Autocorrelation
Autocorrelation is a phenomenon in which errors depend on each other.
are correlated with each other, leading to inefficient t- and F-tests, as well as incorrect estimates.
R. Through the above analysis results, we can see that the Durbin Watson coefficient = 1.778 is in the range (1.6; 2.6). Therefore, the model does not have autocorrelation.
Table 2.13: ANOVA analysis
Model
Sum of squares | df | Mean square | F | Sig. | ||
1 | Regression | 89,336 | 7 | 12,762 | 30,374 | 0.000 |
Balance | 59,664 | 142 | 0.420 | |||
Total | 149,000 | 149 |
Maybe you are interested!
-
Tourism Products and Product Systems of Travel Agencies -
Building a strategy to position teambuilding tourism products in the domestic market at Saigontourist Travel Service Company - 2 -
Tourist Reviews on Diversity in European Tours of Viet Travel Joint Stock Company – Vietnam Travel -
Solution to Diversify European Tourism Products at Viet Travel Joint Stock Company - Vietnam Tourism -
Analyzing Customer Reviews About The Official Website Of Hues Discount Card Program Of Tan Nguyen Mtv Company Limited.
(Source: Data processing results via SPSS software)
We can see that Sig. = 0.000 < 0.05, so we reject the hypothesis H0, which means that this regression model, after being extrapolated to the population, its suitability has been verified. In other words, there is at least one independent variable that affects tourists' decision to choose the 1-day Hue tour product.
Multicollinearity
Multicollinearity is a phenomenon in which independent variables have a very strong correlation with each other. A regression model with multicollinearity will cause many indicators to be distorted, leading to the results of quantitative analysis no longer being meaningful. Therefore, checking this phenomenon is based on the VIF index (Variance inflation factor). According to Hoang Trong & Chu Nguyen Mong Ngoc (2008). When the VIF value exceeds 10, it is a sign of multicollinearity. However, in reality, with research topics that have models and questionnaires using Likert scales, VIF < 2 will not cause multicollinearity. In this case, the VIF values of the independent variables are all less than 2.0, so multicollinearity does not occur. Therefore, the relationship between these independent variables does not significantly affect the results of the regression model explanation.
Table 2.14: Regression analysis results
Variables
Unstandardized coefficient | Standardization factor | t | Significance level Sig. | Multicollinearity statistics | ||||
B | Error | Beta | Tolerance | VIF | ||||
1 | Constant | -1,536E-017 | 0.053 | 0.000 | 1,000 | |||
GC | 0.369 | 0.053 | 0.369 | 6,952 | 0.000 | 1,000 | 1,000 | |
TD | 0.260 | 0.053 | 0.260 | 4,894 | 0.000 | 1,000 | 1,000 | |
Designer | 0.245 | 0.053 | 0.245 | 4,607 | 0.000 | 1,000 | 1,000 | |
KN | 0.291 | 0.053 | 0.291 | 5,479 | 0.000 | 1,000 | 1,000 | |
QC | 0.166 | 0.053 | 0.166 | 3,119 | 0.002 | 1,000 | 1,000 | |
CL | 0.348 | 0.053 | 0.348 | 6,555 | 0.000 | 1,000 | 1,000 | |
DD | 0.320 | 0.053 | 0.320 | 6,033 | 0.000 | 1,000 | 1,000 | |
(Source: Data processing results via SPSS software)
Note:
- GC – Variable representing the factor group “Tour price”
- TD – Variable representing the factor group “Tourism attitude”
- NTK – Variable representing the factor group “Reference group”
- KN – Variable representing the factor group “Travel experience”
- QC – Variable representing the factor group “Tour advertising”
- CL – Variable representing the factor group “Availability and quality of tours”
- DD – Variable representing the factor group “Tour booking location”
- The constant in the regression equation represents the slope coefficient, it does not go with the variable so it does not affect the equation. Especially in models using the Likert scale, this constant has no comment meaning, so the Sig. of the constant, whether it is larger or smaller than 0.05, the constant is negative or positive, is not important.
From the regression results, it shows that the above variables Sig. < 0.05 so the variables still have an impact on the decision to choose the Hue 1-day tour product. In addition, the free constant has Sig. test t = 1.000 > 0.05 so it is not significant in the model.
Unstandardized regression equation
QD = 0.369*GC + 0.348*CL + 0.320*DD + 0.291*KN + 0.260*TD +
0.245*NTK + 0.166*QC Unstandardized regression equation has mathematical meaning
more academic than economic significance, it only reflects the change of the dependent variable "Decision to buy Hue 1-day tour product" when each independent variable changes (conditionally constraining that the remaining independent variables are fixed).
