Imputation of missing data, methodology and applications

• Grant number: 19/00181-1
• Support Opportunities: Regular Research Grants
• Start date: April 01, 2019
• End date: March 31, 2021
• Field of knowledge: Interdisciplinary Subjects
• Principal Investigator: Selene Maria Coelho Loibel
• Grantee: Selene Maria Coelho Loibel
• Host Institution: Instituto de Geociências e Ciências Exatas (IGCE). Universidade Estadual Paulista (UNESP). Campus de Rio Claro. Rio Claro , SP, Brazil
• City of the host institution: Rio Claro
• Associated researchers: Kathleen Fernandes Grego ; Simone Michaela Simons

Abstract

With the globalization process there is an increase in the volume of data produced in recent years and techniques are needed to analyze this data in order to extract useful knowledge. The cheapness of the hardware has increased the capacity of data acquisition, storage and processing, so that researchers can handle an enormous amount of data in the most varied areas of knowledge such as: economics, engineering, sociology, biology, medicine, veterinary, etc. The occurrence of incomplete data is a problem in scientific research, as well as in public databases or corporations. Determining the most appropriate statistical analysis methodology for data sets with incomplete observations is a complex problem. Several techniques of data analysis are not tolerant of missing data, which can lead to bias in the results of inferences and this subject has been discussed in the statistical literature frequently in the last decades. In this project we intend to study the methodologies of data imputation and its applications. In particular, multiple imputation, originally proposed by Rubin for analysis of public databases, will be studied, being a methodology applicable to several types of incomplete datasets considering some adaptations. (AU)