Testing Correlation Between Variables in the Model

2.2.4.2. Testing the correlation between variables in the model

The variables included in the correlation test are: “Responsiveness”, “Tangibles”, “Employee Competence”, “Reliability”, “Empathy”, “Service Quality”, in which “Credit Service Quality” is the dependent variable and the remaining variables are independent variables. If these independent variables are correlated with the dependent variable, the regression analysis will be statistically significant.

H 0 : There is no correlation H 1 : There is correlation

Based on the correlation regression results table (Appendix 2, section 8), we see that the significance level of the dependent variable group with the independent variable groups is less than 0.05. Therefore, we come to the conclusion of rejecting H 0. Thus, the dependent variable "Credit service quality" and the independent variables "Responsiveness", "Tangible means", "Staff capacity", "Trust", "Empathy" have a strong correlation with each other.

2.2.4.3. Model suitability testing

To evaluate the model's suitability, we use the adjusted R 2 value and ANOVA test.

Table 12. Adjusted R-squared test results

Mode

R Square

Adjusted R

Square

Std. Error of the

Estimate

Durbin-Watson

, 820a

,673

,662

,58169369

1,721

Maybe you are interested!

(Source: survey results)

The results in the table above show that the model has an adjusted R2 value of 0.662. Thus, the model's suitability is 66.2%. In other words, 66.2% of the variation in the dependent variable "Credit service quality" is explained by the above 5 observed variables, the rest is due to the impact of other factors outside the model.

Next, the author conducts an F test on the suitability of the regression model, to see whether the dependent variable is linearly related to the entire set of independent variables.

The hypothesis H 0 is stated as:  1 =  2 =  3 =  4 =  5 =  6 =  7

The results of ANOVA analysis (Appendix 2, item 8) show that the Sig value is less than 0.05.

allows to reject the hypothesis H0 , that is, this regression model after being extrapolated to the population, its suitability has been verified, the combination of independent variables explains 66.2% of the changes in the dependent variable.

2.2.4.4. Testing for multicollinearity and autocorrelation

To test for multicollinearity, we rely on the VIF value. In theory, when Toleranco is very small and VIF exceeds 10, it is a sign of multicollinearity. Thus, based on the test results (Appendix 2, section 8), it shows that the variance inflation factor of the variables (VIF) is quite low (less than 10), it can be affirmed that multicollinearity does not occur in this model, or in other words, the independent variables clearly explain the dependent variable.

To test for autocorrelation, we rely on the results of the Durbin – Watson test.

After conducting the test, the Durbin – Watson value result is 1.721 (Table 12) in the range [1.6;2.6]. So the model does not have autocorrelation.

2.2.4.5. Regression analysis

The regression model applied in this study is a multivariate regression model (multiple regression model). In the regression analysis model, the dependent variable is the variable “ satisfaction with credit quality ”. The independent variables are factors extracted from the observed variables from the EFA factor analysis.

Table 13. Regression results

Model

Regression coefficient (Beta)	t	Sig.
Constant	-1,408E-016	0.000	0.000
Responsiveness	0.264	5,624	0.000
Tangible means	0.252	5,368	0.000
Staff capacity	0.385	8,222	0.000
Trust	0.565	12,056	0.000
Empathy	0.268	5,722	0.000

(Source: survey results)

Thus, from the regression results, we can represent the relationship between the dependent variable "Credit service quality" and the independent variables "Responsiveness",

“Tangibles”, “Employee Competence”, “Trust”, “Empathy” through the following equation:

Y=-1.408E-016+ 0.264*X 1 + 0.252*X 2 + 0.385*X 3 + 0.565*X 4 +0.268*X 5

In there:

Y: Credit service quality X 1: Responsiveness

X 2: Tangible means

X 3 : Staff capacity X 4 : Trust

X 5 : Empathy

The regression results show that all 5 factors of the model affect the quality of credit services of VietinBank Ha Tinh branch. In which, the component "Trust" has the largest regression coefficient, or has the most important meaning for the quality of credit services of the bank, followed by the factors "staff capacity", "empathy", "Responsiveness" and finally the factor "tangible means". Current reality shows that the operation of banks also faces many difficulties. Therefore, when choosing a bank to perform credit activities, the trust of the bank is the factor that customers in Ha Tinh are most concerned about, because they always want to have a certain guarantee when performing service activities. Besides, the competition between banks in the area is currently very fierce, therefore, the factor of staff capacity is also focused on by customers. As customer demand increases, it requires more ways and capabilities of service to attract their attention and make them decide to choose your bank's services.

2.2.4.6. Testing hypotheses

Regression equations help us explain and test the hypotheses that have been put forward:

Y=-1.408E-016+ 0.264*X 1 + 0.252*X 2 + 0.385*X 3 + 0.565*X 4 +0.268*X 5

From the equation, it can be seen that the coefficient of the factor “Responsiveness” is 0.264, which is a positive relationship. Thus, when the level of customer satisfaction with “Responsiveness” increases by one unit, the level of satisfaction with the quality of credit services increases.

increased by 0.264 times. With a significance level of Sig. < 0.05, hypothesis H 1 is accepted because there is no basis to reject this hypothesis.

The coefficient of the factor “Tangibles” is 0.252, which is a positive relationship. Thus, when the level of customer satisfaction with “Tangibles” increases by one unit, the quality of credit services increases by 0.252 times. With a significance level of Sig. < 0.05, the hypothesis H 1 is accepted because there is no basis to reject this hypothesis.

The coefficient of the factor "Employee capacity" is 0.385, which is a positive relationship.

dimension, thus, when the level of customer satisfaction with "Employee Competence" increases by one unit, the quality of credit services increases by 0.385 times. With the significance level Sig. < 0.05, hypothesis H 1 is accepted because there is no basis to reject this hypothesis.

The coefficient of the factor “Trust” is 0.565, which is a positive relationship, thus,

When customer satisfaction with “Trust” increases by one unit, credit service quality increases by 0.565 times. With a significance level of Sig. < 0.05, hypothesis H 1 is accepted because there is no basis to reject this hypothesis.

The coefficient of the factor “Empathy” is 0.268, which is a positive relationship. Thus, when the level of customer satisfaction with “Empathy” increases by one unit, the quality of credit services increases by 0.268 times. With the significance level Sig. < 0.05, the hypothesis H 1 is accepted because there is no basis to reject this hypothesis.

Table 14. Conclusion of the model hypotheses

Hypothesis

Content	Conclude
H 1	Responsiveness is positively correlated with credit service quality	Accept
H 2	Tangible means have a positive correlation with quality credit services	Accept
H 3	Employee capacity is positively correlated with credit service quality	Accept
H 4	Reliability is positively correlated with quality. credit services	Accept
H 5	Empathy is positively correlated with quality. credit service volume	Accept

(Source: survey results)

2.2.5. Customer evaluation of factors affecting the quality of bank credit services

Conduct One Sample T-test on the factors: Responsiveness, Tangibles, Staff Competence, Trustworthiness, Empathy with the pair of hypotheses:

H 0: µ = µ 0

H 1: µ ≠ µ 0

When evaluating each customer, we use a Likert scale with 5 levels of evaluation from 1 to 5 corresponding to the levels from strongly disagree to strongly agree. However, when evaluating a large number of customers, the factor groups will be evaluated according to the average value. At this time, the scale will be divided into the following ranges:

1 - 1.8: Strongly disagree 1.8 - 2.6: Disagree 2.6 - 3.4: Neutral

3.4 - 4.2: Agree 4.2 - 5: Strongly agree

2.2.5.1. Responsiveness factor group

To test how customers evaluate the variables of the factor group "responsiveness", we conduct a pair of hypotheses testing:

H 0 : overall mean of the variables in the responsiveness scale = 4 H 1 : overall mean of the variables in the responsiveness scale # 4

Table 15. Results of One Sample T-test for variables of the Responsiveness factor

Criteria

Sig value	Average value	Rating level
Sig value	Average value	M1	M2	M3	M4	M5
Simple procedures and documents	0.000	3,445	7.1	13.5	19.4	47.7	12.3
Fast approval and disbursement time promptly	0.000	3,665	1.9	9.7	24.5	47.7	16.1
NH solves the problem quickly, on time	0.000	3,703	1.3	12.3	25.8	36.1	24.5
NV is willing to answer questions for client	0.003	3,826	1.3	11.6	18.1	41.3	27.7
The bank clearly informs customers. When is the service performed?	0.002	3,735	1.9	13.5	18.1	41.9	24.5
Simple loan conditions, information Clear and attractive credit information	0.002	3,729	4.5	9.0	20.0	41.9	24.5
NV cares about each customer	0.002	3,768	0.6	11.6	17.4	51.0	19.4
NH promptly meets customer needs row	0.000	3,639	3.2	9.7	24.5	45.2	17.4

