schemes of Banks.
Graphs No. 05. Histogram and Normal P-P Plot of Regression
Histogram Normal P-P Plot of Regression

4.10.2. Linear Regression between Employee Commitment and Loan Schemes of
Banks
134
The regression was calculated by taking the total of Employee Commitment and Loan
Schemes of Banks by using SPSS software. In this the Employee Commitment was
Null Hypothesis (H011): It stated that there is no significant impact of Employee

Commitment on Loan Schemes of Banks.
Alternative Hypothesis (H11): It stated that there is no significant impact of Employee

Commitment on Loan Schemes of Banks.
Table No. 4.43. Model Summaryb
1 .859a .738 .737 10.209 1.796
a. Predictors: (Constant), Employee Commitment

b. Dependent Variable: Loan Schemes
Employee Commitment and Loan Schemes of Banks. Durbin Watson value i.e. 1.796
in under standard value 1-3 indicating that there is no auto correlation
1 Regression 121279.494 1 121279.494 1163.537 .000a
Residual 43048.420 413 104.233
Total 164327.913 414
a. Predictors: (Constant), Employee Commitment

The value of F i.e. 1163.537 indicating excellent model fit at the significance level
00.00%.
135
Table No. 4.45. Coefficientsa
1 (Constant) 34.931 2.292 15.241 .000
Employee Commitment 2.505 .073 .859 34.111 .000
a. Dependent Variable: Loan Schemes
Employee Commitment and Loan Schemes of Banks.
4.10.3. Linear Regression between Safety Security and Loan Schemes of Banks
The regression was calculated by taking the total of Safety Security and Loan
Schemes of Banks by using SPSS software. In this the Safety Security was
Null Hypothesis (H012): It stated that there is no significant impact of Safety Security
136
Alternative Hypothesis (H12): It stated that there is no significant impact of Safety
Security on Loan Schemes of Banks.
1 .851a .724 .723 10.488 1.743
a. Predictors: (Constant), Safety Security

Safety Security and Loan Schemes of Banks. Durbin Watson value i.e. 1.743 in under
standard value 1-3 indicating that there is no auto correlation
1 Regression 118902.427 1 118902.427 1081.039 .000a
Residual 45425.487 413 109.989
Total 164327.913 414
a. Predictors: (Constant), Safety Security

00.00%.
1 (Constant) 35.906 2.348 15.295 .000
Safety Security 3.225 .098 .851 32.879 .000
137
Safety Security and Loan Schemes of Banks.
4.10.4. Linear Regression between Safety Security and Loan Schemes of Banks
The regression was calculated by taking the total of Ease Process and Loan Schemes
of Banks by using SPSS software. In this the Ease Process was independent variable
and Loan Schemes of Banks was the dependent variable. Therefore, regression is
calculated by taking dependent and independent variable.
Null Hypothesis (H013): It stated that there is no significant impact of Ease Process on
Loan Schemes of Banks.
Alternative Hypothesis (H13): It stated that there is no significant impact of Ease

Process on Loan Schemes of Banks.
1 .676a .456 .455 14.706 1.837
a. Predictors: (Constant), Ease Process

138
Ease Process and Loan Schemes of Banks. Durbin Watson value i.e. 1.837 in under
standard value 1-3 indicating that there is no auto correlation
1 Regression 75014.833 1 75014.833 346.882 .000a
Residual 89313.080 413 216.254
Total 164327.913 414
a. Predictors: (Constant), Ease Process

00.00%.
1 (Constant) 55.396 3.083 17.970 .000
Ease Process 5.147 .276 .676 18.625 .000
Ease Process and Loan Schemes of Banks.
139
4.10.5. Linear Regression between Sanitation Facility and Loan Schemes of

Banks
The regression was calculated by taking the total of Sanitation Facility and Loan
Schemes of Banks by using SPSS software. In this the Sanitation Facility was
Null Hypothesis (H014): It stated that there is no significant impact of Sanitation

Facility on Loan Schemes of Banks.
Alternative Hypothesis (H14): It stated that there is no significant impact of Sanitation

Facility on Loan Schemes of Banks.
1 .680a .463 .462 14.619 1.864
a. Predictors: (Constant), Sanitation Facility

Sanitation Facility and Loan Schemes of Banks. Durbin Watson value i.e. 1.864 in
under standard value 1-3 indicating that there is no auto correlation
140
1 Regression 76060.555 1 76060.555 355.885 .000a
Residual 88267.358 413 213.722
Total 164327.913 414
a. Predictors: (Constant), Sanitation Facility

00.00%.

1 (Constant) 60.641 2.775 21.851 .000
Sanitation Facility 4.871 .258 .680 18.865 .000
Sanitation Facility and Loan Schemes of Banks.
141
5.10.6. Linear Regression between Promotion Advertising and Loan Schemes of
Banks
The regression was calculated by taking the total of Promotion Advertising and Loan
Schemes of Banks by using SPSS software. In this the Promotion Advertising was
Null Hypothesis (H015): It stated that there is no significant impact of Promotion

Advertising on Loan Schemes of Banks.
Alternative Hypothesis (H15): It stated that there is no significant impact of Promotion

1 .573a .328 .327 16.347 1.676
a. Predictors: (Constant), Promotion Advertising

Promotion Advertising and Loan Schemes of Banks. Durbin Watson value i.e. 1.676
in under standard value 1-3 indicating that there is no auto correlation
1 Regression 53965.732 1 53965.732 201.952 .000a
Residual 110362.182 413 267.221
Total 164327.913 414
a. Predictors: (Constant), Promotion Advertising

142
Table No .4.57. Coefficientsa
1 (Constant) 71.690 2.895 24.766 .000
Promotion Advertising 5.473 .385 .573 14.211 .000
Promotion Advertising and Loan Schemes of Banks.
4.11. Frequency Analysis

4.11.1. Frequency Distribution for Satisfied with loan schemes of banks
The main aim of this frequency analysis is to examine overall performance towards
satisfaction with current loan schemes of bank of both bank i.e. State bank of India
and HDFC Bank.
143
Table No. 4.58. Overall you are satisfied with loan schemes of Banks
Cumulative Percent
Yes
218 52.5
No
Valid
197 47.5
Total 415 100.0
The values in the entire table clearly shown that overall satisfaction with present
personal loan schemes of banks, through the audience vote 52.5% percent people
agree with their statement i.e. satisfied with their present loan schemes rather than
47.55 of customers of State Bank India (SBI) and HDFC banks. The total frequency
of satisfied customer is 218 respondents and not satisfied have 197 total frequencies.
4.11.2. Frequency Distribution for Future Loan Plan
The main aim of this frequency analysis is to check either people willing to take
personal loan future also or not.
The values in entire table of frequency analysis that the maximum people agree with
the statement of yes against the question are you willing to take personal loan in
future also from present bank. The total frequency of yes I willing to take personal
loan to my present banks is 348 respondents with 83.9% and rest of people want to
move another bank for personal loan or they don‟t want to take personal loan to
present bank is 67 total frequencies with 16.1%.
144
Table No. 4.59. Are you willing to take loan in future also in same bank
Cumulative Percent
Yes
348 83.9
No
Valid 67 16.1
415 100.0
Total
B. Data Analysis from Authority Side
4.12. Consistency & Reliability Measurements

