I have a very general question, which I am hoping there is a standard
answer to.
I am hoping to fit regression models (with both binary and continuous
outcomes), but for some of the individuals there is uncertainty (which is
quantified) in their explanatory variable, for example individual i may be
x1 wp. 0.5 x2 wp 0.25 and x3 wp 0.25. I wonder if there is a standard way
to deal with this, my initial thought was to incorportate the probabilities
with some kind of weighting strategy (maybe some adaptation of fweights or
pweights in stata), but I am not sure if this is a common problem an/or
solution.
Any help or suggested references would be greatly appreciated.
