Statistical Inference for the Location and Scale Parameters of the Skew Normal Distribution
- 55 Downloads
In this paper, we consider the problem of estimating the location and scale parameters of the skew normal distribution introduced by Azzalini. For this distribution, the classic maximum likelihood estimators(MLEs) do not take explicit forms. We approximate the likelihood equations and derive explicit estimators of the parameters. The bias and variance of the estimators are investigated and Monte Carlo simulation studies show that the estimators are as efficient as the classic MLEs. We demonstrate that the probability coverages of the pivotal quantities (for location and scale parameters) based on asymptotic normality are unsatisfactory, especially when the sample size is small. The use of unconditional simulated percentage points of these quantities is suggested. Finally, a numerical example is used to illustrate the proposed inference methods.
Key wordsSkew normal distribution maximum-likelihood estimator Monte Carlo simulation probability coverage pivotal quantity
Unable to display preview. Download preview PDF.
- 1.B. Arnold and N. Balakrishnan, Relations, bounds and approximations for order statistics, 53 (2012), Springer Science and Business Media.Google Scholar
- 4.A. Azzalini, A class of distributions which includes the normal ones, Scandinavian Journal of Statistics, (1985), 171–178.Google Scholar
- 5.A. Azzalini, The skew-normal distribution and related multivariate families, Scandinavian Journal of Statistics, (2005), 159–188.Google Scholar
- 9.T. S. Ferguson, A course in large sample theory, Vol. 49 (1996), Chapman and Hall London.Google Scholar
- 10.S. Ghosh and D. Dey, Estimation of the parameters of skew normal distribution by approximating the ratio of the normal density and distribution functions, Dissertations and Theses - Gradworks, (2010).Google Scholar
- 12.N. Henze, A probabilistic representation of the’ skew-normal’ distribution, Scandinavian Journal of Statistics, (1986), 271–275.Google Scholar
- 15.B. Liseo and N. Loperfido, Default bayesian analysis of the skew-normal distribution (2004).Google Scholar
- 18.H. V. Roberts, Data analysis for managers with Minitab: Harry V. Roberts, Scientific Press, (1991).Google Scholar