Here is my basic flow:
State Population State Electorate County Population County Electorate
M 45% 42% 40% ?
F 55% 58% 60% ?
Here is the question I am trying to resolve:
Assume state population is 45% Men and 55% Women. 58% of the state electorate in a primary is women, 42% is male.
Now assume a county in which the population is 40% Men and 60% women. Assume that if any random sample reflecting the state demographics were taken, they electorate would resemble the statewide electorate. My goal is to control for difference in county population to determine what the inferred breakdown of the county demographics are.
Bonus:
How would I resolve a situation with three variables, for example, a state that is 45% Caucasian, 35% African American, and 20% Other? For sake of an example, assume the primary electorate is 60% Caucasian, 30% African American and 10% Other, and the county is 80% Caucasian, 15% African American, and 5% Other.
State Population State Electorate County Population County Electorate
M 45% 42% 40% ?
F 55% 58% 60% ?
Here is the question I am trying to resolve:
Assume state population is 45% Men and 55% Women. 58% of the state electorate in a primary is women, 42% is male.
Now assume a county in which the population is 40% Men and 60% women. Assume that if any random sample reflecting the state demographics were taken, they electorate would resemble the statewide electorate. My goal is to control for difference in county population to determine what the inferred breakdown of the county demographics are.
Bonus:
How would I resolve a situation with three variables, for example, a state that is 45% Caucasian, 35% African American, and 20% Other? For sake of an example, assume the primary electorate is 60% Caucasian, 30% African American and 10% Other, and the county is 80% Caucasian, 15% African American, and 5% Other.