Comparison of Imputation Procedures Once the Data Sets Contain Nonresponses

• Activity Number: 422
• Type: Contributed
• Date/Time: Tuesday, July 31, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Ferry Butar Butar*+ and Keerthi Palitha Dias Bandaranayake and Chiaochih Chang
• Companies: Sam Houston State University and Sam Houston State University and Sam Houston State University
• Keywords: Missing data ; regression ; hot deck ; cold deck

Abstract

Non response is a common obstacle at Statistical surveys. Rather than omitting the non-responses and making inferences, many methods have been developed, in order to get the best out of the situations. Each one has its own advantages as well as disadvantages. Therefore it will be worthwhile to make a study to compare those, by using identical situations. It will produce a platform to make a comparison among different methods. Intensive study over this area may help to investigate in to clusters of experiment types such as whether particular method would be much beneficial for a certain types of experiments. Apart from the methods of handling non-response data, this study would associate large number of published Survey analysis by reputed Statistical establishments in order to identify the key areas of survey interests in the society. Elaborating currently practiced imputation methods at the initial stage of the study would pave the way towards better understanding the outcome of results which would be simulations of complete data sets where the imputations are artificially enforced and the results are compared.