Professional Documents
Culture Documents
7.1 INTRODUCTION
This chapter deals with the analysis and interpretation of the impact of perceived
service quality of Omni-channel e- fulfilment. The data were collected from700
respondents of IT Professionals in Chennai. The responses of the respondents obtained
through the structured questionnaire have been compiled and analyzed using the following
statistical tools in tune with the objectives of the study
1. Correlation analysis
2. Exploratory Factor Analysis (EFA)
3. Confirmatory Factor Analysis (CFA)
4. Structural Equation Modeling (SEM)
5. Multiple Regression Analysis
Interpretation has been made on the basis of the above-mentioned analyses.
Factor analysis is one of the most often used multivariate techniques of research
studies. It is a technique applicable when there is a systematic interdependence among the set
of observed or manifest variables and it would be interesting to find out something more
fundamental or latent, which creates this communality. It seeks to resolve a large set of
measured variables in terms of relatively low categories known as factors. Factor analysis and
reliability tests are applied to ascertain the important factors and to test the reliability of the
data.
132
7.3 ANALYSIS OF PERCEIVED SERVICE QUALITY OF OMNI-CHANNEL
E-
This study explores the important factors that are determining the perceived service
quality of Omni-channel e- fulfilment in Chennai, as depicted in Table 7.1.
Table 7.1
Label Statements
FM7 All the products I requested were in good condition upon delivery
FM8 Omni-channel e-tailers provide delivery agent details with tracking facility.
133
Omni-channel e-tailers effectively communicate about order details and
FM13
shipment through e-mail/mobile.
Omni-channel e-tailers provide information about the return policy and refund
FM14
the amount through e-mail/mobile.
The Omni-channel e-tailer accurately fulfiled my order with one and only one
FM15
delivery attempt.
Order tracking facility is very useful to track the order and also help to
FM17
determine the product delivery status.
Table 7.2
Correlation Matrix
134
Table 7.2
Correlation Matrix ( )
Table 7.3
Sig. .000
Table 7.4
RELIABILITY STATISTICS
N % No. of Items
135
Table 7.5
COMMUNALITIES
136
Table 7.6
137
Table 7.7
Component
1 2 3 4 5 6
Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser
Normalization. a. Rotation converged in 4 iterations
138
Table 7.8
Component Eigen
Variance
1 2 3 4 5 6 value Alpha
FM2 .966
FM8 .960
FM11 .883
FM17 .885
FM5 .974
FM15 .965
3.882 21.565 0.987
FM16 .991
FM18 .993
FM12 .965
FM14 .956
FM1 .985
1.908 10.602 0.972
FM7 .984
FM6 .979
1.887 10.485 0.964
FM9 .981
In factor analysis, the analytical process is based on a matrix of correlation among the
variables. Valuable insights can be gained from an examination of this matrix. If the factors
analysis is to be proper, the variables must be correlated. If the correlation among all the
139
variables is very low and negligible, the factor analysis may not be appropriate. In the present
study, Inter Correlation Matrix shown in Table 7.2 reveals that the correlation among most of the
variables is in good fit and hence the factor analysis is very appropriate for analyzing the
perceived fulfilment service quality of Omni-channel e-tailers in Chennai. The result of the
fitness test regarding factor analysis, based on Inter Correlation Matrix has been presented in
Table 7.2. The KMO (Kaiser-Meyer-Olkin) measure of sampling adequacy has been computed
to determine the suitability of using factor analysis. The values between 0.5 and 1.0 indicate that
factor analysis is suitable or appropriate. The KMO test presented in Table 7.3 indicates that the
calculated value of KMO is 0.766, which shows that the sample is adequate to conduct
Exploratory Fa a significant number of
correlations among the statements. Thus, all the parameters discussed above, support the
application of factor analysis on the data. The scale has also been tested for reliability and the
Table 7.4. Communalities
examined to determine the suitability of data for analysis, (Hair et al., 2010) suggested that
variables with loadings greater than 0.45 are practically significant and support acceptable levels
of explanation. Hence, the criteria of 0.45 have been considered for selecting the variables.
