On missing data

How can we exclude cases pairwise once we are running a MANOVA or Repeated measures ANOVA? Currently, the SPSS default is listwise exclusion which downsizes the sample.


Active Member
Pairwise exclusion does not make sense here since MANOVA models all the dependent variables jointly, as a random vector. The random vector is either observed at a particular data point or not.


TS Contributor
The alternative could be a multilevel model, such as generalized mixed model (in SPSS). These models can handle missing data.

With kind regards