Online Program Home
My Program

Abstract Details

Activity Number: 131 - Simulation and MCMC
Type: Contributed
Date/Time: Monday, July 30, 2018 : 8:30 AM to 10:20 AM
Sponsor: Section on Statistical Computing
Abstract #329621
Title: Evaluating Sufficient Bootstrapping for Confidence Interval Estimates: a Simulation Approach
Author(s): Khairul Islam* and Tanweer Shapla
Companies: Eastern Michigan University and Eastern Michigan University
Keywords: Bootstrapping; Sufficient bootstrapping; Confidence interval; Coverage probability; Confidence length

Recent studies compare sufficient bootstrapping with conventional bootstrapping using point estimates of parameters, along their biases, relative efficiencies, etc. With numerical illustration and simulation, it claims that sufficient bootstrapping performs better than the conventional bootstrapping in certain situations. In real life, confidence interval estimates are preferable to point estimates. Confidence interval estimates take into account the variability of the point estimates for making better inference. In this paper, we provide algorithm to implement sufficient bootstrapping for constructing confidence interval estimates for several parameters such as mean, variance, standard deviation and coefficient of variation for better evaluating the performance of sufficient bootstrapping as compared to the conventional bootstrapping. A simulation study has been undertaken for evaluating confidence interval estimates using the estimated coverage probability and confidence length. This evaluation makes the recommendation for the sufficient bootstrapping stronger.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2018 program