Estudio Estadistico Del Precio De La Vivienda Por Metro Cuadrado En Barcelona

STUDY OF SQUARE METER PRICE

1

STUDY OF HOUSES’ SQUARE METER PRICE INDEX

Summary Introduction Design Analysis Conclusions Appendix

Page 2 Page 3 Page 3 Page 4 Page 15 Page 17

1

STUDY OF HOUSES’ SQUARE METER PRICE

SUMMARY

This text studies the square meter price of Spanish province’s flats. The objective is to obtain an explanatory model to describe the price deviationsover Spain

2

STUDY OF HOUSES’ SQUARE METER PRICE

INTRODUCTION
This project is about the houses’ square meter price. Nowadays, finding a cheap accommodation is very difficult, and this is a real problem for our generation. That’s why we would like to find a model that could explain this high price. Now we have decided the response variable, which is the square meter price of all Spanishprovinces. But there are some restrictions: we have only included data of districts that have offers of more than 50 houses and they cannot be single-family houses. After that, we have to think about some explicative variable which can be important to describe the model. All of them are numerical except one, which is categorical. All those variables are: 1. Gross National Product per capita (GNP)2008 (Millions €): The GNP talks about the richness of people who live in the province. 2. Province’s area 2010 (Km2): In the total area of the province, both if it can be built or not. 3. Activity index 2010 (%): It includes all people who is working or looking for a job. 4. Unemployment index. 2010 (%): It counts people who is not included in the activity rate. 5. Natality index 2010 Births per1000 people. 6. Coast: This is the categorical variable which distinguish if they have coast or not. 7. Population density 2010 (pop/km2) : It is the division between population and area. 8. Foreign population 2010 (%): It is the percentage of foreign people versus all population. 9. Medium mortgage 2010 (1k€/house). 10. Population debt 2010 (1k€/citizen). 11. Taxis service 2010 (citizen/availabletaxis): It counts the number of citizens available per taxi. It is important to comment that we have not founded the data of two provinces, which are, Ceuta and Melilla. That’s the reason why we have excluded them from our study. So, our sample size is 50 at all.

3. DESIGN
All data used in this project have been taken from some reliable websites. First of all, the data of the response have beentaken from the website called “idealista.com”: i. ii. iii. iv. Source: Idealista.com Offer price. Only used buildings Dec2009–dec2010. No repeated houses Single-family houses not included (price deviation) 3

STUDY OF HOUSES’ SQUARE METER PRICE v. vi. vii. Out-market prices dismissed : by district, typical deviation admitted only. Only districts with +50 house offers Sample size: 466districts. 185.855 flats.

Then, all explanatory variables data have been taken from the INE (Instituto Nacional de Estadística). Year 2010 (exception: GNP 2008) Finally, in order to validate our final multiple model, we have extracted 5% of data, which are four provinces: La Rioja, Burgos, A Coruña I Màlaga. They have been extracted randomly from the database by the MINITAB.

4. ANALYSIS
Exploratoryanalysis
To find a model, we have to observe if the explanatory variables seem to have some relation with the response. To do that, first we observe these matrix plots, where we can also distinguish which of them have coast and which not. Those graphs show that there can be a positive and strong relation between the response and the GNP. We can also see this relation with some others, like themedium mortgage, but not as strong as with GNP.

4

STUDY OF HOUSES’ SQUARE METER PRICE

We cannot see clearly some regressions because of the different value between two provinces. This is the case of the density. To solve that, we have to think about some transformations. On the other hand, we have to make sure that the price is normal by doing a normality test. If it is not normal, we...