0

I would like to use EM algorithm to create initial values in spss.

I have used missing value analysis in spss and also selected EM and also I chose 500 times as the maximum number of iterations but after running the spss I just get one set of data! But I also got the EM means and covariance.

My question is how I can use these numbers to create initial values to be able to run imputation model?

Alex Riley
  • 169,130
  • 45
  • 262
  • 238
Ava
  • 67
  • 1
  • 9

1 Answers1

0

The MVA procedure in Statistics produces a single imputed value for each missing case. If you want to use multiple imputation, you need to use the MULTIPLE IMPUTATION procedure.

JKP
  • 5,419
  • 13
  • 5
  • THanks JKP, before multiple imputation in spss we should generate random seeds I would like to use EM algorithm to produce random seed in this way multiple imputation uses data which follow EM distribution instead of random seeds. – Ava May 06 '16 at 16:16
  • I do 't think there is a way to do that. – JKP May 08 '16 at 02:03