4.12.1. Consistency Measurements
Consistency of all the statements in the questionnaires was verified through item total
correlation. Under this methods of every item with total was measured and the
computed value was compared with standard value. The all statements having item to
total correlation lower than the critical value, were declared as inconsistent and
dropped from the questionnaire.
145
Table No. 4.60. Item-Total Statistics
Corrected Item- Cronbach's Status
Total Alpha if Item
Correlation Deleted
It easy to process to obtain personal loans through banks .479 .890 Required
Easy to understand procedures of loan .417 .892 Required
Documentation is relevant under the schemes .537 .889 Required
Lower interest rate influence you to obtain personal loan .467 .891 Required
Beneficiary schemes motivated to obtain loan .466 .891 Required
Family needs is greater purpose to taking loan .396 .893 Required
Employee Empowerment of Banks solve your problem .617 .887 Required
Nearby location banks is only reason to taking loan .498 .890 Required
Banks brochure offering personal loan .555 .888 Required
Agent are helpful to obtain loan from beginning .504 .890 Required
Television or advertising offering to get personal loan .452 .891 Required
It is an essential when you need of money .535 .889 Required
File charges and other processing fee low .575 .888 Required
Banks give you relevant advice about personal loan .414 .892 Required
Bank employee pay 100 percent attention towards you .576 .888 Required
Well dress up employee attract you .579 .888 Required
It remain information of your loan confidential .484 .890 Required
Cleanliness of banks make you happy .548 .888 Required
Other supporting staffs helpful in nature .413 .892 Required
Toiletry facility of banks subsidiary reason for taking loan .480 .890 Required
Banks give you information about other banks facility .450 .891 Required
They solve your problem internally .494 .890 Required
Banks employees are trustworthy about loan .368 .893 Required
Cronbach Alpha had been obtained 0.894 after reliability test through SPSS
which is more than the cut off value (0.7). In above table of consistency measures
mainly two things are considered,
1. Corrected Item to Total Correlation values which is acceptable if greater

than.
2. Secondly Cronbach's Alpha if Item Deleted value for each item is evaluated
and if found value greater than the calculated reliability value (0.894) than that
item should be dropped from the questionnaire and not considered for further
study.
146
So that from above table of consistency of all statements having item to total
correlation lower than the critical value. Obtain results indicate that there is no need to
drop any statement in questionnaire because all statements.
4.12.2. Reliability Test
Cronbach‟s Alpha test had been applied to calculate reliability of all items of Loan
Schemes of bank from the questionnaire. It is considered that the reliability value
more than 0.7 is good and it can be seen that in almost all the reliability values is quite
higher than the standard value.
The Cronbach‟s alpha reliability method was applied to check the reliability of all the
items of the questionnaire. The reliability test measures are as shown below:
Table No. 4.61. Reliability Statistics

Variables Name Cronbach's Alpha N of Items Test Results
Loan Schemes of Data were Highly
.894 23
Banks Reliable
The reliability coefficient value of Loan Schemes of Bank were highly significant i.e.
0.894 that shows high reliability of the questionnaire, so it can be considered good
enough for the present study, the total number of statement have 23 for this study.
4.13. Normality Analysis
There are two main methods of assessing normality: graphically and numerically. The
table below presents the results from two well-known tests of normality, namely the
Kolmogorov-Smirnov Test and the Shapiro-Wilk Test. The Shapiro-Wilk Test is more
appropriate for small sample sizes (< 50 samples), but can also handle sample sizes as
large as 2000. Due to this consideration, Shapiro-Wilk test is used to assessment of
normality as numerical means of assessing normality is done in this study.
147
Table No. 4.62. Tests of Normality
Kolmogorov- Shapiro-Wilk Test results
Smirnova
Statistic df Sig Statistic df Sig. Data is Normally
Loan Schemes of Banks .101 223 .102 .986 223 .093 Distributed
a. Lilliefors Significance Correction
Above table shows that data is normally distributed with significant level of .093. It is
because if the significance value of the Shapiro-Wilk Test is greater than 0.02, the
data is normal. If it is below 0.05 (P<0.05), the data significantly deviate from a
normal distribution.
4.14. Factor Analysis
Factor analysis is a statistical method used to describe variability among observed,

correlated variables in terms of a potentially lower number of unobserved variables
called factors. Factor analysis originated in psychometrics, and is used in behavioral
sciences, social sciences, marketing, product management, operations research, and
other applied sciences that deal with large quantities of data.
Factor analysis searches for such joint variations in response to unobserved latent
variables. The observed variables are modeled as linear combinations of the potential
factors. The information gained about the interdependencies between observed
variables can be used later to reduce the set of variables in a dataset.
Correlation matrix is the starting point of factor analysis. In this analysis inter
correlations between the studied variables are presented. The dimensionality of this
matrix can be reduced by looking for variables that correlate highly with a group of
other variables, but correlate very badly with variables outside of that group (Field;
2000). Then after these variables with high inter correlations could well measure one
underlying variable, which is called a „factor‟. The obtained factor creates a new
dimension which can be visualized as classification axes along with measurement
variables can be plotted (Field; 2000). This projection of the scores of the original
variables on the factor leads to two results: factor scores and factor loadings.
148
Exploratory factor analysis (EFA) is used to identify complex interrelationships
among items and group items that are part of unified concepts. The researcher makes
no "a priori" assumptions about relationships among factors.
4.14.1. K. M. O. measurement of sampling Adequacy and Bartlett’s Test of

Sphericity for Validity Analysis of Loan Schemes of Bank
Table No. 4.63. KMO and Bartlett's Test

Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .897
Approx. Chi-Square 1520.082
Bartlett's Test of Sphericity df 253
Sig. .000
Test results This data is Adequate for Factor Analysis
A KMO measure of sampling reflected KMO value of 0.897 which indicated that the
size of sample is adequate for the present study. The Bartlett‟s test measuring Chi-
Square value 1520.082 which is significant at 0.000% level of significance, indicating
that null hypothesis is rejected. Hence, it shows that item matrix is not an identity
matrix and the value revealed that the data is normally distributed and suitable enough
for factor analysis.
4.14.2. Description of Factors
Principle component factor analysis with Varimax rotation was applied. The raw
scores of 32 items were subjected to factor analysis to find out the factors that
contribute towards Loan Schemes of Bank. After factor analysis 6 factors were
identified. The rotation converged in 7 iterations.
149
Table No. 4.64. Descriptive of Factor Analysis
Variable Eigen % Statement Loading
Value Variance Value
Customer 7.059 30.690 Agent are helpful to obtain loan
.678
Support from beginning
System Nearby location banks is only
.663
reason to taking loan
Television or advertising
.633
offering to get personal loan
Bank employee pay 100 percent
.621
attention towards you
Employee Empowerment of
.601
Banks solve your problem
Trustworthy 1.405 6.110 Agent are helpful to obtain loan .691
Employee from beginning
Commitment Nearby location banks is only
.624
reason to taking loan
Television or advertising
.594
offering to get personal loan
Bank employee pay 100 percent
.495
attention towards you
Employee Empowerment of
.482
Banks solve your problem
Agent are helpful to obtain loan
.463
from beginning
Tenure of 1.312 5.702 Banks employees are trustworthy .723
Loan and about loan
Interest Other supporting staffs helpful in
.614
nature
They solve your problem
.546
internally
Banks give you information
about other banks facility i.e. .539
locker etc.
It remain information of your
.504
loan confidential
Cleanliness of banks make you
.421
happy
Positive Loan 1.198 5.207 Family needs is greater purpose .748
Advertising to taking loan
Beneficiary schemes motivated
.633
to obtain loan
It easy to process to obtain
.498
personal loans through banks
Cleanliness 1.146 4.984 Banks give you relevant advice about .698
and Well personal loan
Easy to understand procedures of loan .573
Dressed
Lower interest rate influence you to
obtain personal loan
.464
150
i. Factor One: Customer Support System
The total Eigen value of this factor is 7.059 and this factor has emerged as the
most important determinant of Loan Schemes of Banks and explains 30.690 % of
variance of Loan Schemes of Banks. The important statement constituting this
factor include; Agent are helpful to obtain loan from beginning 0.678, Nearby
location banks is only reason to taking loan 0.663, Television or advertising
offering to get personal loan 0.633, Bank employee pay 100 percent attention
towards you 0.621 and Employee Empowerment of Banks solve your problem
0.601
ii. Factor Two: Trustworthy Employee Commitment
most important determinant of Loan Schemes of Banks and explains 6.110% of
factor include; Toiletry facility of banks subsidiary reason for taking loan 0.691,
Documentation is relevant under the schemes 0.624, It is an essential when you
need of money 0.594, Banks brochure offering personal loan 0.495, File charges
and other processing fee low 0.482 and Well dress up employee attract you 0.463.
iii. Factor Three: Tenure of Interest & Loan
factor include; Banks employees are trustworthy about loan 0.723, Other
supporting staffs helpful in nature 0.614, They solve your problem internally
0.546, Banks give you information about other banks facility 0.539, It remain
information of your loan confidential 0.504 and Cleanliness of banks make you
happy 0.421.
iv. Factor Four: Positive Loan Advertising
factor include; Family needs is greater purpose to taking loan 0.748, Beneficiary
schemes motivated to obtain loan 0.633 and It easy to process to obtain personal
loans through banks 0.498.
151
v. Factor Five: Cleanliness and Well Dressed
factor include; Banks give you relevant advice about personal loan 0.698, Easy to
understand procedures of loan 0.573 and Lower interest rate influence you to
obtain personal loan 0.464.
4.14.3. Scree Plot

Both the Eigen value rule and the scree plot are accurate if the sample is >250 and the
commonalities are greater than 0.6. Scree plot test is another popular test for the
number of factors is Catell‟s scree criterion (Pallant, 2005). This criterion involves
plotting the Eigen values of the factors, and looking for the point at which the plot
begins to level off from vertical to horizontal. In this case, similar result is suggested
by the scree plot, which would suggest the retention of seven significant factors (Hair
et al., 2003). The results of Catell‟s scree plot are shown in Figure below.
4.14.4. Component Score Covariance Matrix
The result of the analysis was a rotated component matrix consisting of six
components that account for 52.69%of the variance. The breaking point of the Scree-
152
plot, however, was at 5 or 6 factors; extraction of only 5 factors would account, but
enhances the overview of the rotated component matrix considerably.
Table No. 4.65. Component Transformation Matrix

Component Customer Trustworthy Tenure of
Support Employee Loan and Positive Loan Cleanliness and
System Commitment Interest Advertising Well Dressed
Customer Support
.524 .526 .429 .365 .362
System
Trustworthy
Employee .028 -.097 .753 -.632 -.155
Commitment
Tenure of Loan and
-.683 -.157 .329 .242 .584
Interest
Positive Loan
-.048 -.283 .376 .638 -.608
Advertising
Cleanliness and
.505 -.780 -.007 .044 .366
Well Dressed
Fig. No. 11. Factor Model
Trustworthy Employee
Commitment
Tenure of Loan and

Interest
Customer Support 1.405
System
1.312
7.059
Loan Schemes of
Banks
1.146
1.198
Positive Loan Cleanliness and Well

Advertising
Dressed
153
4.15. Correlation between obtain factors of loan schemes of banks
from authority side
The correlation was calculated by taking the total of obtain factors of Loan Schemes of
Banks by using SPSS software. The main purpose of correlation analysis is to find whether
all factors have positive correlation one and another.
Table No. 4.66. Correlations
Customer Trustworthy Tenure of Positive Cleanliness

Support Employee Loan and Loan and Well
System Commitment Interest Advertising Dressed
Customer Support Pearson Correlation 1 .632** .554** .422** .425**

System
Sig. (2-tailed) .000 .000 .000 .000
N 223 223 223 223 223
Trustworthy Employee Pearson Correlation .632** 1 .572** .511** .490**

Commitment
Sig. (2-tailed) .000 .000 .000 .000
N 223 223 223 223 223
Tenure of Loan and Pearson Correlation .554** .572** 1 .430** .433**

Interest
Sig. (2-tailed) .000 .000 .000 .000
N 223 223 223 223 223
Positive Loan Pearson Correlation .422** .511** .430** 1 .464**

Advertising
Sig. (2-tailed) .000 .000 .000 .000
N 223 223 223 223 223
Cleanliness and Well Pearson Correlation .425** .490** .433** .464** 1

Dressed
Sig. (2-tailed) .000 .000 .000 .000
N 223 223 223 223 223
**. Correlation is significant at the 0.01 level (2-tailed).
Hence the correlation clearly indicates that there are Customer Support System,
Trustworthy Employee Commitment, Tenure of Loan and Interest, Positive Loan
154
Advertising and Cleanliness and Well Dressed have positive correlation between each
other.
4.16. Structure Equation Modeling
SEM can best be defined as a class of methodology that seeks to represent hypotheses
about the means, variances, and covariance of observed data in terms of smaller
number of structural parameters defined by a hypothesized underlying model (Kaplan,
2000)8. In other words, SEM is a comprehensive statistical approach to testing
hypotheses about relations among observed and latent variable (Hoyle, 1995) and one
of the more popular statistical methodologies available to quantitative researchers9.
SEM begins with the specification of a model to be estimated. A model is a statistic