Communalities of 18 statements shown in Table 7.5, explain the communalities of all the 18
variables and measure that the amount of variance as a variable, shares with all the other
A large communality means a large amount of the variance, a variable has extracted by the factor
solution. It shows that variables with a comparatively higher value are well-represented in the
common factor space, while the low-value variables are not. Thus, Table 7.5 indicates that the
extracted communalities are high and acceptable for all the variables. It can be seen that the
Exploratory Factor Analysis revealed three underlying dimensions for perceived fulfilment
service quality of Omni-channel e-tailer. These three factors explain 24.468% of total variance.
On the basis of the rotated component matrix, the statements are categorized under respective
factors as shown in Table 7.6. Eigen values for Factor1, Factor2, and Factor3 are 4.404, 3.882
and 2.704 respectively.
140
7.5 NAMING OF THE FACTORS
The naming of the factors has been done on the basis of variables represented in each
case. The explanation of the extracted factors is given below:
Factor 1: ACCESSIBILITY
141
7.6 PERCEIVED SERVICE QUALITY OF OMNI-CHANNEL E-
FULFILMENT SERVICES CONFIRMATORY FACTOR ANALYSIS APPROACH
Confirmatory Factor Analysis (CFA) is used to study the relationships between a set of
observed variables and a set of continuous latent variables. When the observed variables are
categorical, CFA is also referred to as Item Response Theory (IRT) analysis. It provides a more
explicit framework for confirming prior notions about the structure of the domain of contents. It
adds the ability to test constraints on the parameters of the factor model to the methodology of
EFA. With CFA, it is possible to place substantively meaningful constraints on the factor model.
CFA allows testing hypotheses about a particular factor structure. Factor loading between the
first factor and the first observed variable is zero. Unlike EFA, CFA produces many goodness-
of-fit measures to evaluate the model, but it does not calculate factor scores. CFA requires
special purpose software packages like AMOS 23.0. The new 19 variables have been analyzed
with AMOS 23.0 software package, in order to analyze perceived service quality of Omni-
channel e-tailers fulfillment towards Omni-channel e-tailing in Chennai.
Measurement model analysis has been carried out to test the validity and reliability of the
constructs, which build the conceptual model. For this purpose, Confirmatory Factor Analysis
(CFA) has been run using a maximum likelihood estimation technique on AMOS 23.0 software.
The validity of hypothesized model has been assessed in two stages. First of all Goodness-of-fit
(GOF) indices have been examined. After that, construct validity and reliability of the
measurement model have been evaluated. Goodness-of-fit determines the degree to which the
hypothesized model fits the sample data. Multiple fit indices have been used to assess goodness
of fit: Goodness-of-fit (GFI), Adjusted Goodness-of-fit (AGFI), Comparative Fit Index (CFI),
Tucker- -Bonnet Normed Fit
Index (NFI), Parsimonious Normed Fit Index (PNFI), Parsimony Comparative Fit Index (PCFI)
and Incremental Fit Index (IFI). Convergent validity and discriminant validity of the construct
has been tested to examine construct validity.
142
Table 7.9
143
Figure 7.1
144
7.7 TESTING OF HYPOTHESES- STANDARDIZED ESTIMATES
The following Table represents the results of the testing of the hypotheses for measuring
the Fulfilment services offered by Omni-channel e-tailing in Chennai.
Table 7.10
Hypothetical
Hypotheses Results
Relationship
H1: There is a positive impact of Provided
alternatives/choices when the product out of stock and
FM2 Positive Supported
the fulfilment services offered by the Omni-channel e-
tailing in Chennai.
H2: There is a positive impact of provides delivery
agent details with tracking facility and the fulfilment
FM8 Positive Supported
services offered by the Omni-channel e-tailing in
Chennai.
H3: There is a positive impact of Overall checkout
experience from the Omni-channel e-tailers and the
FM10 Positive Supported
fulfilment services offered by the Omni-channel e-
tailing in Chennai.
