when using the MCAR test for analyzing the missing values, do you only put the items or scales for which you know there are missings into the "Incomplete variables" or does the system need the other complete variables of your dataset and analysis to estimate the pattern? I originally thought the "complete variables" tab was only if you wanted to do multiple imputation, but I´m not so sure anymore and have never really dealt with Little´s test before.
Thank you for any help
Hi,
For regression analysis of explainers (Evaluation of MAR explainers), all variables (complete + incomplete) are used.
The basis for imputation is created using the full array of variables you selected.
Including incomplete variables in a separate group allows you to form output tables and pseudo variables only for them.
