Home / Tourism

Summary of Dependent Variable Efa Analysis


Table 4. 3. Summary of EFA analysis of dependent variables


STT

Variable

Factor

1

1

BE1

0.712

2

BE2

0.799

3

BE3

0.857

4

BE4

0.876

Cronbach's Alpha

0.856

KMO

0.742

Bartlett (Sig.)

0.000

Total variance extracted (%)

66.165 %

Maybe you are interested!

(Source: Author's investigation results, October 2018)


Re-evaluate the reliability of the scale through cronbach's alpha analysis after removing the observed variable.

The brand awareness component remained the same as the original and there was no change in the Cronbach's Alpha coefficients.

The Brand Desire component was kept the same as the original observation variables proposed. Therefore, there was no change in Cronbach's Alpha coefficient.

The Brand Loyalty component was kept the same as the original observation variables. Therefore, there was no change in Cronbach's Alpha coefficient.

The Destination Image component was kept the same as the original observation variables. Therefore, there was no change in Cronbach's Alpha coefficient.


Adjustment research model

4.5.1. Adjustment model


After conducting testing and evaluating the scales (through Cronbach's Alpha analysis and exploratory factor analysis (EFA), the measurement scales in the theoretical model were tested and achieved reliability and validity.

4.5.2. Hypotheses after adjustment

Hypothesis H1: Brand awareness has a positive impact on tourism brand equity.

Hypothesis H2: Brand desire has a positive impact on tourism brand equity.

Hypothesis H3: Destination image has a positive impact on tourism brand value.

Hypothesis H4: Brand loyalty has a positive impact on tourism brand equity.


4.5.3. Observed variables after adjustment

After performing the scale reliability assessment, the observed variables of the model will be adjusted to comply with the standards in the scale reliability assessment method to ensure the authenticity and reliability of the variables. The observed variables after adjustment are shown in the following table:


Table 4. 4. Adjusted observation variables


Factor

Variable

Observable variable content


1. Brand awareness

AW1

I know X is a city with developed tourism.

AW2

I can recognize the characteristics of city X.

AW3

I can distinguish city X from other cities.

AW4

I can easily access the tourist attractions of city X.

AW5

I can remember and recognize images of city X.

AW6

I can picture city X when I think of it.


2. Desire

trademark


BI1

I believe that traveling in city X is more worth the money than other cities.

another street

BI2

The possibility of me traveling to city X is very high.

BI3

I often travel to city X.

BI4

I believe, I want to travel in city X.


3. Loyalty and love

effect

LY1

I am a loyal tourist of city X.

LY2

City X is my first choice when traveling.

LY3

I will travel to X and not other cities.


LY4

If other cities have special programs (festivals, discounts)

price…) I will still travel to city X.


4. Destination image

DI1

The infrastructure in city X is very good.

DI2

Tourist attractions in city X meet my needs

DI3

Accommodation and services in city X are very good.

DI4

City X has a high level of security.


DI5

Honesty in selling products to tourists

The schedule in city X is very good.

DI6

In general, city X has high tourism quality.

(Source: Author's investigation results, October 2018)


Regression analysis


After testing the reliability and evaluating the value of the scales in the proposed model, the tourism brand value continues to be tested for its significance in the theoretical model through regression analysis to know the specific weight of each component affecting the overall brand value.

4.6.1. Variable encoding


Before conducting regression, the author codes the variables, the value of the coded variable is calculated by the average of the observed variables, specifically as follows:

Table 4. 5. Variable encoding


STT

Factor

Encryption

1

Brand awareness

AW

2

Brand Desire

BI

3

Brand loyalty

GLASS

4

Destination image

DI

5

Tourism brand value

BEIGE

(Source: Author's investigation results, October 2018)


4.6.2. Correlation analysis


After coding the measurement variables, the author entered the coded variables (AW, BI, LY, DI, BE) into SPSS software to analyze the correlation between these variables. Through the results of the correlation analysis, the author found that the factors AW, BI, LY, DI all have a strong correlation with the factor BE (sig = 0.000), so these variables can be entered into the regression analysis.

