A NEW OPPORTUNITY FOR CLASSIFYING THE METHODS FOR HANDLING MISSING DATA
Author: Deyan Lazarov, Chief Assist. (Burgas Free University)
Keywords: missing data, missing data mechanisms, classification
The article treats the classification of the methods used in analysis of missing data. There are represent several existing classifications like Kalton and Kasprzyk (1986), Rubin and Little (2002) and Ghosh-Dastidar and Schafer (2003). In these classifications can't be found the role of the messing data mechanisms, which lead to choice from the different existing methods. Therefore a new classification is presenting, based on different approaches to handle with missing data like elimination, single imputation and multiple imputation under different mechanisms of missing data.
