Results of Developing a Qualitative Scale on “Investment Intention” Symbol Name of Measured Variable Source

Table 3.22: Potential tourism market - Reliability Statistics

Table 3.23: Potential tourism market - Item-Total Statistics

Source: Author's analysis results using SPSS 22.0 software

Based on the above results, we see that the Cronbach's Alpha coefficient = 0.832, all measurement variables have a total correlation coefficient greater than 0.3, which is acceptable. The Cronbach's Alpha coefficient if the KT6 variable is removed is 0.894, which is greater than 0.832. However, because the KT6 variable is relatively important, we can keep it to test EFA before removing the variable.

c. Reliability testing of the “Tourism infrastructure” scale

The results of running Cronbach's Alpha for the "Tourism Infrastructure" scale are as follows:

Table 3.24: Tourism Infrastructure - Reliability Statistics

Table 3.25: Tourism infrastructure - Item-Total Statistics

Source: Author's analysis results using SPSS 22.0 software

We see that the Cronbach's Alpha coefficient = 0.830; the total correlation coefficient of the measured variables is greater than 0.6, which is very good (according to the requirement, it only needs to be greater than 0.3). The Cronbach's Alpha coefficient if the variable is eliminated is less than 0.830. So this scale's measurement variables for the tourism infrastructure factor are very good, we do not eliminate any variables.

d. Testing the reliability of the “Investment Environment” scale

The results of running Cronbach's Alpha for the " Investment Environment " scale are as follows:

Table 3.26: Investment Environment - Reliability Statistics

Table 3.27: Investment Environment - Item-Total Statistics

Source: Author's analysis results using SPSS 22.0 software

With the above results, we see that the Cronbach's Alpha coefficient = 0.810 is very good. However, the Cronbach's Alpha coefficient if the variables MT1; MT8; MT9; MT10 are removed is greater than 0.810, and these variables have a total correlation coefficient of less than 0.3, which is not satisfactory. However, these variables are very important variables confirmed in the provincial competitiveness index PCI. Therefore, before removing these variables, the author decided to keep them to check the reliability of the scale in EFA analysis. This is appropriate because Nguyen Dinh Tho (2011) once said that if any measurement variable is important and nearly equivalent, we should carefully consider keeping it or pay attention in EFA analysis to check it again before removing that variable. According to the author and experts, the variables MT1; MT8; MT9; MT10 is very important so the author decided to retain and continue to test this variable in the exploratory factor analysis section.

e. Reliability testing of the “Cost advantage” scale

The results of running Cronbach's Alpha for the "Cost Advantage" scale are as follows:

Table 3.28: Cost advantage - Reliability Statistics

Table 3.29: Cost Advantage - Item-Total Statistics

Source: Author's analysis results using SPSS 22.0 software

We see that the Cronbach's Alpha coefficient = 0.662 is the required value greater than 0.6; the total correlation coefficient of the measured variables is greater than 0.3, which is the required value. However, the CP4 variable has a total correlation coefficient of 0.217, which is less than 0.3; at the same time, the Cronbach's Alpha coefficient if the large CP4 variable is removed will reach 0.738. The CP4 variable " low transportation costs " is an additional measurement variable due to in-depth interviews, the survey form does not show this suggestion, and at the same time, because there is already a cheap labor cost variable, the author decided to remove this variable. The scale for removing the CP4 variable is as follows:

Table 3.30: Cost advantage - Reliability Statistics

Source: Author's analysis results using SPSS 22.0 software

After removing the CP4 variable, the scale has a Cronbach's Alpha coefficient of 0.738, greater than 0.7, which is a good scale; at the same time, the total correlation coefficient is greater than 0.5, which is very good; the Cronbach's alpha coefficient if the variable is removed is less than 0.738, so the scale now meets the requirements.

f. Reliability testing of the scale “Attractiveness of investment destination”

Cronbach's Alpha results of the scale "Attractiveness of investment destination"

Table 3.32: Attractiveness of investment destinations - Reliability Statistics

Table 3.33: Attractiveness of investment destinations - Item-Total Statistics

Source: Author's analysis results using SPSS 22.0 software

We see that Cronbach's Alpha coefficient = 0.839 is very good, the total correlation coefficient of the measured variables is greater than 0.5 which is very good. Cronbach's Alpha coefficient if the variable type is less than 0.839, so this scale is very good.

3.3.2.2 Scale testing by exploratory factor analysis EFA

a. KMO and Bartlett test results

Table 3.34: KMO and Bartlett's Test

Source: Author's analysis results using SPSS 22.0 software

The result of KMO coefficient test = 0.850 shows that this research data is very good, meeting the requirements for EFA analysis (Kaiser, 1974; Kaiser and Rice, 1974).

The Bartlett test result has a Sig coefficient = 0.000 < 0.05, which means that the observed variables used to measure the total variable are correlated with each other (Bartlett, 1937; Bartlett, 1950).

b. Exploratory factor analysis with preliminary data

In this study, the author proposes to choose a factor loading coefficient greater than or equal to 0.5. The analysis results show that the factor extraction coefficient Eigenvalue = 1.363 > 1 is satisfactory. The analysis result of the Total Variance Explained coefficient = 60.336% shows that the 5 independent variables explain 60.336% of the change in the dependent variable.

Table 3.35: Preliminary EFA analysis -Rotated Component Matrix a

Source: EFA analysis results from SPSS 22.0 software

With the above EFA analysis results, we can see that all measurement variables of the factors have content values ​​greater than 0.5; convergent values, and discriminant values ​​between factor groups. However, the variables MT1; MT8; MT9; MT10; KT6 in the scale testing section using Cronbach's alpha coefficient have not met the requirements, so the exploratory factor analysis section of the variables MT1; MT8; MT9 are retained, and the variables MT10, KT6 and CP4 are eliminated. Variable MT1: " The locality has available land and space and always creates favorable conditions for transportation."

With the above results, the author asked for more expert opinions on the results of the variables measuring the position change. The experts all agreed that this change was appropriate in terms of content and meaning. Therefore, the author will proceed to the next step of re-testing the scales for the changed scales by analyzing Cronbach's alpha one more time.

3.3.2.3 Re-validation of the new scale using Cronbach's Alpha analysis

a. Reliability testing of the “Resource advantage” scale

The results of running Cronbach's Alpha for the "Resource Advantage" scale are as follows:

Table 3.36: Resource advantages - Reliability Statistics

Table 3.37: Resource Advantage - Item-Total Statistics

Source: Author's analysis results using SPSS 22.0 software