# Population Density vs Income

# Individual Work - GIS-based Research

# Explore if there is a relationship between population density and median income in New York state?

# Histogram of median income in New York state

# The mean of median income is 64403.3 (± 33402.4). And the histogram shows the distribution of median income meets the normal distribution, which means that the data owns a good quality.

# Distribution Map And Hot Spot Map of median income in New York state

# The figure reveals the distribution of median income in New York state. From the map, we could see the median income variable shows a kind of cluster distribution on the map.

# Histogram of population density in New York state

# The figure shows the mean of population density from all censuses is 64.046 (± 86.894) and its distribution in the statistic.

# Distribution Map and Hot Spot Map of median income in New York state

# From the figure, we could see population density’s city-centered distribution on the map, which is similar with median income distribution.

# Comparison of maps of median income and population density in New York state

# The distribution maps of median income and population density reveal the similar distribution of these two variables. To prove if there is a correlation between these two variables, we used a bivariate map and correlation analysis to explore it.

# The figure below shows the median income and population density in New York state. From this map, we could see in the high population density areas (red area), there is a higher median income (bigger blue circle). By contrast, there is a lower median income (small blue circle) in the low population density areas (green area). The map shows a possible relationship between median income and population density.

# Correlation Analysis

# To prove this hypothesis, Minitab software was used to test the correlation between these two variables. The table below reveals the result of calculation, and the correlation index is -0.207523(P-value <0.001). In statistic, when correlation index is bigger than 0.7 or less than -0.7, we may conclude that there is a strong correlation between two variables. Because of the correlation index is only -0.207523, we may not conclude there is a strong correlation between median income and population density, the conclusion is different from our observation from the maps.

# Correlation between median income and population density in New York state at the country level.

# Explore possible regression model.

# Limitation

Difference between living location and working location

The median income data are from households, which is based on people’s living location. However, the area (census tract area, country, city) of people’s living location may be different from the working location. As a result, even people living in areas with low population density, they may receive high income by working in areas

with a high population density. It is pretty possible especially considering that two situations: 1) census tract area is dived pretty small in the data we used, 2) lower house density in the wealthy community around cities.

More specific attributes of population density, such as education, age, and family size

In this research, we only explored the population density in one area, but we did not explore more attributes of population density, such as density of education, age and family size. It is important to know what attribute contribute more or less in population density, which could help us explore more deeply in the possible correlation between population density and median income.

# Conclusion

This project explored the possible correlation between median income and population density in New York state and generated several maps to show the distribution of these two variables. The results showed both median income and population density have a statistically significant cluster in terms of geography. Even though there is a similar distribution of median income and population density on the map, the correlation analysis could not conclude a strong correlation between these two variables

.

As discussed in the limitation part, this research did not handle the difference of living location and working location, which may cause a mistake in the correlation analysis. To solve this problem, one future plan is to adjust the basic track area. For example, we may define an area including most location areas and working areas. As a result, the population density and income would come from one basic area. And also, another possible future work is to explore more attributes of population density, such as density of population with different education, age backgrounds.

# References

Arouri, M., Youssef, A. B., & Nguyen, C. (2017). Does urbanization reduce rural poverty? Evidence from Vietnam. Economic Modelling, 60, 253-270. doi:10.1016/j.econmod.2016.09.022

BoÌˆdeker, M., Finne, E., Kerr, J., & Bucksch, J. (2018). Active travel despite motorcar access. A city-wide, GIS-based multilevel study on neighborhood walkability and active travel in Germany. Journal of Transport & Health,9, 8-18. doi:10.1016/j.jth.2018.03.009

Wu, D., & Rao, P. (2016). Urbanization and Income Inequality in China: An Empirical Investigation at Provincial Level. Social Indicators Research,131(1), 189- 214. doi:10.1007/s11205-016-1229-1.