Methodology
In this work, the three methods used were the Pearson correlation, one-way ANOVA and linear regression. Since I wanted to test if the housing prices had an influence in the number of complaints coming from after/before hours constructions, it made sense to first see if these variables were related. Since the result was positive, I then tested if these differences were statistically significant. Using a one-way ANOVA made sense because there were more than two groups and this test helps in identifying the difference in means from multiple groups. Finally, the use of linear regressing helped in identifying how much the variable percentage of noise complaints contributed to the housing prices group.
Conclusions
Based on the first two tests performed, the results indicated that there is a high correlation between the percentage of noise complaints and the housing prices groups and when the ANOVA was performed, we could confirm that, in fact, there are differences in the means from each of the housing prices groups. However, the linear regression model gave us a low R-squared (36%), meaning that the variable does not contribute much to the model.
Future work
I believe that in order to improve results, it would be necessary to add more variables that could help improve the model performance.
Bibliography
Bello, J.P., Silva, C. T. , Nov,O., DuBois, R.L., Arora, A., Salamon, J., Mydlarz, C., and Doraiswamy,H., “SONYC: A system for the monitoring, analysis and mitigation of urban noise pollution,” CoRR, vol. abs/1805.00889, 2018.