According to the standardized coefficients of the independent variables in the model, the values are as follows: “Tour price” = 0.369; “Tour availability and quality” = 0.348; “Tour booking location” = 0.320; “Travel experience” = 0.291; “Travel attitude” = 0.260; “Reference group” = 0.245; “Tour advertising” = 0.166;. We have the following regression model:
QD = 0.369*GC + 0.348*CL + 0.320*DD + 0.291*KN + 0.260*TD + 0.245*NTK + 0.166*QC
Or to write it more clearly:
Decision to choose Hue 1-day tour product = 0.369*Tour price + 0.348*Tour availability and quality + 0.320*Tour booking location + 0.291*Travel experience + 0.260*Travel attitude + 0.245*Reference group + 0.166*Tour advertisement
The standardized regression equation has more economic meaning than mathematical meaning, indicating which factor has the greatest impact (with the largest standardized regression coefficient), and which factor has the weakest impact.
Meaning of regression coefficients in the model
Based on the results of correlation coefficient testing and multivariate regression, it was shown that 7 groups of explanatory variables influence tourists' decision to choose Hue 1-day tour products, described by the following model:
Tour price
Tour availability and quality
0.348
Tour booking location
0.320
Travel experience
0.291
Decide to choose Hue 1 day tour product
0.260
Travel attitude
0.245
Reference group
0.166
Tour advertising
0.369
Figure 2.3: Post-regression correction model
Regression model results
The coefficient '..i shows the influence of factors on tourists' purchasing decisions for the Hue 1-day tour product at Dai Bang Advertising and Tourism Services Joint Stock Company. Thereby, it shows how each dependent variable influences. At the same time, the (+) sign in the regression coefficient shows that there is a positive relationship between the independent variables in the regression model above and the dependent variable. The meaning of the coefficient of each independent variable is shown through the regression results as follows:
First is “Tour price” with a standardized coefficient of 0.369. This is the explanatory variable with the largest coefficient in the regression model, with a significance level of Sig. t = 0.000 < 0.05, so this variable has the largest influence on the dependent variable “Decision to choose Hue 1-day tour product”. Because there is a positive relationship between them.
As explained above, therefore when the “Tour Price” increases by one unit,
“Decision to choose Hue 1 day tour product” increased by 0.369 units.
The second is “Availability and quality of tour” with a standardized coefficient of 0.348. This is the second largest coefficient in the regression model, with a significance level of Sig. t = 0.000 < 0.05, so this variable has the second influence on the dependent variable “Decision to choose Hue 1-day tour product”. When “Availability and quality of tour” increases by 1 unit, “Decision to choose Hue 1-day tour product” increases by 0.348 units.
Third is “Tour booking location” with a standardized coefficient of 0.320. This is the third largest coefficient in the regression model, with a significance level of Sig. t = 0.000 < 0.05, so this variable has the third influence on the dependent variable “Decision to choose Hue 1-day tour product”. When “Tour booking location” increases by one unit, “Decision to choose Hue 1-day tour product” increases by 0.320 units.
Fourth, “Travel experience” has a standardized coefficient of 0.291 with a significance level Sig. t test = 0.000 < 0.05, so it is statistically significant. At the same time, the positive relationship with the dependent variable is shown when “Travel experience” increases by one unit, “Decision to choose Hue 1-day tour product” increases by 0.291 units.
Fifth is “Tourism attitude” with a standardized coefficient of 0.260 with a significance level Sig. test t = 0.000 < 0.05, so it is statistically significant. At the same time, the positive relationship with the dependent variable is shown when “Tourism attitude” increases by one unit, “Decision to choose Hue 1-day tour product” increases by 0.260 units.
Sixth, the "Reference Group" has a standardized coefficient of 0.245 with a significance level of Sig. t test = 0.000 < 0.05, so it is statistically significant. At the same time, the positive relationship with the dependent variable is shown when "Travel experience" increases by one unit, "Decision to choose Hue 1-day tour product" increases by 0.245 units.
Seventh, “Tour advertising” has a standardized coefficient of 0.166 with a significance level of Sig. t = 0.000 < 0.05, so it is statistically significant. At the same time, the positive relationship with the dependent variable is shown when “Tour advertising” increases by one unit, “Decision to choose Hue 1-day tour product” increases by 0.166 units.
