28/09/2021

Master’s Thesis in Banking Service Development Economics at Nam Viet Commercial Joint Stock Bank – 8

The most surveyed sample is long-term customers (over 3 years of using Navibank services, accounting for 36.7%).

And customers are relatively satisfied with Navibank’s , so 63.9% of respondents are willing to recommend Navibank services to others and will continue to transact with Navibank when there is a need in the future.

Table 2.7 : Descriptive statistics of satisfaction evaluation factors Descriptive Statistics

WOMEN Minimum Maximum Mean Std. Deviation
TT1 166 2 5 3.36 .689
TT2 166 1 5 3.20 .618
TT3 166 1 5 3.20 .674
TT4 166 2 5 3.42 .787
TT5 166 1 5 3.08 .713
PH1 166 2 5 3.31 .736
PH2 166 2 5 3.27 .818
PH3 166 2 5 3.28 .695
PH4 166 2 5 3.17 .609
DB1 166 1 5 3.13 .740
DB2 166 1 5 3.16 .708
DB3 166 2 5 3.25 .711
DB4 166 2 5 3.07 .666
CT1 166 2 5 2.94 .768
CT2 166 2 5 3.33 .708
CT3 166 2 5 3.27 .812
CT4 166 2 5 2.96 .664
HH1 166 1 5 2.73 .646
HH2 166 1 4 2.75 .655
HH3 166 1 5 2.82 .690
HH4 166 1 4 2.72 .659
HH5 166 2 5 2.71 .652
HL1 166 2 5 3.27 .689
HL2 166 2 5 3.16 .699
HL3 166 2 5 3.01 .713
Valid N (listwise) 166        

Maybe you are interested!

Through the statistical table describing the factors to assess the level of satisfaction, we see that the factors that are most satisfied by customers are TT4 – Navibank keeps information well for customers (3.42), TT1 – Navibank always performs fulfill its commitment to you about banking services (3.36).

Besides, empathy and tangible factors are the two factors that customers are least satisfied with such as whether CT1 – Navibank calls or sends greeting cards to you on birthdays, holidays, New Year… (2.94). ), CT4 – Regular sweepstakes or gift giving programs for you (2.96), HH4 – Convenient parking for customers (2.72), HH5 – Wide transaction network (2.71)

In summary, through the survey, we find that tangible factors and empathy are the two factors that are least satisfied by customers. To explain this problem, it is that Navibank does not really care about customers. Currently, Navibank only has a policy to care about customers with large deposit balances, depositing terms of three months or more. will receive birthday gifts, Tet gifts… The policy is that, but the implementation is not uniform, there are some PGDs still not implementing, or forgetting to implement. As for other customers, Navibank currently does not have any specific policies or programs to take care of them.

Currently, customer care at Navibank is only spontaneous, not really professional. Navibank also does not often have promotions, lucky draws and if there are, it is very monotonous, mainly for deposit services, sometimes there are promotions for card services, but promotions This does not bring much enjoyment to customers. Especially, Navibank rarely invests in giving gifts to customers. Customers have not found that their needs are understood, nor have they felt the importance of Navibank’s benefits for them.

It is possible that in the process of communicating with customers, Navibank staff have not thoroughly understood the needs of customers, they have only met specific needs from customers. Therefore, customers feel that Navibank is no different from other banks. Regarding the tangible factor, the parking space is not convenient. Currently, in most transaction offices, there is no parking space for customers, leading to inconvenience for customers, especially when dealing with large amounts of money. The fact that the transaction network is not widespread is also a disadvantage in terms of geography of Navibank.

2.6.3.2 Evaluation of the scale by Cronbach Alpha reliability coefficient:

The scales were evaluated through Cronbach’s Alpha coefficient and exploratory factor analysis (EFA). Cronbach’s Alpha coefficient must be between 0.6 and 1.0 to ensure that the variables in the same group of factors are significantly correlated.

According to the survey results obtained:

Trust consists of 05 observed variables from TT1 to TT5. After checking the reliability coefficient of Cronbach Alpha for the first time, the observed variable TT5 has a total correlation coefficient of less than 0.3, so it is excluded from the model. In the second test, all components have variables with a total correlation coefficient greater than 0.3, so all variables are accepted. The Cronbach Alpha coefficient is quite high from 0.6 and above, so the scales are satisfactory.

