Hi there,

I am estimating propensity scores using logistic regression on an imputed dataset (M=10). I want to estimate propensity scores for each imputation separately then combine the propensity scores to get the overall average. However I'm having some issues with estimating the propensity scores for each imputation in STATA version 16. Is there another method of obtaining propensity scores for each imputation?

Please see my code below and error message I'm getting:

mi estimate, i(1): logistic assessed_psych gender agegroup curtreat pretreat prev_selfharm
imputations() must contain at least 2 and at most M=10 integer numbers between 1 and 10
r(122);

predict propensity_m1


Many thanks.