# Regression on Survey analysis across 50 states

#### vkredi

##### New Member
I ran a survey across all 50 states in the US. The goal is to see if there is a correlation between people's love for the state and the population growth % over 5 years.

The survey has 15 questions - all on Likert scale (1 to 5 Strongly Disgaree to Strongly Agree) - questions about people's love for the state. The sample has 30 respondents who were born and stayed in the state and 30 who moved into the state in the last 5 years. So a total of 60 residents from each state.

I take an average of averages for each question for each state. So I have a "Love Score" for each state. I then regress this with the population growth%.

Question1 - The final linear regression model has only 50 data points. Is this still a good model (although the overall sample size for the survey is 50 * 60 = 3000).

Question2 - I don't understand multilevel analysis yet but does this regression need multilevel analysis? If so, could you please point to some resources I can read?

Question3 - How do we draw conclusions for each of the levels? for example: Born in-state vs out-of-state, Large vs small state land size, etc.

Thank you very much and really appreciate your help! - Dave