(Source: survey results)

Note: M1: Strongly disagree to M5: Strongly agree

Through the above data table, we can see that the Sig. value of all variables in the Responsiveness factor is less than 0.05, so there is not enough basis to reject the hypothesis H 0 . For the factor group " responsiveness", the factors are rated quite highly by customers and relatively evenly. All factors are in the "agree" range. In the 8 variables listed above, the variable " Employees are willing to answer questions for customers " has the largest average value of 3.826, the number of customers who rate at the level of agree or higher accounts for 69% of the total sample. In fact, credit officers here always respond well to customer requests, enthusiastically answering customers' questions about loan conditions, how to make a profile, loan procedures, interest rates, etc. The variable " Simple procedures and documents " is also in the "agree" range, but has the lowest average value, only 3.445. In reality, to carry out credit activities, customers need to prepare many types of required related documents such as mortgaged assets, loan application, income proof, etc. This sometimes causes difficulties in the process of carrying out credit activities.

2.2.5.2. Tangible means factor group

Hypothesis:

H 0 : the overall mean value of the variables in the “Tangibles” scale = 4

H 1 : overall mean value of variables in the “Tangibles” scale # 4

The results are as follows:

Table 16. Results of One Sample T-test for variables of the Tangibles factor

Criteria

Sig value	Average value	Rating level
Sig value	Average value	M1	M2	M3	M4	M5
NH has modern facilities and equipment full work	0.000	3,516	9.7	7.7	18.1	50.3	14.2
Convenient trading location	0.000	3,723	2.6	9.0	20.6	49.0	18.7
Spacious and convenient garage	0.000	3,452	1.3	14.2	33.5	40.0	11.0
Neat and polite staff uniform	0.000	3,645	5.2	7.7	21.3	49.0	16.8
There are diverse credit products, meeting meet customer needs	0.002	3,742	2.6	9.7	21.3	43.9	22.6
Competitive bank interest rates, with timely adjustment	0.000	3,658	3.9	10.3	23.9	40.0	21.9

(Source: survey results)

Note: M1: Strongly disagree to M5: Strongly agree

The Sig. values of the factors in this factor group are all less than 0.05, so there is not enough basis to reject H 0. The factors are evaluated relatively evenly and are all evaluated in the range of 3.4 to 4.2, meaning that customers evaluate them in the agreement range. The criteria in the Tangible means factor that are evaluated at the level of agreement or higher mostly account for over 60%. The factors "Convenient transaction location" and "Have diverse credit products, meeting customer needs" have the highest average value. Customers mostly agree with these two opinions. Currently, the bank has many credit product packages such as car loans, home loans, business loans, etc., which meet the diverse needs of customers. In addition, the Bank is located on a major road, convenient for searching and transactions. The factor "Spacious and convenient parking lot" is rated the lowest by customers. The rate of customers who rated from the level of agreement was only 51%. In fact, the transaction office is located next to a main road, so the corridor for customers' parking is limited, there is no separate parking lot of the bank, if customers go by car, they have to park on the roadside. Therefore, this factor is not highly appreciated by customers. The bank needs to come up with reasonable solutions to overcome this problem.

2.2.5.3. Employee capacity factor group

Hypothesis:

H 0 : the overall mean value of the variables in the scale “Employee Competence” = 4 H 1 : the overall mean value of the variables in the scale “Employee Competence” # 4

Table 17. Results of One Sample T-test for variables of the factor Employee Competency

Criteria

Sig value	Average value	Rating level
Sig value	Average value	M1	M2	M3	M4	M5
Highly qualified staff	0.001	3,768	0.6	8.4	21.3	52.9	16.8
Polite and respectful communication staff	0.000	3,394	1.9	12.3	40.6	34.8	10.3
NV works with professional style	0.000	3,510	1.9	9.0	37.4	39.4	12.3
NV makes customers comfortable, creates give confidence	0.000	3,613	1.9	14.2	20.6	47.1	16.1
NV has professional ethics, loyalty real, reliable	0.002	3,484	1.3	18.7	31.6	27.1	21.3
Employees know how to coordinate and help each other. customer service	0.000	3,594	1.3	7.7	34.2	43.9	12.9