statement about the relations among variables. The specification is the exercise of
formally stating a model and it is central to the SEM approach. Indeed, no analysis
can take place until the researcher has specified a model of the relations among the
variables to be analyzed (Hoyle, 1995). In SEM model specification involves
formulating a statement about a set of parameters. Parameters typically are specified
as either fixed or free. Fixed parameters are not estimated from the data and their
value that the investigator believes to be nonzero (Hoyle, 1995). There are two
components of the general structural equation model: the measurement model is that
components of the general model in which latent variables are prescribed. The
structural model is that component of the general model that prescribes relations
between latent variables and observed variables that are not indicators of latent
variables (Hoyle, 1995). A fundamental consideration when specifying models in
SEM is identification. Identification concerns whether a single unique value for each
and every free parameter can be obtained from the observed data10 (Hoyle, 1995).
Once a model has been specified, the next task is to obtain estimates of the free
parameters from a set of observed data. Although single-stage least square methods
such as those used in standard ANOVA or multiple regressions can be used to derive
parameter estimates, iterative methods such as maximum likelihood or generalized
8
9
Hoyle, R.H. (1995) (ed.). Structural equation modeling: concept, issue and applications. Newbury
Park CA
10
Ibid
155
least square are preferred. The index of fit provides a perspective on the fit of
structural equation models. The basic idea behind indices is that the fit of the model is
compared to the fit of some baseline model that usually specifies complete
independence among the observed variables. Some of these indices include normed fit
index (NFI), Tucker-Lews Index (TLI), and comparative fit Index (CFI). According to
Kaplan, (2000), these indices are typically scaled to lie between zero and one, with
one representing perfect fit relative to the baseline model. He noted that 0.95 is
indicative of good fit relative to the baseline model. In other words, the value of 0.95
is considered evidence that the target model fit is a god fit to the data relative to the
baseline model11.
4.16.1. Goodness of Indices
Goodness of fit indices (or fit indices) indices the goodness of fit between the
hypothesized model and the observed data, in this sub-chapter, the most commonly
used goodness of fit indices, and then cut off (suggested) value of those indices are
presented.
4.16.2. Chi-Square
Chi-square (χ2) is a traditional measure of overall model fit (Howell, 1997), Chi-
square test the validity of the specifications of factor loading, factor covariance, and
error variances for the studied model12 (Byrne, 2010, P 79). The Chi-square statistics
is associated with a probability. Low probability indicates a poor fit of the model
(Byrne, 2001, p 80), for a good model fit, the probability should be no significant, that
is greater than .05 (Hatcher, 1994, p. 339). There is also a guideline for the ratio of
chi-square and degree of freedom (df)13. According to hatcher (1994, p 339), the chi-
square df ratio should be at least 2. The use of chi-square has major drawback; for
example, with larger sample sizes the chi-square can reject a valid14 (Bentler and
Bonnet, 1980; Kline, 1998).
11
12
Byrne, B. M. (2010). Structural equation modeling with AMOS: Basic concepts, applications, and
programming (2nd ed.). Mahwah, NJ: Erlbaum
13
Hatcher, L. (1994). A step-by-step approach to using the SAS® System for factor analysis and
structural equation modeling Cary, NC: SAS Institute Inc.
14
Bentler, P. M. (2005). EQS 6 structural equations program manual. Encino, CA: Multivariate
Software.
156
4.16.2. Confirmatory Factor Analysis
Confirmatory factor analysis (CFA) was used to help the selection process of the scale
items. The use of CFA requires knowledge of the underlying latent variable structure
(Byrne, 2001, p6). A model or relationship of the observed variables must be
specified before the factor analysis, whose task then is to test goodness of fit between
the model and observed data (Byrne, 2001, p.6; Hatcher, 1994, p.289). in essence,
CFA is a way to test how measured variables represent a smaller number of constructs
(Hair et al., 2010 p.693). The analysis helps to identify the factor loading of individual
items. Cross loadings can be studied with help of CFA. The analysis also helps to
define the optional number of items.
Model fits was evaluated by using the normed chi-square, Tucker-Lewis Index (TLI),
Comparative Fit Index (CFI) and Root-Mean-Square-Error of Approximation
(RMSEA). The first indicated a poor fit of the model to the data. There model
improvement techniques were applied (Anderson & Gerbing, 1998), and factor
loading and significance levels were examined.
A χ2 statistic is listed in the column between the tools and the path diagram. The value
of χ2was 109.43 and degree of freedom (df) was found 78 at significant probability
level of 0.00, such insignificant results was possible due to the large sample size. The
value of other estimates were found as below mentioned
Table No. 4.67. Confirmatory Factor Analysis
Model Estimate Obtain Value Standard Value

Chi-Square 109.43
Df 78
P-Value 0.00000 >.05
RMSEA 0.297 <.08
NFI 0.903 >.90
IFI 0.932 >.90
CFI 0.899 >.90
RMR 0.375 <.08
GFI 0.897 >.90
AGFI 0.941 >.90
157
TLI 0.924 >.90
PNFI 0.521 >.50
PGFI 0.496 >.50
Here values which indicate satisfactory adjustment for a model loan schemes of banks
(CMIN/DF), values 1.0 or 3.0 or, at most up to 5.0; for CFI values similar than .90
and for RMSEA, value lower than .06 or up to .08 from above table, it was evident
that‟s of the value of Normed Fit Index (NFI) was found .90, comparative fit index
(CFI) was found .899, IFI was found .932, goodness of fit index (GFI) was found
.897, Adjusted goodness of the fit index (AGFI) was found .941, Tucker-Lewis index
(TLI) was found .924, root mean square of residual and root means squares residual
(RMR) was found .375 and the value for root mean square error of approximation
(RMSEA) was found .297.
158
Fig. No. 12. CFA (Confirmatory Factors Analysis)
All value were under the satisfactory limit and shown a good model fit while the
value of parsimonious Normed fit index (PNFI) and Parsimonious goodness of fit
index (PCFI) were found high than the cutoff value which was possible because of
large sample size.
4.17. t-Test between Perception of Demographic Variables and Loan

4.17.1. Independent Samples Test between perception of Staff of HDFC Bank
and State Bank of India and Loan Schemes
T-test was applied to find out difference between perception of Staff of HDFC Bank
and State Bank of India for Loan Schemes of Banks.
159
H010. There is no relationship between perception of Staff of HDFC Bank and State
Bank of India on Loan Schemes of Banks.
Bank N Mean Std. Std. Error
Deviation Mean
Loan Schemes of HDFC Bank 112 82.5625 13.34017 1.26053
Banks State Bank of India 111 78.9550 14.20652 1.34842
Group statistics table explore the mean value of HDFC Bank‟s Staff individuals is
82.5625, for std. deviation is 13.34017 and for std error mean is 1.26053 whereas for
the mean value of State Bank of India‟s Staff individuals is 78.9550, for std deviation
is 14.20652 and for std error mean is 1.34842.

