*4.1. R Program Data Analysis*

As a result of analysis using the macroscopic variable data and microscopic variable data in Section 3, the regression model itself was found to be valid, but the overall independent variables did not significantly affect the dependent variable. Therefore, using

**Figure 16.** Analysis flow diagram for the estimation of real estate land prices.

programs to determine which factors are related to land prices.

*Symmetry* **2021**, *13*, x FOR PEER REVIEW 16 of 25

**Figure 15.** Number of crimes per quarter nationwide.

**tion of Future Land Prices in the Real Estate Market** 

the backward elimination method and the stepwise selection method, the analysis was conducted based on the smallest AIC. in this analysis. Figure 16 shows the real estate data analysis through the R and Python programs to determine which factors are related to land prices.

**4. Analysis of Factors from Macroscopic and Microscopic Perspectives for the Predic-**

Prior to the data analysis, the land prices in the real estate market were analyzed based on apartments where land price transactions occurred smoothly in the living areas of each region. The land price of each regional living area was set as the independent variable, and the dependent variables are the economic activity participation rate of the young people, the economic growth rate, the unemployment rate of the young people, interest rates, the application rate of the comprehensive real estate tax and policy, local building regulations, foreign exchange reserves, the growth in the trading area, and violent crimes. The analysis was carried out with a total of 10 out of the 13 variables mentioned above; when we focused on the quarterly data from 2015 to 2018, the transportation, education, and public facilities sectors fluctuated. These were not used as variables because their effect was not big. As we all know, transportation, education, and public facilities are variables with a huge impact on real estate land prices. Unlike other areas, Gangnam-gu is a region which has been developed since the 1970s, so it has complete transportation, education, and public facilities. Thus, these three variables were not used

**Figure 16. Figure 16.** Analysis flow diagram for the estima Analysis flow diagram for the estimation of real estate land prices. tion of real estate land prices.

From the results above, it was confirmed that the dependent variable explains the large LP (Land Price) fluctuation of 92% and 99.8%. In addition, as a result of testing whether there is a significant relationship between the dependent variable and the set of all of the independent variables, the two were confirmed to be related at a significance level of 95%.

Based on Figure 17, we were able to ascertain the influence of the dependent variables and the independent variables. First, in the case of Apgujeong, we confirmed that dependent variables such as the unemployment rate, interest rate, comprehensive real estate tax, and violent crime rate affect land prices. In the case of the Second Samsung Living Area, dependent variables such as the economic growth rate, policies, land construction regulations, foreign exchange reserves, and violent crime rates were confirmed to have an effect. In the case of the third Daechi Living area, we confirmed that dependent variables such as the unemployment rate, interest rate, foreign exchange reserves, commercial growth, and violent crime rate have an effect. For the Yeoksam living sphere, eight dependent variables—the youth economic activity participation rate, economic growth rate, interest rate and policy, land construction regulation, foreign exchange reserves, commercial growth, and violent crime rate—have an effect. In the case of the fifth Gaepo living area, dependent variables such as the youth economic activity participation rate, interest rate, comprehensive real estate tax, policy, foreign exchange reserves, and commercial growth have an effect. Finally, in the case of the Suseo living areas, the youth economic activity participation rates, comprehensive real estate tax, and policy dependent variables were confirmed to have an effect.

As a result, we were able to confirm whether the ten factors affect land prices. Nonetheless, four factors policy, interest rate, violent crime rate, and foreign exchange reserves were found to have more influence than the other factors. In addition, the F TEST confirmed that the dependent variables of the six regional living spaces were significant together with the independent variables.
