Home / Domestic Tour

% > 50 %. This Shows That 64.233 % Of The Variation In The Data Is Explained By 6 Factors.


According to table 4.14, the Eigenvalue criterion has a value of 1.476 and 26 variables are divided into 6 groups of component factors. The total variance extracted is 65.045%>50% which is satisfactory, showing that the 6 groups of factors above can explain the influence level of 26 variables is 65.045%. Analyzing the component rotation matrix, the factor loading coefficient represents the relationship between factors and variables. This large coefficient indicates that the factors and variables are closely related to each other. Therefore, the observed variables have factor loading weights

<0.5 and with distribution Δ<0.3 will be eliminated.

Table 4.15. First component rotation matrix




Ingredient

1

2

3

4

5

6

NC2

.814






NC5

.806






NC1

.731






NC3

.710






NC4

.683






CL5


.790





CL3


.756





CL4


.754





CL1


.702





CL2


.641





TK1



.752




TK4



.749




TK3



.736




TK2



.714




TH4

.545


.687




GC1




.795



GC4




.775



Maybe you are interested!

% &gt; 50 %. This Shows That 64.233 % Of The Variation In The Data Is Explained By 6 Factors.


GC3




.742



GC2




.718



MK1





.774


MK4





.740


MK2





.734


MK3





.713


TH1






.832

TH2






.789

TH3






.756

Extraction method: Principal Component Analysis.

Rotation method: Varimax with Kaiser Normalization.

Source: Data analysis – appendix 4


Eliminate variable TH4 because Δ<0.3.



after:

Because variable TH4 does not satisfy the condition, the second retest has the following result:


Table 4.16. KMO and Bartlett coefficients for the second time


KMO and Bartlett's Test

KMO coefficient (Kaiser-Meyer-Olkin).

.873

Bartlett's test model.

Chi-square value

2499.958

Degrees of freedom

300

Sig (P-value)

.000

Source: Data analysis – appendix 4


KMO = 0.884 so factor analysis is appropriate.


>> Sig. (Bartlett's Test) = 0.000 (sig. < 0.05) shows that the observed variables are correlated with each other in the population.


Table 4.17. Second extracted total variance


Ingredient

Initial Eigenvalues

Index after deduction

Index after rotation

Total

%

variance

Accumulate

%

Total

%

variance

Accumulate

%

Total

%

variance

Accumulate

%

1

7.302

29,210

29,210

7.302

29,210

29,210

3.129

12,518

12,518

2

2,153

8,612

37,822

2,153

8,612

37,822

3,089

12,357

24,874

3

1,851

7.404

45,226

1,851

7.404

45,226

2,644

10,577

35,451

4

1,750

7,000

52,226

1,750

7,000

52,226

2,506

10,024

45,476

5

1,574

6,294

58,520

1,574

6,294

58,520

2,496

9,984

55,460

6

1,428

5,713

64,233

1,428

5,713

64,233

2,193

8,773

64,233

Extraction method: Principal Component Analysis.

Source: Data analysis – appendix 4


Eigenvalues ​​= 1.428 > 1 represents the portion of variation explained by each factor, then the extracted factor has the best summary of information.

Total extracted variance: Rotation Sums of Squared Loadings (Cumulative %) =

64.233 % > 50 %. This shows that 64.233 % of the variation in the data is explained by 6 factors.

Table 4.18 Final component rotation matrix



Ingredient

1

2

3

4

5

6

NC2

.824






NC5

.805






NC1

.741






NC3

.703






NC4

.671






CL5


.792







Ingredient

1

2

3

4

5

6

CL3


.759





CL4


.754





CL1


.700





CL2


.641





GC1



.797




GC4



.776




GC3



.742




GC2



.720




TK1




.766



TK4




.741



TK3




.723



TK2




.715



MK1





.776


MK4





.742


MK2





.735


MK3





.711


TH1






.832

TH2






.789

TH3






.756

Extraction method: Principal Component Analysis.

Rotation method: Varimax with Kaiser Normalization.

Source: Data analysis – appendix 4


After performing Principal components extraction and Varimax rotation methods, the remaining 25 variables are distributed into 6 factor groups as follows:

Needs: NC1, NC2, NC3, NC4, NC5


Reference: TK1, TK2, TK3, TK4


Brand: TH1, TH2, TH3


Quality: CL1, CL2, CL3, CL4, CL5 Price: GC1, GC2, GC3, GC4

Marketing: MK1, MK2, MK3, MK4


Table 4.19. EFA factor analysis results table for domestic tour selection decision

KMO and Bartlett's Test

KMO coefficient (Kaiser-Meyer-Olkin).

.818

Bartlett's test model.

