I have some independent values, e.g. percentage of head of the household with low low education per zone, percentage of unemployed people per zone [...] The percentage is related to the total of people in each zone, and not from the total population of the city. I think if I use the absolute value of each variable instead of percentage it might be better. I other side if I use the absolute values it will give me some correlation (multicollinearity) among independent variable, since that zone with more population tends to have also higher number of people with low education, unemployed people and so on.

Thank you in advance for your cooperation,

Patrik Silva

More Patrik Silva's questions See All
Similar questions and discussions