Table 2.8: Cronbach Alpha coefficient of the scale components

  Scale Mean if Item Deleted Scale Variance if Item Deleted Corrected Item-Total Correlation Cronbach’s Alpha if Item Deleted

Confidence (TT) when not eliminating variable TT5: Cronbach’s Alpha = 0.716

TT1 12.90 3.930 .478 .666
TT2 13.06 4.093 .494 .663
TT3 13.06 3.608 .643 .600
TT4 12.85 3.559 .516 .651
TT5 13.19 4.383 .274 .746

Confidence (TT) after eliminating variable TT5: Cronbach’s Alpha = 0.746

TT1 9.83 2.715 .526 .695
TT2 9.98 2.903 .522 .699
TT3 9.98 2.769 .516 .700
TT4 9.77 2.311 .606 .649

Feedback (PH): Cronbach’s Alpha = 0.775

PH1 9.72 2.968 .562 .730
PH2 9.76 2.766 .551 .743
PH3 9.74 2.848 .683 .667
PH4 9.86 3.349 .544 .742

Guarantee (DB): Cronbach’s Alpha = 0.750

DB1 9.49 2.760 .511 .713
DB2 9.45 2.746 .562 .683
DB3 9.36 2.850 .504 .715
DB4 9.54 2.771 .610 .659

Sympathy (CT): Cronbach’s Alpha = 0.750

CT1 9.56 2.478 .629 .531
CT2 9.17 2.893 .497 .623
CT3 9.23 3.026 .320 .740
CT4 9.54 2.977 .512 .618

Tangibility (HH): Cronbach’s Alpha = 0.783

HH1 11.00 4.012 .544 .747
HH2 10.98 4.266 .421 .786
HH3 10.91 3.998 .493 .764
HH4 11.01 3.685 .678 .701
HH5 11.02 3.727 .669 .705

Satisfaction (HL): Cronbach’s Alpha = 0.723

HL1 6.17 1.406 .588 .581
HL2 6.28 1.499 .500 .687
HL3 6.43 1.410 .545 .634

2.6.3.3 Evaluation of the scale by exploratory factor analysis EFA

Factor analysis explores independent variables: v A necessary condition for the application of factor analysis is that the variables must be correlated with each other (measured variables reflect different aspects of the same common factor). should use Bartlett test to consider whether the variables have correlation in the population and the sufficient condition to conduct factor analysis is that the KMO (Kaiser-Mayer-Olkin) value must be from 0.5 to 1 (followed by Hoang Trong and Chu Nguyen Mong Ngoc, 2005). Extraction of representative factors by observed variables is done by method of extracting Principal components with Varimax rotation and factor loading factor must be > 0.5. Components with Eighenvalues ​​> 1 and total variance extracted ≥ 0.5 are considered as representative variables.

Satisfaction scale is measured by 22 observed variables, after checking the reliability by Cronbach alpha analysis, the TT5 variable is removed, leaving 21 observed variables. EFA exploratory factor analysis was used to re-evaluate the degree of convergence of 21 observed variables according to the components.

The results of the first analysis show that the KMO coefficient is quite high at 0.792 and the significance level is zero, indicating that the factor analysis is appropriate. After using the method of factor extraction and rotation, the factor analysis method showed that CT3 has a load factor <0.5 (Appendix 3.1) so it was excluded from the analysis. we proceed to run the EFA again.

Table 2.9: Result of factor analysis for the 2nd time exploratory scale

KMO and Bartlett’s Test

Rotated Component Matrixa

  Component        
  1 2 3 4 5
HH4 .829        
HH5 .811        
HH1 .723        
HH3 .625        
HH2 .607        
DB4   .788      
DB2   .683      
DB1   .675      
DB3   .675      
PH3     .831    
PH2     .772    
PH4     .722    
PH1     .661    
TT2       .774  
TT4       .732  
TT3       .728  
TT1       .622  
CT4         .831
CT1         .801
CT2         .690

Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization. a. Rotation converged in 5 iterations.

After eliminating unsatisfactory variables, EFA exploratory factor analysis results in customer satisfaction remaining 5 components with 20 observed variables. The EFA analysis conditions are satisfied: KMO = 0.792>0.5, sig.<0.05, the factor upload coefficients are all >0.5. Total extracted variance = 60.148% (greater than 50%) shows that the extracted variance is satisfactory. Thus, with 5 components TT, PH, DB, CT, HH will explain 60.148% variation of the data.

Factor analysis explores the dependent variable

Table 2.10: Results of factor analysis to explore the satisfaction scale

 KMO and Bartlett’s Test

Component Matrixa

  Component
  1
HL1 .834
HL3 .805
HL2 .768

Extraction Method: Principal Component Analysis. a. 1 components extracted.

The EFA analysis conditions are satisfied:

– KMO = 0.671 > 0.5, Sig. < 0.05

– Total variance extracted = 64,420 > 50%

– The upload factors are all > 0.5

Extracted a single dependent factor with three observed variables HL1, HL2, HL3.

2.6.3.4 Multiple linear regression analysis, testing hypotheses and research models

Pearson correlation coefficient test:

Before conducting multiple linear regression analysis, we need to test the linear correlation between the variables in the model: between the dependent variable and each independent variable and between the independent variables. Using the Pearson correlation coefficient to quantify the closeness of the linear relationship between two quantitative variables: the closer the absolute value of Pearson’s coefficient is to 1, the more closely these two variables have a linear correlation. (Hoang Trong & Chu Nguyen Mong Ngoc, 2005).

