11-15-2011 01:20 AM
I have a question on the SAS proc MIANALYZE.
I have a dataset containing missing data, and I used proc MI (MCMC) to generate 5 imputations
Then I ran logistic regression to the imputed dataset (by_imputation), and then ran proc MIANALYZE to combine the results.
One of my variables, Education, was in 3 categories (<high school (ref), high school, >high school). The proc MIANALYZE give me the p-value of each dummy comparing to the ref categories (HS vs <HS; >HS vs <HS). However, I would like to know the overall p-value (Type-3 analysis) of Education itself to the outcome instead of two separate p-values (Education1, Education2).
The Type-3 result normally appears in the un-imputed logistic regression, but I couldn’t find it in the proc MIANALYZE.
How could I obtain an overall p-value for the variable Education in this situation?
I tried treating Education as continuous when I ran logistic and then proc MIANALYZE; however, the result doesn’t seem appropriate. Initially, when Education was treated as categorical, Education1 was significant (HS vs <HS), and the Type 3 (Education overall) in the each imputed logistic was significant (whereas Education2 - >HS vs <HS was not significant). However, when Education was treated as continuous, its p-value became very insignificant, probably due to it’s not an increasing trend (HS – the middle category was causing the significance).
I looked at the SAS proc MIANALYZE manual, and it only showed categorical variables (they call it classification) in 2 categories but not 3.
Thank you for your help.