Embedded Files
Analysis Statement
In addition to time series analysis, we also want to know what's the relationship between the change of rental price and the change of population.
We choose rental price for all homes in metro level and get population data from the United States Census Bureau. Since the data are from different resource, we need do some preparation before we build analysis models. Cleaning the data includes matching the metro area names from two data files, linking the MSA in data file to matching MSA in tableau, and calculating the changes for both rental prices.
Population data from 2010 and estimated population in 2018 drove us to use the rental prices from December 2010 and December 2018 to calculate the change in rental price. This method stays consistent with population data. We noticed there are three MSAs which have missing data for rental in December 2010, in this case, we choose rental prices in January 2011 for those locations. Then we choose 150 MSAs by population size from largest to smallest. There are two main considerations: first, the accumulated population of top 150 MSA are nearly 238 million, composing more than 70% of total US populations, which can be adequately representative; second, high proportion of the MSAs after top 150 have no corresponding objects from rental price dataset.
The data used for the demographic analysis can be accessed here.
We used R to analyze the relationship between change of rental price and change of population, by building Linear Regression Models and Polynomial Regression Models. We evaluated the performance for each model and chose the one that has best balance between Prediction accuracy and interpret ability. Linear Regression Models satisfy a good balance with the two metrics. From the linear model, we conclude that the change of rental price has a significant positive correlation with the change of population. Given metropolitan area, the rental price is like to increase by 6.7 percent with the same population, for 1 percent more population increase, the rental price will increase 1.4 percent. The model details and visualization are showed below.
Data Spatial Patterns
Data Spatial Patterns
2018 Estimate Population (Top 150)
2018.12 Rental Price (Top 150)
Percentage of Population Change (2010.12 - 2018.12)
Percentage of Rental Price Change (2010.12 - 2018.12)
For more Maps and Graphs, please see Graphs page
Page updated
Google Sites
Report abuse