Linear regression analysis:

The factors extracted in factor analysis are used for linear regression analysis to test the research model and accompanying hypotheses. Statistical hypothesis tests applied at the 5% level of significance.

The study performed multiple linear regression by Enter method: all variables were included once and related statistical results were considered.

The multiple regression equation for the proposed research model after preliminary research is as follows:

Y = 5*X5 + Eb4*X4+b3*X3+ b2*X2 + b1*X1+ b0 + b

Inside:

+Y is the dependent variable that represents the predictive value of customer satisfaction.

5 are regression coefficients b4, b3, b2, b1, b0 ,b+

+ X1, X2, X3, X4, X5 are the independent variables in the order of assuming: Trust, Feedback, Assurance, Empathy, Tangibility, Satisfaction.

+E are unknown factors

Test your hypotheses:

To test the fit of the model, the researchers use the coefficient of determination R2 (R-square) to assess the fit of the research model, the coefficient of determination R2 is proven to be a function of zero. decreases with the number of independent variables included in the model. Besides, it is necessary to check that there is no multicollinearity with the variance inflation factor (VIF): VIF<2. The higher the standardized beta coefficient of any variable, the greater its impact on customer satisfaction (following Hoang Trong and Chu Nguyen Mong Ngoc, 2005).

Table 2.11 : Results of running multiple linear regression

Variables Entered/Removedb

Model Variables Entered Variables Removed Method
1 PH, HH, CT, TT, DBa   Enter

a. All requested variables entered.

b. Dependent Variable: HL

Model Summary

Model R R Square Adjusted R Square Std. Error of the Estimate
1 .727a .528 .513 .39211

a. Predictors: (Constant), PH, HH, CT, TT, DB

ANOVAb

Model Sum of Squares Df Mean Square F Sig.
1 Regression Residual Total 27.500 24.600 52.100 5 160 165 5.500 .154 35.773 .000a

b. Dependent Variable: HL

Regression linear regression analysis method with 5 components of customer satisfaction included at the same time (enter) shows that the regression model is suitable to use to test the hypothesis (sig. = 0.000) . The level of explaining the relationship between components by this regression method gives the results R2 (R-quare) = 0.528 and adjusted R2 (Adjusted R-quare) = 0.513, showing that the model’s compatibility is 51.3. % or about 51.3% variance of satisfaction is explained by five components: trust, feedback, assurance, empathy, tangibles.

According to the correlation matrix (Table 2.13), it shows that the correlation coefficient between satisfaction variable – HL (dependant variable) with each independent variable is greater than 0.3, so the dependent variable has a linear correlation with the independent variables. create. The analysis results also show that some independent variables are correlated with each other, so when analyzing regression, it is necessary to pay attention to the problem of multicollinearity.

Table 2.12 : Pearson Correlations . test results

**. Correlation is significant at the 0.01 level (2-tailed).

Table 2.13: Results of regression coefficient analysis

Coefficientsa

a. Dependent Variable: HL

The variance inflation factor (VIF) of all factors is less than 2, showing that these independent variables are not closely related, so there is no multicollinearity phenomenon. Therefore, the relationship between the independent variables does not significantly affect the explanatory results of the regression model.

On the other hand, through Anova analysis, we see that the sig value is 0.000 < 0.05, which means that the hypothesis H0 is rejected (β1=β2=β3=β4=β5=0).

Regression coefficients with positive signs show that the factors in the regression model positively affect customer satisfaction. Regression coefficient of which factor is larger will have more impact on customer satisfaction. Therefore, the order of impact level of factors on customer satisfaction is as follows:

Table 2.14: Order of impact levels of factors on customer satisfaction

Number Beta coefficient Factor
first. 0.345 Guarantee
2. 0.262 Sympathy
3. 0.197 The tangible
4. 0.146 Feedback
5. 0.132 Trust

Thus, after running the regression and testing the statistical hypotheses, we draw the regression model as follows:

Satisfaction = 0.345* Assurance + 0.262* Empathy + 0.197 * Tangibles +0.146* Feedback + 0.132* Trust + E

2.6.3.5 Anova . test

Analyze the difference between the attributes of the research object and the dependent variable (Customer satisfaction) in the research model. A summary analysis table is presented below (for details see Appendix 5 – ANOVA Analysis)

Table 2.15: Analysis of differences by attributes of research subjects

Attributes Levene Statistics (sig.) Anova Analytics (sig.)
Sex .631 .993
Age .287 .372
Income .421 .001
Used Time .71 .000
Service Introduction .218 .000
Future needs .147 .001

Analysis of differences by sex

This test shows whether the variance of customer satisfaction is equal or different between Female and Male. Sig of Levene statistic = 0.631 (>0.05), so at 95% confidence level, hypothesis H0: “Equal variances” is accepted, and hypothesis H1: “Different variances” is rejected. Therefore, the results of ANOVA analysis can be used.

Leave a Message


By reply you accept the Privacy Policy *

Home Contact