SMABS 2004 Jena University
SMABS 2004 Home Organization About Jena Sponsors Links Imprint / Contact SMABS Home

European Association of Methodology

Department of methodology and evaluation research

Jena University

Contributions: Abstract

Adequacy of the non-parametric bootstrap for the construction of a confidence interval for the determination coefficient under conditions of non-normality

Alicia López-Jáuregui Paula Elosua Oliden
Universidad del Pas Vasco
Spain

The determination coefficient is the most popular index used to evaluate the adjustement for the linear regression model. Unfortunately, the complexity of its sample distribution, which depends on unknown parameters, makes it difficult to estimate its precision and to construct confidence intervals. In such a case, the bootstrap method (Efron, 1979) has revealed its usefulness.

The aim of this study is to extend previous work on the adequacy of the non-parametric bootstrap in the construction of confidence intervals for R2 (López Jáuregui y Elosua, 2003), including the consideration of non-normal data, via Monte Carlo simulations.

MCYT, code project BSO2002-00490