Hi!
I have a covariate, Ar, that is a measure of Arsenic level in human blood. This covariate contains some missing, due to values being below the Level of Detection (that is, the lowest concentration level of a substance that can be determined to be statistically different from a blank value with a stated confidence level.) I would like to do multiple imputation to replace missing values with values being constrained between 0 and Level of Detection, using, I believe, some sort of distribution-based MI methods for left-censored data (A paper that did something similar, is Chen et al. 2011: 10.1289/ehp.1002124, who used a nonlinear optimization routine by Newton-Raphson ridge method in SAS IML) . Does anyone know how to do this in Stata 15?
Best,
Kjell
I have a covariate, Ar, that is a measure of Arsenic level in human blood. This covariate contains some missing, due to values being below the Level of Detection (that is, the lowest concentration level of a substance that can be determined to be statistically different from a blank value with a stated confidence level.) I would like to do multiple imputation to replace missing values with values being constrained between 0 and Level of Detection, using, I believe, some sort of distribution-based MI methods for left-censored data (A paper that did something similar, is Chen et al. 2011: 10.1289/ehp.1002124, who used a nonlinear optimization routine by Newton-Raphson ridge method in SAS IML) . Does anyone know how to do this in Stata 15?
Best,
Kjell
