Munich Personal RePEc Archive

Normality Testing- A New Direction

Islam, Tanweer ul (2008): Normality Testing- A New Direction.

[img]
Preview
PDF
MPRA_paper_16452.pdf

Download (349Kb) | Preview

Abstract

Abstract

This paper is concerned with the evaluation of the performance of the normality tests to ensure the validity of the t-statistics used for assessing significance of regressors in a regression model. For this purpose, we have explored 40 distributions to find the most damaging distribution on the t-statistic. Power comparisons are conducted to find the best performing test against these distributions. It is found that Anderson-Darling statistic is the best option among the five normality tests, Jarque-Bera, Shapiro-Francia, D’Agostino & Pearson, Anderson-Darling & Lilliefors.

UB_LMU-Logo
MPRA is a RePEc service hosted by
the Munich University Library in Germany.