Linear Regression with Suburb Level Data

Dear all,

I am looking to discover what factor(s) is causing the high density of believers of a certain religion in several suburbs. In other suburbs, the densities do not reach that high level. So I am curious to find out what's driving these believers to pick those few suburbs -- is it ethnic background? Or is it income? Or is it age?

I am trying to use linear regression, but I don't have access to data at individual level, only at suburb level -- they are all just summary statistics, for example 56% = Irish, 30% = Italian... average income = $800... median age = 40, etc, etc.

Could anybody advise me on:
  1. is it alright to proceed with this type of regression?
  2. how could I improve my analysis?

Thank you in advance.


Well-Known Member
How many row data do you have? (suburbs)
I think that it isn't clear what are the dependent variables and what are the independent variables? for example, a high level of believers may encourage more believers to come.


Well-Known Member
Hi Jack,

Okay, the required sample size depends on the number of independents variables.

You can try and see if you get significant results and that you meet the linear regression assumptions (Linearity, Residual normality, continuous or ordinal independent variables (Xi), Homoscedasticity etc)