[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: prediction after regress

From   "Eva Poen" <>
Subject   Re: st: prediction after regress
Date   Fri, 22 Aug 2008 10:04:44 +0100


Yes, this does what you want. Alternatively, you could temporarily
recode your dummy variable and use -predict-:

replace insurance = 1-insurance
predict counterfactual if e(sample)
replace insurance = 1-insurance

Hope this helps,

2008/8/22 Shehzad Ali <>:
> Hi listers,
> I am trying to get counterfactual predictions after regress. Basically,
> people in my sample are either insured or uninsured. I want to predict how
> much insured would spend if they were uninsured (assuming their other
> characteristics remained the same) and vice versa for the uninsured. Can
> -adjust- help in this situation? Here is what I was trying but not sure if
> this is right:
> reg y $xvar insurance_dummy
> adjust insurance=1 if insurance==0, gen(counter_1)
> adjust insurance=0 if insurance==1, gen(counter_0)
> I would appreciate any help.
> Thank you,
> Shehzad
*   For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index