Chi-square value

384,375

Degrees of freedom

6

Sig (P-value)

.000



Encryption

Component

1

QD1

.846

QD4

.819

QD3

.818

QD2

.812

Source: Data analysis – appendix 4


The four observed variables of “Choice Decision” were performed by Principal components extraction method and Varimax rotation. KMO coefficient = 0.818 > 0.5, factor analysis is suitable for research data. Bartlett test result is 384.375 with sig < 0.05 significance level (rejecting the Ho hypothesis that the observed variables are not correlated with each other in the population) so the hypothesis of the factor model is not suitable and will be rejected, this proves that the data for factor analysis is completely suitable.


4.3.2. Multiple regression analysis

After conducting an exploratory factor analysis (EFA) of the scale of factors determining tourists' choice of domestic tours, the author continued to analyze the factors affecting tourists' choice at Lua Viet Travel Company. The dependent variable is the decision and the independent variables are the factors of demand, reference, brand, quality, price and marketing.

Call the independent variables including 6 variables: "Demand", "reference", "brand", "quality", "price", "marketing".

Call the dependent variable: “Choice decision” To analyze the regression, the author calls:

+ Factor 1: NC is demand (the average of variables NC1, NC2, NC3, NC4, NC5)

+ Factor 2: TK is the reference (the average of variables TK1, TK2, TK3,

TK4)


+ Factor 3: TH is the brand (the average of variables TH1, TH2, TH3)

+ Factor 4: CL is quality (the average of variables CL1, CL2, CL3,

CL4, CL5)

+ Factor 5: GC is the price group (the average of variables GC1, GC2, GC3, GC4)

+ Factor 6: MK is convenience (the average of variables MK1, MK2, MK3, MK4)

Let QD be the decision to choose a domestic tourism program of tourists (the average of variables QD1, QD2, QD3, QD4). The function of factors affecting tourists' decision is:

QD = 𝜷 1NC + 𝜷 2TK + 𝜷 3TH + 𝜷 4CL + 𝜷 5GC + 𝜷 6MK

The independent variables will be measured through the results of multiple regression analysis as shown in the table below.


Model

R

R squared

R squared difference

adjust

Std. Error of the Estimate

Durbin-Watson

1

.856

.732

.725

.38942

2,094

Table 4.20. Multivariate regression results Summary model


Source: Data analysis – appendix 4


The adjusted R-squared is 0.725 = 72.5 %. Thus, the independent variables included in the regression affect 72.5 % of the change in the dependent variable.


Table 4.21. Regression analysis results


Model

Zero factor

standard

Coefficient

standard

t

Sig.

Collinearity statistics

B

Std.

Error

Beta

Tolerance

VIF


1

(Constant)

-.131

.138


-.949

.344



NC

.243

.032

.291

7,695

.000

.769

1,300

MK

.082

.029

.104

2,777

.006

.787

1,271

CL

.183

.031

.234

5,843

.000

.691

1,448

TK

.083

.031

.107

2,685

.008

.700

1,428

GC

.200

.032

.243

6,249

.000

.728

1,373

TH

.203

.028

.272

7.152

.000

.764

1,308

Regression equation:

QD = 0.291*NC + 0.104*MK + 0.234*CL + 0.107*TK + 0.243*GC + 0.272*TH

From the table above, we can see that the VIF value is used to check for multicollinearity of independent variables. If the VIF value is less than 10, there is no multicollinearity.


collinearity. According to the linear regression results, the VIF values ​​of the independent variables are much smaller than the allowable value of 10, so there is no multicollinearity between the independent variables.

The correlation coefficient R has been shown to be a non-decreasing function of the number of independent variables included in the model (6 variables).

R2 = 0.732 has shown the reality of the model

Adjusted R2 from R2 of 0.725 was used to more accurately reflect the goodness of fit of the multivariate regression model, and adjusted R2 was also independent of the exaggerated bias of R2 .

Thus, the adjusted R2 is 0.725, showing that the model's compatibility with the observation station is very large and the dependent variable of tourists' decision-making behavior is completely explained by the 6 independent variables in the model: demand, reference, brand, quality, price and marketing.

Table 4.22. Results of meta-regression analysis


STT

Symbol

Group name

Beta coefficient

Coefficient

Standardized Beta

Sig coefficient

1

NC

Demand

.243

.291

.000

2

MK

Travel agency marketing

.082

.104

.006

3

CL

Quality provided for

client

.183

.234

.000

4

TK

Consult

.083

.107

.008

5

GC

Price

.200

.243

.000

6

TH

Company Brand

.203

.272

.000

Source: Author's synthesis

The summary table above shows:

+ The NC group (demand) has the highest standardized Beta coefficient (0.291) and has the strongest impact on tourists' decision to choose domestic tourism programs, because

Comment


Agree Privacy Policy *