Centering Data According to Age Group after MI

[analyze28]

Hi,

I need some advice.  I have conducted multiple imputations on my data, but now need to centre the continuous variables so that I can form an interaction for GZLM analysis.  As I am stratifying my analysis by age, I have written syntax accordingly so that the data would be centered by imputation and age group i.e. Original data 20's/Imputation 1 20's/Imputation 2 20's and so forth.

I'm having trouble writing the syntax that would allow me to run this for each section separately.  I have tried using the DATA, SELECT CASES, but that doesn't work.  Can anyone tell me how to run this without having to split the dataset down to Imputations and Age (which would result in 6 separate datasets for each age group!).

Many thanks

Danica

[David Marso]

Take a look at the AGGREGATE command.

analyze28 wrote

Hi,

I need some advice.  I have conducted multiple imputations on my data, but now need to centre the continuous variables so that I can form an interaction for GZLM analysis.  As I am stratifying my analysis by age, I have written syntax accordingly so that the data would be centered by imputation and age group i.e. Original data 20's/Imputation 1 20's/Imputation 2 20's and so forth.

I'm having trouble writing the syntax that would allow me to run this for each section separately.  I have tried using the DATA, SELECT CASES, but that doesn't work.  Can anyone tell me how to run this without having to split the dataset down to Imputations and Age (which would result in 6 separate datasets for each age group!).

Many thanks

Danica