4.6.3. Regression analysis


After coding the measured variables and analyzing the correlation between the variables, the author conducted regression analysis with the Enter method. According to this method, 04


Independent variables (AW, BI, LY, DI) and one dependent variable (BE) will be entered into the model at the same time and give the following results:

Table 4.6. Summary of regression model


Tissue

image

R

R

square

R squared

correction

Standard error

estimate

Durbin-

Watson

1

0.843a​

0.710

0.706

0.59841

1,933

(Source: Author's investigation results, September-October 2018)


Table 4.7. Results of Anova analysis in regression


Model

Total average

direction

Df

Square

medium

F

Sig.


1

Regression

256,003

4

64,001

178,727

0.000 b

Remainder

104,563

292

0.358



Total

360,566

296




(Source: Author's investigation results, September-October 2018)


The results of multiple linear regression showed that the model had a coefficient of determination R2 of 0.843 and an adjusted R2 of 0.706.

F test (Anova table) shows the significance level p=0.000<0.05. Thus, this regression model is suitable, or in other words, the brand equity component variables explain about 70.6% of the variance of the overall brand equity variable.


Table 4.8. Regression weights



Model

Unstandardized regression coefficients

Standard regression coefficient

chemical


t


Sig.

Multicollinearity statistics


B

Standard deviation


Beta


Tolerance


VIF

1

Constant

number

-0.006

0.259


-0.023

0.982



AW

0.021

0.031

0.022

0.683

0.495

0.026

0.040

BI

0.233

0.041

0.202

5,695

0.000

0.483

0.316

DI

0.294

0.025

0.428

11,898

0.000

0.677

0.571

GLASS

0.356

0.032

0.437

10,982

0.000

0.737

0.541

(Source: Author's investigation results, September-October 2018)


All variables have variance inflation factor VIF <10, which proves that there is no multicollinearity in the model.

In the weight table above, we see that the components BI, LY and DI have a positive impact on the dependent variable BE because the regression weights of these 3 components are statistically significant p<0.05. If we compare the impact of these 3 variables on the dependent variable BE, we see that the Beta coefficient of BI is 0.104, LY is 0.405 and DI is 0.414, meaning that of the 3 components, DI and LY have the strongest impact, followed by BI. However, the AW component in the regression weight table is not statistically significant (p=0.495 >0.05).

The AW component in the regression weight table is not statistically significant in accurately reflecting the brand value of the tourism industry in particular. The cities included in the survey are all large and famous tourist cities widely known in the domestic tourism market. The ease of recognizing tourist destinations has long been grasped and learned by domestic tourists. Therefore, measuring brand value cannot be assessed through the level of recognition for a tourism brand because the destinations all have similar levels of recognition. Therefore, the brand recognition factor for the tourism industry in Vietnam is not meaningful in terms of


statistics as well as reality. However, it can be seen that in this study, the author chose 5 typical survey destinations and most of them were heard or experienced by domestic tourists, so the AW component is not statistically significant.

Assumptions about the normal distribution of residuals



Figure 4.1. Normal distribution chart of residuals


(Source: Author's investigation results, September-October 2018)


Based on the graph, it can be said that the normal distribution of the residuals is approximately normal (Mean=- 2.1E-15) and the standard deviation Std.Dev = 0.993, which is close to 1. Therefore, it can be concluded that the assumption of normal distribution of the residuals is not violated.


We can use the PP plot chart to test this hypothesis:


Figure 4.2. PP plot chart


(Source: Author's investigation results, September-October 2018)


Based on the PP plot, it can be seen that the observed points are not too far from the expected line, so we can conclude that the normal distribution assumption is not violated. In addition, through the scatterplot, it can be seen that there is uniform dispersion.

Comment


Agree Privacy Policy *