Bayesian Approach to Handling Informative Nonresponse

• Grant number: 13/17746-5
• Support type: Scholarships in Brazil - Post-Doctorate
• Effective date (Start): January 01, 2014
• Effective date (End): July 13, 2015
• Field of knowledge: Physical Sciences and Mathematics - Probability and Statistics - Applied Probability and Statistics
• Principal researcher: Julio Michael Stern
• Grantee: Anna Sikov
• Home Institution: Instituto de Matemática e Estatística (IME). Universidade de São Paulo (USP). São Paulo , SP, Brazil

Abstract

In this research we propose to investigate the problem of informative nonresponse from the Bayesian perspective. On one hand, this research will constitute a natural extension to the approach developed by A. Sikov in her Ph.D dissertation. On the other hand, it will employ the results, obtained by the Bayesian research group at IME-USP. Application of the Bayesian approach permits solving the problems of hypothesis testing and model identification, which arise due to complexity of the models, being used for handling informative nonresponse. Implementation of the Full Bayesian Significance Test (FBST), which is based on the Bayesian measure of evidence for precise null hypothesis, elaborated by the Bayesian research group, seems a promising solution. In this research we also propose to modify the Bayesian approach by incorporating calibration constraints which utilize additional information not contained into the model, and to investigate the performance of the FBST in this case. In this research we assume a model for the outcome variable under complete response and a model for the response probability, which is allowed to depend on the outcome and auxiliary variables. The two models define the model holding for the outcomes observed for the responding units. We will specify prior distribution for all unknown parameters, indexing the model for the responding units and apply MCMC simulations. All the developments will be illustrated using simulations and real data set.

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

SIKOV, A.; STERN, J. M. Application of the full Bayesian significance test to model selection under informative sampling. STATISTICAL PAPERS, v. 60, n. 1, p. 89-104, FEB 2018. Web of Science Citations: 0.