Test for
Equality of
Variances

(2- Difference Difference Interval of the
tailed) Difference
Lower Upper
Equal
variances .068 .794 1.955 221 .050 3.60755 1.84533 -.02915 7.24424
Loan assumed
Schemes Equal
of Banks variances
1.954 219.866 .050 3.60755 1.84585 -.03029 7.24538
not
assumed
Value of F is .068 (<.5) with significant level of .794 which means F value is not significant
hence Equal variances assumed. Value of T is 1.955 which is insignificant at .050 levels. It
means significant difference between perception of HDFC Bank‟s staff and State Bank of
India‟s Staff individuals on Loan Schemes of Bank. Hence the null hypothesis was
rejected by the results.
160
H011. There is no relationship between married and unmarried bank customer on

Loan Schemes of Banks.
Grade N Mean Std. Deviation Std. Error Mean
Clerk 181 80.9116 13.78538 1.02466
Loan Schemes of Banks
Manager 42 80.1429 14.35877 2.21561
table explore the mean value of Clerk individuals is 80.9116, for std deviation is
13.78538 and for std error mean is 1.02466 whereas for the mean value of Managers
individuals is 80.1429, for std deviation is 14.35877 & for std error mean is 2.21561.
Table No.6.71. Independent Samples Test
Levene's Test t-test for Equality of Means
for Equality
of Variances

tailed) Difference
Lower Upper
Equal variances
Loan .223 .637 .323 221 .747 .76875 2.37959 -3.92084 5.45833
assumed
Schemes of
Equal variances
Banks .315 59.791 .754 .76875 2.44107 -4.11448 5.65197
not assumed
levels. It means no significant difference between perception of Clerks and Managers
individuals on Loan Schemes of Bank. Hence the null hypothesis was accepted by the
results.
161
4.17.3. Independent Samples Test between Perception Gender and Loan Schemes
of Banks by Staff
T-test was applied to find out difference between perception of Male and Female staff
H012. There is no relationship between perception of Male and Female staff on Loan
Schemes of Banks.
Gender N Mean Std. Deviation Std. Error Mean
Male 118 81.4746 14.88802 1.37055
Female 105 79.9714 12.64185 1.23372
Group statistics table explore the mean value of perception of Male Staffs individuals
is 81.4746, for std deviation is 14.88802 and for std error mean is 1.37055 whereas for
the mean value of perception of Female Staffs s individuals is 79.9714, for std
deviation is 12.64185 and for std error mean is 1.23372.
for Equality
of Variances

tailed) Difference
Lower Upper
Equal variances
Loan 3.282 .071 .807 221 .420 1.50315 1.86163 -2.16566 5.17196
assumed
Schemes of
Equal variances
Banks .815 220.532 .416 1.50315 1.84404 -2.13105 5.13734
not assumed
Value of F is 3.282 significant level of .071 (P>0.05) with which means F value is not
levels. It means no significant difference between perception of Male and Female Staff
results.
162
Schemes of Banks by Staff
staff for Loan Schemes of Banks.
H013. There is no relationship between perception of married and unmarried staff on


Marital N Mean Std. Std. Error
Status Deviation Mean
Loan Schemes of Married 147 81.4898 13.96148 1.15152
Banks Unmarried 76 79.3684 13.66049 1.56697
Group statistics table explore the mean value of perception of married staffs
individuals is 81.4898, for std. deviation is 13.96148 and for std. error mean is
1.15152 whereas for the mean value of perception of unmarried staffs individuals is
79.3684, for std deviation is 13.66049 and for std error mean is 1.56697.
for Equality of
Variances

tailed) Difference
Lower Upper
Equal variances
Loan .455 .501 1.083 221 .280 2.12137 1.95818 -1.73772 5.98047
assumed
Schemes
Equal variances
of Banks 1.091 154.702 .277 2.12137 1.94458 -1.71998 5.96273
not assumed
significant hence Equal variances assumed. Value of T is 1.083 which is insignificant at .280
levels. It means no significant difference between perception of married and unmarried staffs
results.
163
4.17.5. Independent Samples Test between Perception Residential and Loan
T-test was applied to find out difference between perception of urban and rural staff
H014. There is no relationship between perception of urban and rural staff on Loan
Schemes of Banks
Residential N Mean Std. Std. Error
Area Deviation Mean
Loan Schemes of Urban 170 80.9647 14.19763 1.08891
Banks Rural 53 80.1321 12.84978 1.76505
Group statistics table explore the mean value of perception of Urban staffs individuals
is 80.9647, for std deviation is 14.19763 and for std error mean is 1.08891 whereas for
the mean value of perception of Rural staffs individuals is 80.1321, for std deviation is
12.84978 and for std error mean is 1.76505.
for Equality of
Variances
F Sig. t df Sig. (2- Mean Std. Error 95% Confidence Interval

tailed) Difference Difference of the Difference
Lower Upper
Equal
-
variances .693 .406 .381 221 .704 .83263 2.18556 5.13984
Loan 3.47458
assumed
Schemes of
Equal
Banks -
variances not .401 94.886 .689 .83263 2.07392 4.94994
3.28468
assumed
levels. It means no significant difference between perception of Urban and Rural staffs
164
results.
4.17.6. Independent Samples Test between Perception Job Types and Loan
T-test was applied to find out difference between perception of Permanent and
Temporary staff for Loan Schemes of Banks.
H015. There is no relationship between perception of Permanent and Temporary staff

on Loan Schemes of Banks

Types of N Mean Std. Std. Error
Job Deviation Mean
Loan Schemes of Permanent 110 79.6091 14.44578 1.37735
Banks Temporary 113 81.8938 13.24317 1.24581
Group statistics table explore the mean value of perception of permanent staffs
individuals is 80.9647, for std deviation is 14.19763 and for std error mean is 1.08891
whereas for the mean value of perception of Temporary staffs individuals is 80.1321,
for std deviation is 12.84978 and for std error mean is 1.76505.

for Equality
of Variances

tailed) Difference
Lower Upper
Equal variances
Loan .344 .558 -1.232 221 .219 -2.28471 1.85501 -5.94049 1.37106
assumed
Schemes
Equal variances
of Banks -1.230 218.183 .220 -2.28471 1.85719 -5.94504 1.37561
not assumed
significant hence Equal variances assumed. Value of T is -1.232 which is insignificant at .219
165
levels. It means no significant difference between perception of permanent and temporary
staffs individuals on Loan Schemes of Bank. Hence the null hypothesis was accepted by
the results.
4.18. ANOVA Test of Equality of Error Variance
H016. There is no relationship between Age groups of bank staff on Loan Schemes of
Banks.
H017. There is no relationship between Educations of bank staff on Loan Schemes of

Banks.
H018. There is no relationship between Monthly Income of bank staff on Loan

Schemes of Banks.
H019. There is no relationship between Religion levels of bank staff on Loan Schemes
of Banks.
Table No. 4.80. Levene's Test of Equality of Error Variancesa

Dependent Variable: Loan Schemes of Banks
F df1 df2 Sig.
2.217 122 100 .000
Tests the null hypothesis that the error variance of the dependent variable is equal
across groups.
a. Design: Intercept + Age + Education + MonthlyIncome + Reliagion + Age *
Education + Age * MonthlyIncome + Age * Reliagion + Education *
MonthlyIncome + Education * Reliagion + MonthlyIncome * Reliagion + Age *
Education * MonthlyIncome + Age * Education * Reliagion + Age *
MonthlyIncome * Reliagion + Education * MonthlyIncome * Reliagion + Age *
Education * MonthlyIncome * Reliagion
Levene's Test of Equality of Error Variances is calculated to evaluate homogeneity of

groups on error variances. The value of F found insignificant at .150 levels of
significance which is unacceptable. The result has no importance for selecting Post
hoc test as the appropriate tests applicable on groups having different variances on the
variables of test will be selected while selecting options.
166
4.18.1. Tests of Between-Subjects Effects of different groups
The model used for ANOVA has bad fit as indicated by adjusted R square value of
0.522 the model fit was tested using F test and the value of F was found to be .894
which is insignificant at .723% level of significance, indicating that 52.2 % errors in
the dependent variable are explained and overall model fit was low.
Table No. 4.81. Tests of Between-Subjects Effects

Source Type III Sum of Squares df Mean Square F Sig.

a
Corrected Model 22270.960 122 182.549 .894 .723
Intercept 308724.025 1 308724.025 1512.692 .000
Age 993.854 4 248.463 1.217 .308
Education 1187.037 4 296.759 1.454 .222
Monthly Income 200.228 5 40.046 .196 .963
Religion 490.849 4 122.712 .601 .663
Age * Education 398.018 6 66.336 .325 .922
Age * Monthly Income 995.693 7 142.242 .697 .674
Age * Religion 1330.628 10 133.063 .652 .766
Education * Monthly Income 2170.266 12 180.856 .886 .563
Education * Religion 1052.918 12 87.743 .430 .948
Monthly Income * Religion 3121.021 16 195.064 .956 .510
Age * Education * Monthly Income 662.671 6 110.445 .541 .776
Age * Education * Religion 2938.400 6 489.733 2.400 .033
Age * Monthly Income * Religion 133.719 4 33.430 .164 .956
Education * Monthly Income * Religion 2221.609 6 370.268 1.814 .104
Age * Education * Monthly Income * Religion .000 0 . . .
Error 20408.914 100 204.089
Total 1497371.000 223
Corrected Total 42679.874 222
a. R Squared = .522 (Adjusted R Squared = -.062)
4.18.2. Post Hoc
In statistics, the Bonferroni correction is a method used to counteract the problem of

multiple comparisons. It is considered the simplest and most conservative method to
control the family wise error rate. It is named after Italian mathematician Carlo Emilio
Bonferroni for the use of Bonferroni inequalities, but modern usage is credited to
Olive Jean Dunn, who first used it in a pair of articles written in 1959 and 1961
167
Bonferroni
(I) Age Groups (J) Age Groups Mean Difference (I-J) Std. Error Sig. 95% Confidence Interval
Lower Bound Upper Bound
28-37 3.3689 2.25263 1.000 -3.0977 9.8354
38-47 4.6143 3.03852 1.000 -4.1082 13.3368

18-27
48-57 5.0000 7.37725 1.000 -16.1775 26.1775
57 & Above 8.5000 10.26869 1.000 -20.9778 37.9778

18-27 -3.3689 2.25263 1.000 -9.8354 3.0977
38-47 1.2454 2.73934 1.000 -6.6183 9.1091
28-37
48-57 1.6311 7.25914 1.000 -19.2073 22.4696
57 & Above 5.1311 10.18418 1.000 -24.1041 34.3664
18-27 -4.6143 3.03852 1.000 -13.3368 4.1082
28-37 -1.2454 2.73934 1.000 -9.1091 6.6183
38-47
48-57 .3857 7.54012 1.000 -21.2593 22.0308
57 & Above 3.8857 10.38632 1.000 -25.9298 33.7012
18-27 -5.0000 7.37725 1.000 -26.1775 16.1775
28-37 -1.6311 7.25914 1.000 -22.4696 19.2073
48-57
38-47 -.3857 7.54012 1.000 -22.0308 21.2593
57 & Above 3.5000 12.37202 1.000 -32.0158 39.0158
18-27 -8.5000 10.26869 1.000 -37.9778 20.9778
28-37 -5.1311 10.18418 1.000 -34.3664 24.1041

57 & Above
38-47 -3.8857 10.38632 1.000 -33.7012 25.9298
48-57 -3.5000 12.37202 1.000 -39.0158 32.0158

Bonferroni correction is the basic type of simultaneous inference that aims to control
the family wise error rate. A significant statistical research was done in the field from
early 1960s until late '90s, and many improvements were offered. Most notably are
the Holm–Bonferroni method, which offers a uniformly more powerful test procedure
(i.e., more powerful regardless of the values of the unobservable parameters), and the
Hochberg (1988) method, guaranteed to be no less powerful and is in many cases
more powerful when the tests are independent (and also under some forms of positive
dependence).
The Bonferroni correction can be somewhat conservative if there are a large number
of tests and/or the test statistics are positively correlated. Bonferroni correction
controls the probability of false positives only. The correction ordinarily comes at the
168
cost of increasing the probability of producing false negatives, and consequently
reducing statistical power. When testing a large number of hypotheses, this can result
in large critical values. In addition, in certain situations where one wants to retain, not
reject, the null hypothesis, and then Bonferroni correction is non-conservative. Hence
the null hypothesis H016 was accepted by the test.

Bonferroni
(I) Education Level (J) Education Level Mean Difference Std. Sig. 95% Confidence Interval
(I-J) Error Lower Upper
Bound Bound
Graduation 4.9032 3.49246 1.000 -5.1224 14.9288
Post Graduate -1.7571 2.79773 1.000 -9.7884 6.2742

Higher Secondary Higher than Post
.7273 2.65858 1.000 -6.9046 8.3591
Graduation
Other -8.5227 7.46061 1.000 -29.9395 12.8941

Higher Secondary -4.9032 3.49246 1.000 -14.9288 5.1224
Post Graduate -6.6603 3.27838 .448 -16.0714 2.7508
Graduation Higher than Post
-4.1759 3.16046 1.000 -13.2485 4.8966
Graduation
Other -13.4259 7.65383 .825 -35.3974 8.5456
Higher Secondary 1.7571 2.79773 1.000 -6.2742 9.7884
Graduation 6.6603 3.27838 .448 -2.7508 16.0714
Post Graduate Higher than Post
2.4844 2.37034 1.000 -4.3201 9.2888
Graduation
Other -6.7656 7.36282 1.000 -27.9017 14.3705
Higher Secondary -.7273 2.65858 1.000 -8.3591 6.9046
Higher than Post Graduation 4.1759 3.16046 1.000 -4.8966 13.2485
Graduation Post Graduate -2.4844 2.37034 1.000 -9.2888 4.3201
Other -9.2500 7.31108 1.000 -30.2376 11.7376
Higher Secondary 8.5227 7.46061 1.000 -12.8941 29.9395
Graduation 13.4259 7.65383 .825 -8.5456 35.3974

Other Post Graduate 6.7656 7.36282 1.000 -14.3705 27.9017
Higher than Post

9.2500 7.31108 1.000 -11.7376 30.2376
Graduation

169
Bonferroni
(I) Income (J) Income Mean Difference (I- Std. Sig. 95% Confidence Interval
Monthly Monthly J) Error Lower Upper
Bound Bound
10001-20000 -.2418 2.21418 1.000 -6.9006 6.4170
20001-30000 2.5114 4.76740 1.000 -11.8259 16.8486
0-10000 30001-40000 1.4729 4.24480 1.000 -11.2927 14.2385
40001-50000 3.0883 4.24480 1.000 -9.6773 15.8539
50000 & Above -.5386 3.53888 1.000 -11.1813 10.1040

