Kmo Test and Bartlett's Test for Independent Variables

By occupation : it can be seen that the survey subjects of the research topic have the main occupation of "Business" with 44 responses (accounting for 36.7% of the total 120 survey subjects. The second group is the group of subjects "Office staff, civil servants" (accounting for 34.2%) and "General workers" (accounting for 18.3%). The remaining groups are a few groups "Students" (with 7 responses) and "Other" group (only 6 responses).

By income : the results of the table above, it is easy to see that the majority of customers have an income from 5 to 10 million/month. Of which, the income level of "5 to 10 million/month" has the highest rate with 43 responses (accounting for 35.8% of the total 120 surveyed subjects), followed by the group with an income level of "10 to 15 million/month" with 40 subjects (accounting for 33.3% of the total 120 surveyed subjects), ranked 3rd is the group "Over 15 million/month" accounting for 16.7% and the least is the group "Under 5 million/month" with only 17 responses (accounting for 14.2%).

2.2.1.2. Customer behavioral characteristics

Table 2.5: Customer behavioral characteristics

Criteria

Quantity (120)	Structure (100%)	Accumulation rate (100%)
According to the number of times using travel services at Global Travel company
First time	69	57.5	57.5
Second time	41	34.2	91.7
More than 2 times	10	8.3	100
According to the information source that helps customers know the company
Promotional and introductory activities of the company	99	82.5
Via the Internet	93	77.5
Friends and relatives recommend	66	55
Other	45	37.5
According to the reason why customers choose the company's services
Reasonable price	38	31.7	31.7
Prestigious brand	33	27.5	59.2
Good service quality	25	20.8	80
Rich services, diversity	14	11.7	91.7
Enthusiastic and attentive staff unique	5	4.2	95.8
Other	5	4.2	100

Maybe you are interested!

(Source: Author's investigation results in 2020)

According to the number of times using tourism services at Global Travel Company : tourism is one of the industries that is most difficult to retain customers when more and more tourism companies are emerging, so it is easy to see that most customers only use the service from 1 to 2 times mainly. Based on the results of the survey, it is easy to see that most customers only use tourism services for the "First time" with 69 responses (accounting for 57.5% of the total 120 surveyed subjects). Coming in second is the group using tourism services for the "Second time" with 41 responses and accounting for 34.2%. Meanwhile, only 8.3% of the total 120 surveyed subjects have used tourism services at Global Travel Company "More than twice".

According to the source of information that helps customers know the company : out of a total of 120 surveyed customers, it can be seen that the majority of customers know the company Global Travel "Through the Internet", with 93 subjects accounting for 77.5%. This is a fairly high number, which is understandable, because today when customers are exposed to Internet media every day and information technology is increasingly modern and developed, searching and choosing a travel company is also easier and more complete. In addition, because the company always promotes marketing and communication activities, the fact that they know the company Global Travel through "Promotional activities, introducing the company" accounts for the highest rate with 99 subjects choosing and accounting for 82.5%. Next is the group that knows the company "Through friends and relatives' introduction" with 55% of the total 120 surveyed subjects. The rest is the "Other" hedgehog with only 37.5%.

According to the reason why customers choose the company's services : when there are more and more travel companies emerging like today, the choice of travel services as well as the price at a company will be greatly influenced by the company. Therefore, it is easy to see that out of 120 surveyed subjects, 38 subjects have the reason "Reasonable price", accounting for 31.7%. Coming in second is the group with the reason "Reputable brand" with 33 responses, accounting for 27.5%. In addition, a few customers choose travel services at Global Travel company with the reason "Good service quality" accounting for 20.8%, "Services

"Rich and diverse services" accounted for 11.7% and the remaining 2 groups were "Enthusiastic staff".

“Conscientious” and “Other” accounted for only 4.2% of the total 120 survey subjects.

2.2.2. Testing the reliability of the scale Cronbach's Alpha

Before proceeding to the data analysis steps, the study conducted a reliability test of the scale through the Cronbach's Alpha coefficient. Cronbach's Alpha must be performed first to eliminate irrelevant variables (Garbage Items) before conducting the EFA exploratory factor analysis.

The research topic uses a scale of 5 independent variables: "Trust", "Responsiveness", "Empathy", "Tangibles", "Service capacity".

Variables with a Corrected Item Total Correlation greater than 0.3 and a Cronbach's Alpha greater than 0.6 will be accepted and included in the next steps of analysis. Specifically:

o Cronbach's Alpha coefficient greater than 0.8: high correlation coefficient.

o Cronbach's Alpha coefficient from 0.7 to 0.8: acceptable.

o Cronbach's Alpha coefficient from 0.6 to 0.7: acceptable if the scale is new.

During the reliability testing process, the observed variables all have correlation coefficients.

The total variable coefficient is greater than 0.3 so no variable is removed from the model.

