Good morning to everybody,
I am doing a quantile estimation and I want to include the inverse ratio of mills in the following regression:
qrprocess income female school exp exp exp2 ethnicity head civil i.regions i.branch size private condition [iweight= fexp] if area==2 & age_range==1, q(0.10 0.25 0.50 0.75 0.90) vce(bootstrap, reps(1000) )
But I was reading a bit and I have to calculate manually to later include in my estimate:
*Step 1: Model selection
probit prob_work school age age2 marital boss i.regions woman [iw=fexp] if private==1
predict xb if e(sample), xb
*Step 1.2: Calculate the Inverse Mills Ratio
gene mills = normalden(-xb)/(1-normal(-xb))
label variable mills "inverse Mills ratio" //put label
qrprocess income female school exp exp exp2 ethnicity head civil i.regions i.branch size private condition mills [iweight= fexp] if area==2 & age_range==1, q(0.10 0.25 0.50 0.75 0.90) vce(bootstrap, reps(1000) )
Greetings,
Kathy
I am doing a quantile estimation and I want to include the inverse ratio of mills in the following regression:
qrprocess income female school exp exp exp2 ethnicity head civil i.regions i.branch size private condition [iweight= fexp] if area==2 & age_range==1, q(0.10 0.25 0.50 0.75 0.90) vce(bootstrap, reps(1000) )
But I was reading a bit and I have to calculate manually to later include in my estimate:
*Step 1: Model selection
probit prob_work school age age2 marital boss i.regions woman [iw=fexp] if private==1
predict xb if e(sample), xb
*Step 1.2: Calculate the Inverse Mills Ratio
gene mills = normalden(-xb)/(1-normal(-xb))
label variable mills "inverse Mills ratio" //put label
qrprocess income female school exp exp exp2 ethnicity head civil i.regions i.branch size private condition mills [iweight= fexp] if area==2 & age_range==1, q(0.10 0.25 0.50 0.75 0.90) vce(bootstrap, reps(1000) )
Greetings,
Kathy
