R-Squared turned out to be much higher after setting entropy index as the dependent variable. It kind of tells that low-income housing density can affect a census tract's racial diversity in a negative way. Based on the model, it shows that a 1% increase in affordable housing density will decrease the entropy index by 0.12% at 95% confidence level. Plus, it is interesting to see that there is a positive relationship between Gini index and entropy index. In other words, It says that a census tract with racial segregation issues usually also has income inequity issues.

Limitation

Since affordable housing has existed for many decades, the 7-year time frame used cannot sufficiently reflect the things changes occurring in the past. Secondly, there might be multicollinearity existing among different independent variables. For this reason, the coefficient result noted from the regression analysis could be biased. As well, much of the census tract level data show very high levels of margin of error. Since the data I used could be very biased, the model has been created cannot be perfect anyway.

Reference

1.Social explorer — ACS neighborhood feature table by census tract