The results of Cronbach's Alpha test are summarized in the table below:

Table 2.6: Reliability test of independent variables scale

Variable

Total variable correlation coefficient	Cronbach's Alpha coefficient if variable type
1. Reliability: Cronbach's Alpha = 0.760
TINCAY1	0.619	0.669
TINCAY2	0.556	0.705
TINCAY3	0.520	0.724
TINCAY4	0.539	0.714
2. Responsiveness: Cronbach's Alpha = 0.784
DAPUNG1	0.574	0.741
DAPUNG2	0.581	0.739
DAPUNG3	0.611	0.721
DAPUNG4	0.604	0.725
3. Service capacity: Cronbach's Alpha = 0.802
PHUCVU1	0.658	0.730
PHUCVU2	0.626	0.749
PHUCVU3	0.648	0.736
PHUCVU4	0.536	0.789
4. Empathy: Cronbach's Alpha = 0.798
DONGCAM1	0.645	0.730
DONGCAM2	0.665	0.718
DONGCAM3	0.596	0.753
DONGCAM4	0.535	0.782
5. Tangibles: Cronbach's Alpha = 0.761
HUUHINH1	0.566	0.702
HUUHINH2	0.571	0.700
HUUHINH3	0.539	0.719
HUUHINH4	0.568	0.701

(Source: Author's investigation results in 2020)

Through the summary table of the results of the reliability test of the scale above, it can be concluded that the scale used in the study is appropriate and reliable, ensuring the step of exploratory factor analysis EFA.

Table 2.7: Reliability test of dependent variable scale

Variable

Variable correlation coefficient total	Cronbach's Alpha coefficient if variable type
Satisfaction: Cronbach's Alpha = 0.730
HAILONG1	0.550	0.665
HAILONG2	0.581	0.681
HAILONG3	0.556	0.640

(Source: Author's investigation results in 2020) The results of the reliability assessment of the factor "Satisfaction" for Cronbach's Alpha coefficient = 0.730. The total correlation coefficient of the 3 observed variables is greater than 0.3 and the Cronbach's Alpha coefficient if the variable is eliminated is less than 0.730, so the dependent variable "Satisfaction" is retained and ensures reliability to perform the analysis steps.

2.2.3. Exploratory Factor Analysis (EFA)

2.2.3.1. KMO test and Bartlett's Test of independent variables

To apply factor analysis, it is necessary to test the suitability of the data for the factor analysis method. This test is performed through two quantities: KMO (Kaiser-Meyer-Olikin Meansure of Sampling Adequacy) and Barlett's Test of Sphericity.

The KMO value is an indicator used to examine the appropriateness of EFA. Test content: the KMO coefficient must satisfy the condition 0.5 ≤ KMO ≤ 1, proving that the EFA exploratory factor analysis step is appropriate in this study.

The results are as follows:

- KMO value of 0.704 is greater than 0.5 and less than 1, showing that EFA analysis is appropriate.

- The significance level Sig. of Bartlett's Test is less than 0.05, so the observed variables included in the research model are correlated with each other and suitable for exploratory factor analysis EFA.

Table 2.8: KMO test and Bartlett's Test of independent variables

KMO and Bartlett's Test

KMO value (Kaiser-Meyer-Olkin Measure of Sampling Adequacy		0.704
Statistical quantity Bartlett's Test of Sphericity	When squared(Chi- Square)	777,534
	Standard deviation(Df)	190
	Significance level(Sig.)	0.000

(Source: Author's investigation results in 2020)

2.2.3.2. Exploratory factor analysis EFA independent variables

In this study, when analyzing the EFA, the topic uses the Principal Components analysis method with the number of factors (Number of Factor) determined in advance as 5 according to the proposed research model. The purpose of using this method is to reduce the data, limit the violation of multicollinearity between factors in the analysis of the next regression model.

The selected factor rotation method is Varimax procedure: rotate the original factors to minimize the number of variables with large coefficients on the same factor to enhance the ability to explain the factor. Variables with factor loading coefficient < 0.5 will be eliminated from the research model, only variables with factor loading coefficient > 0.5 will be included in the next analysis.

In this study, the factor loading coefficient (Factor Loading) must satisfy the condition of greater than or equal to 0.5. According to Hair & ctg (1998), Factor Loading is the indicator to ensure the practical significance of EFA, Factor Loading > 0.3 is considered the minimum level. This study chooses the Factor Loading value > 0.5 with a sample size of 120.

Table 2.9: Extracting independent variable factors

Observation variable

Group of factors
	1	2	3	4	5
PHUCVU1	0.826
PHUCVU3	0.802
PHUCVU2	0.799
PHUCVU4	0.703
DONGCAM2		0.838
DONGCAM1		0.798
DONGCAM3		0.759
DONGCAM4		0.715
DAPUNG4			0.802
DAPUNG3			0.798
DAPUNG2			0.756
DAPUNG1			0.726
TINCAY1				0.806
TINCAY4				0.759
TINCAY2				0.749
TINCAY3				0.699
HUUHINH1					0.786
HUUHINH4					0.756
HUUHINH2					0.754
HUUHINH3					0.735
Eigenvalue coefficient	3,282	2,956	2,366	2,026	1,849
Variance progressive (%)	16,412	31,192	43,022	53,151	62,395

(Source: Author's investigation results in 2020)

Comment