Abstract
In this paper, an approach to generate imputed values for count variables to incorporate missing data mechanism uncertainty is proposed. For multiple imputation, a distribution is considered in such a manner that it can reflect missing data mechanism uncertainty. For combining the parameter estimation of these imputed data sets the rules of nested multiple imputation are used. The performance of the multiple imputations is investigated using some simulation studies. Also, a real data set is analyzed using the proposed approach.
Recommended Citation
Farahani, E. J. and Baghfalaki, T.
(2018).
Multiple-Model Multiple Imputation for Longitudinal Count Data to Address Uncertainty in Missingness Mechanism,
Applications and Applied Mathematics: An International Journal (AAM), Vol. 13,
Iss.
1, Article 6.
Available at:
https://digitalcommons.pvamu.edu/aam/vol13/iss1/6