H4: There is a positive impact of logical sequence of
steps for completing an order and the fulfilment
FM11 Positive Supported
services offered by the Omni-channel e-tailing in
Chennai.
H5: There is a positive impact of order tracking
facility and the fulfilment services offered by the FM17 Positive Supported
Omni-channel e-tailing in Chennai.
H6: There is a positive impact of products are
delivered directly at home and the fulfilment services FM5 Positive Supported
offered by the Omni-channel e-tailing in Chennai.
H7: There is a positive impact of fulfilment accuracy,
one delivery attempt and the fulfilment services FM15 Positive Supported
offered by the Omni-channel e-tailing in Chennai.
H8: There is a positive impact of On-time delivery and
the fulfilment services offered by the Omni-channel e- FM16 Positive Supported
tailing in Chennai.
H9: There is a positive impact of good post-purchase FM18 Positive Supported
145
experience and the fulfilment services offered by the
Omni-channel e-tailing in Chennai.
H10: There is a positive impact of good interaction
with customer care and the fulfilment services offered FM12 Positive Supported
by the Omni-channel e-tailing in Chennai.
H11: There is a positive impact of shipment details
sent to mobile/E-mail and the fulfilment services FM13 Positive Supported
offered by the Omni-channel e-tailing in Chennai.
H12: There is a positive impact of information about
return policy/refund sent to e-mail/mobile and the
FM14 Positive Supported
fulfilment services offered by the Omni-channel e-
tailing in Chennai.
Table 7.11
REGRESSION WEIGHTS
Measured Latent
Estimate S.E. R2 C.R. P
variable variable
FM17 <--- FAC1 .865 .033 0.82 26.225 ***
FM11 <--- FAC1 .865 .033 0.81 26.080 ***
FM10 <--- FAC1 .966 .029 0.94 33.384 ***
FM8 <--- FAC1 .989 .028 0.98 36.119 ***
FM2 <--- FAC1 .918 .027 0.95 33.496 ***
FM18 <--- FAC2 .991 .027 0.98 37.100 ***
FM16 <--- FAC2 .987 .028 0.99 36.818 ***
FM15 <--- FAC2 .979 .029 0.95 33.570 ***
FM5 <--- FAC2 .958 .028 0.96 34.662 ***
FM14 <--- FAC3 .964 .029 0.94 33.090 ***
FM13 <--- FAC3 .817 .028 0.88 29.174 ***
FM12 <--- FAC3 .969 .028 0.97 34.471 ***
*** Significant at 1% level
146
Table 7.12
FM2 .918
FM8 .989
FM10 .966
FM11 .865
FM17 .865
FM5 .958
FM15 .979
FM16 .987
FM18 .991
FM12 .969
FM13 .817
FM14 .964
147
Table 7.13
Acceptable Calculated
Factors value value
CR AVE CR AVE
Accessibility 0.956 0.815 CR > AVE
Post-sale services > 0.7 > 0.5 0.987 0.951
Communication quality 0.950 0.864
Table 7.14
Table 7.15
148
From the above analysis, all the measured variables with the latent variable of successful
operation are disclosed, for analyzing the fulfilment services offered by the Omni-channel e-
tailing in Chennai and also they have a positive relationship with the significance at 1 percent
and 5 percent level. The analysis of the model suggests that all the measured variables have a
high impact on the fulfilment services offered by Omni-channel e-tailing in Chennai. Table 7.11
shows the regression coefficient of the exogenous variables. It is concluded that all the measured
variables are significant at 1 percent level. Hence, it is concluded that all the new variables have
a high impact on the fulfilment services offered by the Omni-channel e-tailers in Chennai. In
order to establish convergent validity of the construct, standardized factor loadings have been
reviewed, According to (Hair et al., 2010), for establishing convergent validity, standardized
loadings should be 0.50 or higher, average variance extracted (AVE) should be 0.50 or greater
and composite reliability (CR) should be 0.70 or more. All of them are found to be greater than
0.50. All the factor item loadings ranged from 0.817 to 0.9913 (Table 7.12). Thus, all are above
the minimum limit of 0.50. Average variance explained for three dimensions (Accessibility,
Post-sale services and Communication quality) are 0.815, 0.951 and 0.864 respectively. Thus, all
AVEs for each dimension of fulfilment services exceed the variance attributable to measurement
error (i.e., AVE>0.50). Thus, Convergent validity of three-dimensional fulfilment service scales
has been tested and established. It is the extent to which, a construct is truly uni-dimensional and
is distinct from other constructs. In other words, if the inter-correlations among the set of
indicators, which are supposed to measure the different latent constructs, are not too high. They
exhibit discriminant validity. The discriminant validity of the constructs has been computed and
depicted in Table 7.14.
