Hello I know there aren't many people here on Statalist who are keen on propensity score matching, But I thought I might as well try my luck.
I went on a course on propensity scores using matching and iptw techniques.
I had taken my own notes but lost these as I left them on the train
however.....I Still have a copy of my do file found below...
I had written to calculate my ATT from IPTW (look at section A) one would enter the following code in red... But this just gives the coefficients.
Can anyone please help me to remember where to get the overall ATT from. I'm sure I wrote these on my course notes but I lost them !
I would like to compare the ATT from IPTW with ATT from matching, as they should be similar.
WITH PSM - using -psmatch2 - one obtains a ATT automatically
With IPTW ... one performs the following steps
I went on a course on propensity scores using matching and iptw techniques.
I had taken my own notes but lost these as I left them on the train
however.....I Still have a copy of my do file found below... I had written to calculate my ATT from IPTW (look at section A) one would enter the following code in red... But this just gives the coefficients.
Can anyone please help me to remember where to get the overall ATT from. I'm sure I wrote these on my course notes but I lost them !
I would like to compare the ATT from IPTW with ATT from matching, as they should be similar.
WITH PSM - using -psmatch2 - one obtains a ATT automatically
With IPTW ... one performs the following steps
Code:
logit $treatment $covariatelist
predict pscore_iptw
//Create Propensity score weights
//weight for control
gen attweight=pscore_iptw/(1-pscore_iptw) if $treatment == 0
//creating weight for treatment
replace attweight = 1 if $treatment == 1
//Apply weights to means
mean $xlist if $treatment == 0 [pw=attweight] //control
mean $xlist if $treatment == 1 [pw=attweight] //treatment
//Calculate att - section A
reg $ylist $treatment $xlist [pw=attweight]
