Advised missing data method

Longitudinal analysis > Missings in predictor/covariate > MAR > <10% missing data > >75% missing item scores

The imputation method should be applied to the total scores directly. With this relatively small percentage of subjects with missing data, a stochastic regression imputation of the total scores may perform satisfactory. Even though, the missing data uncertainty is not incorporated in this method, in this situation where a small percentage of subjects have missing total scores (<10%), the effect of not including this uncertainty is negligible. Nevertheless, multiple imputation of the total scores would be more corect and can be applied in these situations as well.