comparatively studies for missing data in regression analysis

ABSTRACT

The statistical analysis with missing data is an important applied problem because missing data are commonly encountered in practice and most data analysis procedures were not designed for missing data. In this paper, we introduce a recent comparison for the methodologies that handle missing data problems. Some real examples of missing data problems, the pattern of missing data and the mechanisms that lead to missing data will be introduced in the case of regression analysis. A comparative study among the complete case [CC] method, the available case [AC] method, the least squares [LS] on imputed data and the maximum likelihood [ML] are presented. A real data and GPA scores for undergraduate students, College of Administrative Sciences at King Saud University in 1998, is used to make an artificial missing data in three different types of the mechanisms Missing At Random [MAR], Missing Completely At Random [MCAR], and Missing Not At Random [MNAR] Then we reanalyze the new data set, using a Monte Carlo simulation, for the different methods to make some comparisons