Academic Journals Database
Disseminating quality controlled scientific knowledge

Multiple Imputation by Chained Equations: An Overview of Conceptual and Operational Aspects and Software: Review

Author(s): Jennifer L. BEAUMONT | Hakan DEMİRTAŞ

Journal: Turkiye Klinikleri Journal of Biostatistics
ISSN 1308-7894

Volume: 5;
Issue: 1;
Start page: 29;
Date: 2013;
Original page

Keywords: Missing data | multiple imputation | chained equations

Missing data are prevalent in nearly all areas of research. It is of utmost importance that appropriate methods are used to obtain statistically valid inferences in the presence of missing data. Multiple imputation relies on the creation of multiple sets of plausible values for the missing data. We provide background on missing data and its consequences, and describe the fundamental concepts underlying multiple imputation. We then describe a flexible implementation of multiple imputation, called multiple imputation by chained equations, that allows the analyst to specify a separate conditional regression for each variable with missing data. We conclude with an overview of software options for multiple imputation by chained equations.
Save time & money - Smart Internet Solutions      Why do you need a reservation system?