2.2.4 Customer reviews of tourism products of Dai Bang Tourism Service and Media Joint Stock Company.
After determining the factors that actually influence customers' choice decisions as well as the level of influence, we proceed to analyze customers' evaluation of each group of factors through the results of the interview survey that the study had collected before. The research questionnaire uses the Likert scale with 5 levels of customer interest as follows:
1
2 | 3 | 4 | 5 | |
Strongly disagree | Disagree | Neutral | Agree | Totally agree |
In this study, the hypothesis H0 is: The average rating score of tourists for the factors "Tour price", "Tour availability and quality", "Tour booking location", "Travel experience", "Travel attitude", "Reference group" and "Tour advertising" is equal to 3 (95% confidence level), which is at a neutral level.
If Sig. in the One-Sample test table < 0.05, it means rejecting H0, or in other words, other factors are neutral. Then use the average value in the One-Sample Statistics table to examine the value and draw conclusions.
If Sig. in the One-Sample test table > 0.05, it means accepting H0,
or in other words, the average rating with factors at a neutral level.
2.2.4.1 Customer reviews for the “tour price” group
Table 2.15 One Sample T-test of tour price group
One Sample T-test
Criteria | Mean | Sig.(2-tailed) |
Reasonable tour price | 4.15 | 0.000 |
The company has many incentives for tours. | 3.61 | 0.000 |
Diverse payment methods | 3.97 | 0.000 |
Assumption:
H0: µ = 3 (Sig. >0.05) H1: µ ≠ 3 (Sig. < 0.05)
Based on the above results, it can be seen that all three criteria GC1, GC2 and GC3 have a significance level of 0.000, accepting the hypothesis H1: µ ≠ 3. So we will consider the average value of these criteria to draw conclusions.
For criterion GC1, the average value is 4.15, proving that customers feel that the price of the 1-day Hue tour is reasonable and highly agreed by customers.
As for the criteria GC2 and GC3, the values are 3.61 and 3.97, reflecting that the company has preferential policies for tours. However, it is necessary to add more payment methods and complete preferential policies for tours.
2.2.4.2 Customer reviews for the group “Tour availability and quality”
Table 2.16 One Sample T-test Group Tour Availability and Quality
One Sample T-test
Criteria | Mean | Sig.(2-tailed) |
Tours are always available and varied. | 3.93 | 0.000 |
Tour has many attractive destinations | 4.03 | 0.000 |
Tour quality is guaranteed | 3.76 | 0.000 |
Assumption:
H0: µ = 3 (Sig. >0.05) H1: µ ≠ 3 (Sig. < 0.05)
Based on the above results, it can be seen that all three criteria CL1, CL2 and CL3 have a significance level of 0.000, accepting the hypothesis H1: µ ≠ 3. So we will consider the average value of these criteria to draw conclusions.
For criterion CL2, the average value is 4.03, proving that customers feel that the tour has many attractive destinations that are highly appreciated by customers. Showing that the company makes efforts to exploit locations and does a good job of creating a 1-day Hue tour.
CL1 criteria is quite agreed by customers but not yet completed.
all when the mean value is 3.93
The CL3 criterion is rated lower at 3.76. The company needs to create satisfaction. Review previous trips to gain experience for the next trip to be guaranteed.
2.2.4.3 Customer reviews for the “Tour booking location” group
Table 2.16 One Sample T-test for Tour Booking Location Group
One Sample T-test
Criteria | Mean | Sig.(2-tailed) |
Convenient tour booking location | 4.20 | 0.000 |
Tours can be booked by phone. | 3.74 | 0.000 |
Tours can be booked online. | 3.51 | 0.000 |
Assumption:
H0: µ = 3 (Sig. >0.05) H1: µ ≠ 3 (Sig. < 0.05)
Based on the above results, it can be seen that all three criteria DD1, DD2 and DD3 have a significance level of 0.000, accepting the hypothesis H1: µ ≠ 3. So we will consider the average value of these criteria to draw a conclusion.
For criterion DD1, with an average value of 4.20, this criterion has achieved a very good value of 4.20. The location of the tour booking for customers is convenient for customers to book tours.
Criteria DD2 and DD3 have average values of 3.74 and 3.51 respectively, showing that the company needs to try harder to improve tour booking activities via phone and internet to ensure the most convenience for customers.
2.2.4.4 Customer evaluation of the “Travel experience” group Table 2.17 One Sample T-test of the Travel experience group
One Sample T-test
Criteria | Mean | Sig.(2-tailed) |
I have a lot of experience joining Hue 1 day tour | 3.91 | 0.000 |
I have joined the Hue 1 day tour more than 2 times. | 3.83 | 0.000 |
I was satisfied with the 1 day Hue tour. | 3.87 | 0.000 |
Assumption: