mi predict problem

[John Franklin]

Dear all,

I am having problems trying to get the baseline survivor function after mi predict.
I am using the following code:
mi predict basesurv using miest, basesurv

and get the following error:
option basesurv not allowed r(198)

Is there a way to calculate predicted probabilities after mi predict command? I can only calculate XB after mi predict, but I would like to calculate 10-year risks for each person in my dataset. Therefore I first need baseline survivor values.. Any suggestions are very welcome..

Thanks!

[daniel klein]

When mi predict produces results, it predicts the quantity of interest in each of the imputed datasets and combines the results (at the observation level) applying "Rubin's rules"; the latter is just another way of saying that you average the M predicted values to obtain your point estimate (and combine the within- and between imputation variances in a certain way to obtain standard errors). Rubin's rules yield valid results if the quantity of interest satisfies certain conditions. One of these conditions is that the sampling distribution of the quantity of interest is (close to) normal. Linear predictions tend to follow the normal distribution; non-linear predictions, such as the baseline survivor function, might not follow a normal distribution. Hence, mi predict supports linear predictions but not non-linear predictions.

If you are confident that the sampling distribution of your baseline survivor function is close enough to normal, you can look into mi predictnl.

Best
Daniel