(Source: survey results)

Note: M1: Strongly disagree to M5: Strongly agree

In this group of factors, all the components also have a significance level of less than 0.05. Therefore, there is not enough basis to reject H 0. The factor that employees always communicate politely and respectfully with customers is only rated at 3.394, which is in the neutral range. In fact, the work of credit officers is a lot, plus the pressure of work makes them feel tired. That affects their psychology when communicating with customers. In fact, during the research process, in some cases, a customer asked too many questions about credit conditions and procedures, the staff tried to answer but still did not satisfy the customer's questions, so the staff could not keep calm. Therefore, customers rated this factor at the level of disagreement. The remaining factors were rated by customers at the level of agreement. The factor "Highly qualified staff" has a fairly high average value (3.768), the percentage of customers who rate it from the level of agree or higher accounts for 69.7%. Employees at the bank are a team of carefully recruited personnel with good professional expertise, solid knowledge of credit as well as understanding of social information channels to best meet customers' requests and questions.

2.2.5.4. Trust factor group

H 0 : the overall mean value of the variables in the scale “ Trust ” = 4 H 1 : the overall mean value of the variables in the scale “ Trust ” # 4 Table 18. One Sample T-test for variables of the factor Trust

rely

Criteria

Value Sig.	Value TB	Rating level
Value Sig.	Value TB	M1	M2	M3	M4	M5
NH ensures information confidentiality client	0.001	3,458	3.9	11.6	29.0	45.8	9.7
NH performs accurate service, less error prone	0.000	3,477	0.6	10.3	41.3	36.1	11.6
NH provides timely service points and as promised	0.000	3,503	2.6	11.6	28.4	47.7	9.7

(Source: survey results)

Note: M1: Strongly disagree to M5: Strongly agree

The factors " Bank ensures customer information confidentiality", "Bank performs services accurately, with few errors", "Bank provides services on time and as committed" all have Sig. values less than 0.05, so there is not enough basis to reject H 0 . The average values of these factors are relatively equal. This shows that customers are quite satisfied with the reliability of the bank. However, the satisfaction level is only at the lower end of the "agree" rating range. Looking at the statistical table, it can be seen that customers rate the criteria at the neutral level or lower quite high. The factor "Bank performs services accurately, with few errors" has a neutral rating of 41.3% of customers. Therefore, banks need to promote solutions to further enhance customers' trust and confidence in the bank.

2.2.5.5. Empathy factor group

Hypothesis:

H 0 : the overall mean value of the variables in the scale “ Trust ” = 4 H 1 : the overall mean value of the variables in the scale “ Trust ” # 4 Table 19. One Sample T-test for the variables of the factor

Empathy

Criteria

Value Sig.	Value TB	Rating level
Value Sig.	Value TB	M1	M2	M3	M4	M5
Regularly have chapters customer care	0.001	3,561	2.6	12.3	28.4	40.0	16.8
NH understands the needs and benefits customer benefits	0.000	3,535	14.8	-	31.0	40.0	14.2
Bank working hours convenient for customers	0.000	3,619	2.6	11.6	25.2	42.6	18.1

(Source: survey results)

Note: M1: Strongly disagree to M5: Strongly agree

From the results of the table above, it can be seen that there is not enough evidence to reject the hypothesis H 0 (because the significance level of the factors in the Empathy scale is less than 0.05). The factors are all rated in the value range from 3.4 to 4.2. That means that most customers agree with these factors. “In the three factors of this scale, customers

The factor “Bank working hours are convenient for customers” has the highest average value (3.619), the percentage of customers who agree and strongly agree is more than 60%. This shows that customers are quite satisfied with the bank's working hours. The remaining two factors are evaluated quite equally. The average value is both greater than 3.5, indicating that customers quite agree with the factors “Regularly have customer care programs” and “Employees understand the needs and benefits of customers”. However, for the factor “Employees understand the needs and benefits of customers”, up to 14.8% of customers rated as strongly disagreeing. The bank needs to find out the specific reasons from these customers to have appropriate solutions.