0-10000 .2418 2.21418 1.000 -6.4170 6.9006
20001-30000 2.7532 4.79503 1.000 -11.6672 17.1735
10001-20000 30001-40000 1.7147 4.27581 1.000 -11.1442 14.5736
40001-50000 3.3301 4.27581 1.000 -9.5288 16.1889
50000 & Above -.2968 3.57601 1.000 -11.0512 10.4575
0-10000 -2.5114 4.76740 1.000 -16.8486 11.8259
10001-20000 -2.7532 4.79503 1.000 -17.1735 11.6672
20001-30000 30001-40000 -1.0385 6.00900 1.000 -19.1096 17.0327
40001-50000 .5769 6.00900 1.000 -17.4942 18.6481
50000 & Above -3.0500 5.53294 1.000 -19.6895 13.5895
0-10000 -1.4729 4.24480 1.000 -14.2385 11.2927
10001-20000 -1.7147 4.27581 1.000 -14.5736 11.1442
30001-40000 20001-30000 1.0385 6.00900 1.000 -17.0327 19.1096
40001-50000 1.6154 5.60342 1.000 -15.2361 18.4668
50000 & Above -2.0115 5.08956 1.000 -17.3176 13.2945
0-10000 -3.0883 4.24480 1.000 -15.8539 9.6773
10001-20000 -3.3301 4.27581 1.000 -16.1889 9.5288
40001-50000 20001-30000 -.5769 6.00900 1.000 -18.6481 17.4942
30001-40000 -1.6154 5.60342 1.000 -18.4668 15.2361
50000 & Above -3.6269 5.08956 1.000 -18.9330 11.6792
0-10000 .5386 3.53888 1.000 -10.1040 11.1813
10001-20000 .2968 3.57601 1.000 -10.4575 11.0512
50000 & Above 20001-30000 3.0500 5.53294 1.000 -13.5895 19.6895
30001-40000 2.0115 5.08956 1.000 -13.2945 17.3176
40001-50000 3.6269 5.08956 1.000 -11.6792 18.9330

170
Bonferroni
(I) Religion (J) Religion Mean Difference Std. Error Sig. 95% Confidence Interval
(I-J) Lower Bound Upper Bound
Muslim -.8043 3.95934 1.000 -12.1702 10.5616
Sikh .2411 3.67584 1.000 -10.3110 10.7932

Christian
Hindu 1.4084 3.32334 1.000 -8.1317 10.9486
Other 2.7582 3.90528 1.000 -8.4526 13.9689

Christian .8043 3.95934 1.000 -10.5616 12.1702
Sikh 1.0455 3.38251 1.000 -8.6645 10.7555
Muslim
Hindu 2.2128 2.99569 1.000 -6.3868 10.8123
Other 3.5625 3.63053 1.000 -6.8595 13.9845
Christian -.2411 3.67584 1.000 -10.7932 10.3110
Muslim -1.0455 3.38251 1.000 -10.7555 8.6645
Sikh
Hindu 1.1673 2.60951 1.000 -6.3237 8.6583
Other 2.5170 3.31906 1.000 -7.0108 12.0449
Christian -1.4084 3.32334 1.000 -10.9486 8.1317
Muslim -2.2128 2.99569 1.000 -10.8123 6.3868
Hindu
Sikh -1.1673 2.60951 1.000 -8.6583 6.3237
Other 1.3497 2.92386 1.000 -7.0436 9.7431
Christian -2.7582 3.90528 1.000 -13.9689 8.4526
Muslim -3.5625 3.63053 1.000 -13.9845 6.8595

Other
Sikh -2.5170 3.31906 1.000 -12.0449 7.0108
Hindu -1.3497 2.92386 1.000 -9.7431 7.0436

4.19. Regression Analysis
4.19.1. Linear Regression between Customer Support System and Loan Schemes
of Banks
The regression was calculated by taking the total of Customer Support System and
Loan Schemes of Banks by using SPSS software. In this the Customer Support
System was independent variable and Loan Schemes of Banks was the dependent
variable. Therefore, regression is calculated by taking dependent and independent
variable.
171
Null Hypothesis (H020): It stated that there is no significant impact of Customer
Support System on Loan Schemes of Banks.
Alternative Hypothesis (H20): It stated that there is no significant impact of Customer

Support System on Loan Schemes of Banks.
1 .806a .650 .648 8.22373 2.046
a. Predictors: (Constant), Customer Support System

Customer Support System and Loan Schemes of Banks. Durbin Watson value i.e.
2.046 in under standard value 1-3 indicating that there is no auto correlation.
Table No.4.87. ANOVAb
1 Regression 27733.710 1 27733.710 410.082 .000a
Residual 14946.165 221 67.630
Total 42679.874 222
a. Predictors: (Constant), Customer Support System

1 (Constant) 27.269 2.699 10.105 .000
Customer Support System 2.958 .146 .806 20.250 .000
172
Customer Support System and Loan Schemes of Banks.
4.19.2. Linear Regression between Trustworthy Employee Commitment and

The regression was calculated by taking the total of Trustworthy Employee

Commitment and Loan Schemes of Banks by using SPSS software. In this the
Trustworthy Employee Commitment was independent variable and Loan Schemes of
Banks was the dependent variable. Therefore, regression is calculated by taking
dependent and independent variable.
Null Hypothesis (H021): It stated that there is no significant impact of Trustworthy

Employee Commitment on Loan Schemes of Banks.
Alternative Hypothesis (H21): It stated that there is no significant impact of

Trustworthy Employee Commitment on Loan Schemes of Banks.
173
1 .862a .743 .742 7.04815 2.120
a. Predictors: (Constant), Trustworthy Employee Commitment
Trustworthy Employee Commitment and Loan Schemes of Banks. Durbin Watson
value i.e. 2.120 in under standard value 1-3 indicating that there is no auto correlation.
1 Regression 31701.384 1 31701.384 638.158 .000a
Residual 10978.490 221 49.676
Total 42679.874 222
a. Predictors: (Constant), Trustworthy Employee Commitment
Table No 4.91. Coefficientsa

Std.
Model B Error Beta t Sig.
1 (Constant) 26.747 2.190 12.214 .000
Trustworthy Employee 2.598 .103 .862 25.262 .000
Commitment
Trustworthy Employee Commitment and Loan Schemes of Banks.
174
Graphs No.14. Histogram and Normal P-P Plot of Regression
4.19.3. Linear Regression between Tenure of Loan and Interest and Loan
Schemes of Banks
The regression was calculated by taking the total of Tenure of Loan and Interest and
Loan Schemes of Banks by using SPSS software. In this the Tenure of Loan and
Interest was independent variable and Loan Schemes of Banks was the dependent
variable.
Null Hypothesis (H022): It stated that there is no significant impact of Tenure of Loan
and Interest on Loan Schemes of Banks.
Alternative Hypothesis (H22): It stated that there is no significant impact of Tenure of

Loan and Interest on Loan Schemes of Banks.
1 .804a .647 .645 8.25975 1.768
a. Predictors: (Constant), Tenure of Loan and Interest

