/*--------------------------------------------------
ex5wt.do

To calculate weights for the Ayr and Arran health survey

The first command will not be needed if 
data read in already
----------------------------------------------------*/
 use "C:\Documents and Settings\gillian raab\My Documents\aprojects\peas\web\exemp5\data\ex5wt.dta", clear

. 
/*------------------------------------------------
run a regression model for non response

Only the final model is shown here
not the model building steps

To get categorical variables in Stata a set of
dummy variables need to be created

first calculate variables making sex numeric
and calculating interaction variables
-------------------------------------------------------*/
generate sex=0
replace sex=1 if match(gender,"F")
generate agegrpsq=agegrp*agegrp
generate sexage=agegrp*sex
generate sex2age=agegrp*agegrp*sex
generate sexinc=sinc*sex
glm ayra_num agegrp sex sexage sex2age sinc sexinc sacc, f(bin npop) link(logit)
predict phat
/*------------------------------------
Stata predicts number not probability
so need to multiply weight by npop
---------------------------------------*/
generate gweight=npop/phat
/*--------------------------------------
next step is to merge this one back with the main data file for 
analysis (not shown om this site
--------------------------------------------*/