Missing Data
摘要
It is important to realise that in most studies not for every subject, all data is available. In other words, in almost every study there will be missing data. Regarding missing data a distinction can be made between missing completely at random (MCAR), missing at random (MAR) and missing not at random (MNAR). In this chapter it is discussed how it can be explored which missing data mechanism occurs and how to deal with missing data. In light of this, multiple imputation is briefly discussed. Finally, the chapter contains an advise in which situation imputation should be used or should not be used.