175
1 Regression 27602.497 1 27602.497 404.590 .000a
Residual 15077.377 221 68.223
Total 42679.874 222
a. Predictors: (Constant), Tenure of Loan and Interest
1 (Constant) 22.893 2.930 7.814 .000
Tenure of Loan and Interest 2.626 .131 .804 20.114 .000
Tenure of Loan & Interest and Loan Schemes of Banks.
Graphs No.15. Histogram and Normal P-P Plot of Regression
176
4.19.4. Linear Regression between Positive Loan Advertising and Loan Schemes
of Banks
The regression was calculated by taking the total of Positive Loan Advertising and
Loan Schemes of Banks by using SPSS software. In this the Positive Loan
Advertising was independent variable and Loan Schemes of Banks was the dependent
variable.
Null Hypothesis (H023): It stated that there is no significant impact of Positive Loan
Alternative Hypothesis (H23): It stated that there is no significant impact of Positive

Loan Advertising on Loan Schemes of Banks.
1 .685a .469 .466 10.12970 1.750
a. Predictors: (Constant), Positive Loan Advertising

he obtain value of R Square is 0.469 indicating that 46.9% relationship between

1 Regression 20002.875 1 20002.875 194.939 .000a
Residual 22676.999 221 102.611
Total 42679.874 222
a. Predictors: (Constant), Positive Loan Advertising

177
1 (Constant) 44.083 2.714 16.246 .000
Positive Loan Advertising 3.698 .265 .685 13.962 .000
Positive Loan Advertising and Loan Schemes of Banks.
4.19.5. Linear Regression between Cleanliness and Well Dressed and Loan
Schemes of Banks
The regression was calculated by taking the total of Cleanliness and Well Dressed and
Loan Schemes of Banks by using SPSS software. In this the Cleanliness and Well
Dressed was independent variable and Loan Schemes of Banks was the dependent
variable.
178
Null Hypothesis (H010): It stated that there is no significant impact of Cleanliness and
Well Dressed on Loan Schemes of Banks.
Alternative Hypothesis (H10): It stated that there is no significant impact of

Cleanliness and Well Dressed on Loan Schemes of Banks.

1 .676a .458 .455 10.23531 1.665
a. Predictors: (Constant), Cleanliness and Well Dressed
1 Regression 19527.587 1 19527.587 186.400 .000a
Residual 23152.287 221 104.761
Total 42679.874 222
a. Predictors: (Constant), Cleanliness and Well Dressed

1 (Constant) 43.347 2.825 15.343 .000
Cleanliness and Well Dressed 3.766 .276 .676 13.653 .000
Cleanliness and Well Dressed and Loan Schemes of Banks.
179
4.20. Frequency Analysis

4.20.1. Frequency Analysis for Overall Satisfaction for SBI
The main aim of this frequency analysis is to examine overall loan customer
satisfaction towards different service of SBI. There are two major categories in this
analysis i.e. yes and no if the customer happy with bank facility they choose yes and
remaining customer who not satisfied they select no.
Frequency Percent
Yes 61 54.5
No 51 45.5
Valid
Total 112 100.0
The values in the entire table clearly shown that overall satisfaction with present
personal loan schemes of SBI quite good because only 54.5% customers of bank are
happy with present personal loan schemes of bank and some people willing to take
loan in future also but there are remaining 45.5% customer of the SBI are not happy
with current schemes from the bank authority side.
180
4.20.2. Frequency Analysis for Overall Satisfaction for HDFC
Similarly the main aim of this frequency analysis is to examine overall loan customer
satisfaction towards different service of HDFC. The data collected into two major
categories in this analysis i.e. yes and no if the customer happy with bank facility they
choose yes and remaining customer who not satisfied they select no.
Frequency Percent
Yes 86 76.8
No 26 23.2
Valid
Total 112 100.0
The HDFC customer are very happy with their present personal schemes of bank
because the survey we found 76.8% people say yes the personal loan schemes
inclusive other facility HDFC banks have maximum satisfied customers tested
through values in the entire table of frequency analysis. According to HDFC authority
they have maximum satisfied customers and they willing to take loan in future also.
181

09 - Chapter 4

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

09 - Chapter 4

Uploaded by

Copyright:

Available Formats

Chapter – 4

A. Data Analysis for Customer Questionnaire

4.1. Detail of Demographic Profile of Respondents

4.1.1. Gender-wise Frequency Distribution

Table No.4.1. Gender-wise Frequency Distribution

4.1.2. Age Groups-wise Frequency Distribution

Total 415 100.0

4.1.3. Residential Area-wise frequency Distribution

According to residential area of bank customer-wise frequency distribution, in below

Table No. 4.3. Residential Area

4.1.4. Education Level-wise frequency Distribution

Table No.4.4. Education Level

4.1.5. Income Monthly-wise frequency Distribution

According to monthly income-wise frequency distribution we found the income

4.1.6. Occupation-wise frequency Distribution

Total 415 100.0

4.1.7. Religion-wise frequency Distribution

Total 415 100.0

4.1.8. Marital Status-wise frequency Distribution

Total 415 100.0

4.2. Consistency & Reliability Measurements

Table No. 4.9. Item-Total Statistics

4.2.2. Reliability Test

Table No. 4.10. Reliability Statistics

4.3. Normality Analysis

Table No. 4.11. Tests of Normality

4.4. Factor Analysis

Factor analysis is a statistical method used to describe variability among observed,

Exploratory factor analysis (EFA) is used to identify complex interrelationships

4.4.1. K. M. O. measurement of sampling Adequacy and Bartlett’s Test of

4.4.2. Description of Factors

i. Factor One: Service Quality

ii. Factor Two: Employees Commitment

iii. Factor Three: Safety and Security

iv. Factor Four: Ease Process

v. Factor Five: Sanitation Facility

vi. Factor Six: Promotion through Advertisements (Ads)

Fig. 02. Factor Model of Loan Schemes of Bank

4.4.3. Scree Plot

4.4.4. Component Score Covariance Matrix

Table No. 4.14. Component Transformation Matrix

Service Quality .515 .480 .463 .322 .326 .282

Extraction Method: Principal Component Analysis.

4.5. Structure Equation Modeling

SEM begins with the specification of a model to be estimated. A model is a statistic

4.5.1. Goodness of Indices

4.5.3. Confirmatory Factor Analysis of Corporate Social Responsibility

Model Estimate Obtain Value Standard Value

Service Employee Safety Ease Sanitation Promotion

Table No. 4.17. Group Statistics

4.7.3. Independent Samples Test between Perception Residential Area of People

Table No. 4.21. Group Statistics

Levene's t-test for Equality of Means

4.7.4. Independent Samples Test between Perception of Family and Loan

4.7.5. Independent Samples Test between Perception of Head of Family and

Table No. 4.26. Independent Samples Test

Levene's t-test for Equality of Means

H06. There is no relationship between Age groups of bank customers on Loan

H07. There is no relationship between Occupations of bank customers on Loan

H08. There is no relationship between Monthly Income of bank customers on Loan

H09. There is no relationship between Education levels of bank customers on Loan

Table No. 4.27. Levene's Test of Equality of Error Variancesa

Dependent Variable: Loan Schemes of Bank