Statistical Inference for Generalized log Burr Distribution | ||||
The Egyptian Statistical Journal | ||||
Article 9, Volume 43, Issue 2, December 1999, Page 209-222 | ||||
Document Type: Original Article | ||||
DOI: 10.21608/esju.1999.314531 | ||||
![]() | ||||
Abstract | ||||
In this article, statistical inference for the generalized log Burr type-XII distribution is considered. Reparametrizing the generalized log Burr type XII distribution gives a location-scale model. The inference procedures include two-sided interval estimation for the location and scale parameters and one-sided (lower) tolerance bounds for percentiles. The standard approximate intervals, for a parameter that are based on the normal approximation of the maximum likelihood estimators are widely used when the sample size is large enough. However, they are often inaccurate for small samples. In this article four bootstrap methods are used to obtain improvements over the standard intervals for the generalized log Burr type- XII distribution. Numerical comparisons via simulation studies are presented. The results suggest the preference of the bootstrap intervals to the standard ones for small and moderate sample sizes. In particular, the bootstrap-t is found to be superior to all other methods in terms of coverage. | ||||
Keywords | ||||
Bootstrap Methods; Interval Estimation; Generalized Log Burr Distribution; Location-Scale Model; Reparameterization; Statistical Inference | ||||
Statistics Article View: 34 |
||||