Online Program

Sequential Probability Ratio Test Subject to Incomplete Data with Covariate Information

*Ofer Harel, University of Connecticut 
Ellis Shaffer, University of Connecticut 
Jun Yan, University of Connecticut 

Keywords: Incomplete data, sequential analysis, SPR test

Sequential methodology is becoming more prevalent in the design of biomedical and health related studies. However, the impact of incomplete data in sequential methodology is mostly ignored. We propose a sequential probability ratio test (SPRT) subject to incomplete data with covariate information. We evaluate our procedure via simulation study and illustrate its applicability using a data example.