I found it very interesting to note that the average population size for the metropolitan areas with the largest mutual information values was significantly higher than the average population size for those with the lowest mutual information values.
Conversely, the average population size was actually higher for the micropolitan areas with the lowest MI as compared to those with the highest MI. That the averages for the micropolitan areas are so close is most likely due to the fact that there is an upper and lower bound for population size built into the definition of micropolitan areas. However, we do not see a concentration of areas with high MI near the upper bound nor do we see a concentration of low MI areas near the lower bound.
This would suggest a somewhat parabolic relationship between MI and population size, but this relationship may be highly affected by bin size for micropolitan and metropolitan areas. You noted that faster growing areas in Texas had high MI values, but has any work been done to determine the relationship between income segregation and absolute population?
I found it very interesting to note that the average population size for the metropolitan areas with the largest mutual information values was significantly higher than the average population size for those with the lowest mutual information values.
Conversely, the average population size was actually higher for the micropolitan areas with the lowest MI as compared to those with the highest MI. That the averages for the micropolitan areas are so close is most likely due to the fact that there is an upper and lower bound for population size built into the definition of micropolitan areas. However, we do not see a concentration of areas with high MI near the upper bound nor do we see a concentration of low MI areas near the lower bound.
This would suggest a somewhat parabolic relationship between MI and population size, but this relationship may be highly affected by bin size for micropolitan and metropolitan areas. You noted that faster growing areas in Texas had high MI values, but has any work been done to determine the relationship between income segregation and absolute population?