Online Program Home
My Program

Abstract Details

Activity Number: 651
Type: Contributed
Date/Time: Thursday, August 4, 2016 : 8:30 AM to 10:20 AM
Sponsor: Section on Statistical Computing
Abstract #321117 View Presentation
Title: Uniform- and Triangular-Based Third-Order Power Method Distributions Using a Doubling Technique
Author(s): Mohan Dev Pant* and Todd Christopher Headrick
Companies: The University of Texas at Arlington and Southern Illinois University Carbondale
Keywords: Monte Carlo ; Simulation ; L-moments ; L-correlation ; Estimation
Abstract:

Power method (PM) polynomials have been used for simulating non-normal distributions in a variety of settings such as toxicology research, price risk, business-cycle features, microarray analysis, computer adaptive testing, and structural equation modeling. A majority of the applications associated with the PM polynomials are based on the method of matching conventional moments (e.g., skew and kurtosis). However, estimators of skew and kurtosis can be (a) substantially biased, (b) highly dispersed, or (c) influenced by outliers. To address this limitation, two families of third-order PM distributions are developed through the method of L-moments (Hosking, 1990) using a doubling technique (Morgenthaler & Tukey, 2000) and contrasted with the method of moments in the contexts of estimation of parameters. The methodology is based on simulating uniform- and triangular-based third-order PM distributions with specified values of L-skew and L-kurtosis. Monte Carlo simulation results indicate that the estimators based on method of L-moments are superior to their conventional moment-based counterparts.


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

Back to the full JSM 2016 program

 
 
Copyright © American Statistical Association