2.2.6. Testing the difference in credit service quality assessment between customer groups

2.2.6.1. Independent Sample T Test

Conduct an audit of the average assessment level of credit service quality for factors: gender and age

Hypothesis:

H 0 : The mean values of the two groups are equal H 1 : The mean values of the two groups are different

Differences in average ratings of credit service quality between gender groups

Based on the Independent Sample T Test results table in section 10 of Appendix 2, we see that the Sig value in Levene's Test is 0.078, which is greater than 0.05, so it can be concluded that the two gender groups have equal variance in the overall variance. Looking at the results of the test column on the equality of the average of the two populations in the homogenous variance row, we see that the Sig value = 0.08 > 0.05, so there is not enough basis to reject the hypothesis H 0 . It can be concluded that the average value of the two groups is equal or there is no difference in the average assessment level of credit service quality between the two groups of men and women in the overall population. Therefore, the bank's credit activities can be carried out synchronously for these two groups of subjects.

Differences in the influence of age on the assessment of credit service quality

The test results (Appendix 2, item 10) show that, in the test for homogeneity of variance, the Sig. value = 0.428 > 0.05. Therefore, the age groups are equal in terms of overall variance. The test results for the difference in the influence of age on the assessment of credit service quality show that the Sig value of this test is 0.530, greater than 0.05, we accept the hypothesis H 0 . That is, there is no difference in the assessment of credit quality between the age groups of customers.

2.2.6.2. One Way Anova Test

We conduct a One Way Anova test for the factors: occupation and income with the pair of hypotheses:

H 0 : There is no difference in the assessment of credit service quality between groups in the same factor

H 1 : There is a difference in the assessment of credit service quality between groups in the same factor.

Differences in the influence of occupation on the assessment of credit service quality

The variance test results table (Appendix 2, item 11) shows that with the significance level sig.= 0.837 > 0.05, it can be said that the variance of customer credit quality assessment between occupational groups is homogeneous. Thus, the ANOVA analysis results in the table can be used well.

The results of ANOVA analysis on the differences between occupational groups show that the Sig. value = 0.354 >0.05. Therefore, there is not enough basis to reject the hypothesis H 0 or to conclude that there is no difference in customer assessment of credit service quality between occupational groups.

Differences in the impact of income on credit service quality assessment

The Sig. value in the variance test table of credit quality assessment among the 3 income groups is 0.113, which is greater than the significance level α = 0.05. Therefore, it can be concluded that the income groups have equal variance. The ANOVA test results can be used.

Through the One Way Anova test table (Appendix 2, item 11), it shows that the Sig value.

= 0.179 > 0.05. Therefore, we accept the hypothesis H 0 and conclude that there is no difference between income groups in assessing the quality of credit services at VietinBank Ha Tinh branch.

CHAPTER 3

ORIENTATION AND SOLUTIONS TO IMPROVE CREDIT SERVICE QUALITY AT VIETINBANK

HA TINH BRANCH

1.1. Credit activity orientation of Vietinbank Ha Tinh

1.1.1. Specific objectives of Vietinbank Ha Tinh

Specific goals of Vietinbank Ha Tinh in 2015

- Mobilized capital by the end of the year reached 2,226 billion VND (including capital mobilized in the area, capital outside the area, capital of financial institutions, social insurance). Of which, mobilized savings capital is 1,550 billion VND.

- Year-end outstanding debt reached VND 2,793 billion (including co-financed loans)

- Actual interest collected is 100% as committed in the credit contract.

- Improve credit quality, strive to keep bad debt ratio below 1.5%.

- Profit strives to reach 82,830 million VND

1.1.2. Credit activity orientation

Actively increase credit growth on the basis of:

- Expand customer base, targeting commercial customers in all economic sectors. Increase outstanding loans to individual and household customers; focus on lending capital to businesses in industries with financial potential and effective production and business; resolutely reduce outstanding loans, even stop credit relationships with ineffective and loss-making customers;

- Expanding credit portfolio: Developing retail and consumer loans, financial proof loans, import-export business loans, micro-enterprise loans, etc.

- Strengthen marketing, target potential credit market segments, expand transaction network by opening new transaction offices.

- Apply flexible loan interest rates and service fees within VietinBank's allowable limits for each specific customer.