From the above Table, it is shown that the entire test has a range of 0 to1. The
Comparative Fit Index (CFI) scores 0.966, Normed Fit Index (NFI) scores 0.962, Tucker-Lewis
Index (TLI) scores 0.954, Parsimonious Normed Fit Index (PNFI) scores 0.729, Parsimony
Comparative Fit Index (PCFI) scores 0.731, Relative Fit Index (RFI) scores 0.950, Incremental
Fit Index (IFI) scores 0.966, Adjusted Goodness of Fit Index (AGFI) scores 0.811 and Goodness
of Fit Index (GFI) scores 0.879. This shows that the new model is fit for further research and
indices to indicate a close fit of the model.
149
7.8 MULTIPLE REGRESSION ANALYSIS
In order to estimate the degree and the extent of the interrelationship between dependent
variables and the number of independent variables, multiple regression techniques are generally
used. To measure the impact of perceived service quality of Omni-channel e- fulfilment in
Chennai, the regression technique has been applied to compute R2 from the following model.
0 1X1 + 2 X2 n Xn + µ
The regression co-efficient and overall variations are tested respectively by computing t
value and F ratios. The goodness of fit of the estimated equation is worked out with the help of
R-squared and R-adjusted square values.
X1 = Accessibility
X2 = Post-sale services
X3 = Communication quality
150
7.9 MULTIPLE REGRESSION ANALYSIS PERCEIVED SERVICE QUALITY OF
OMNI-CHANNEL E-
Significance level
Table 7.16
MODEL SUMMARY
Table 7.17
ANOVA
Table 7.18
REGRESSION CO-EFFICIENT
Unstandardized Standardized
Model Coefficients Coefficients t Sig.
B Std. Error Beta
(Y1) 62.187 .089 696.712 .000
(X1) 5.307 .089 .690 59.410 .000
(X2) 4.151 .089 .540 46.476 .000
(X3) 2.870 .089 .373 32.135 .000
Source: Compiled from Collected data
151
Table R= 0.952
indicates a high degree of R2 = 0.906 , how much of the total
variation in the dependent variable can be explained by the independent variables. In this case,
90.6% can be explained, which means that the multiple regression explains high variance in the
data. Table 7.17 presents the ANOVA results, which show, how well the regression equation fits
the data (i.e., predicts the dependent variable). Table 7.17 clearly indicates that the regression
model predicts the dependent variable significantly well (p<0.05). Table 7.18 shows the multiple
regression coefficients including the intercept and the significance levels. Table 7.18 states that
Beta expresses the relative importance of each independent variable in standardized terms. The
regression co-efficient Table provides the intercept and the slope for the following regression
equation.
Regression fitted:
The analysis of variance of multiple regression models for Omni-channel fulfilment services
indicates that the overall significance of the model is well fitted. The co-efficient of
determination R2 value shows that these variables put together, explain the variations of
fulfilment services to the extent 95.6 %.
It is concluded that the p-value 0.000 is less than the 0.05, so the null hypothesis has been
. There exists enough evidence to conclude that the
fulfilment service quality regression line is not zero. Hence, Accessibility, Post-sale services and
Communication quality are useful as predictors of perceived services quality of Omni-channel e-
fulfilment on Omni-channel e-tailing.
7.10 SUMMARY
152