Analiza mocy wybranych testów jednorodności czasów trwania dla populacji o rozkładzie Weibulla
The Power Analysis of Tests for Comparing Survival of Weibull Distributed Populations
Author(s): Andrzej Sokołowski, Sabina Denkowska, Kamil Fijorek, Marcin SalamagaSubject(s): Economy, Business Economy / Management, ICT Information and Communications Technologies, Socio-Economic Research
Published by: Główny Urząd Statystyczny
Keywords: survival function; comparing survival; Monte Carlo research
Summary/Abstract: Recently, tests for comparing survival distributions become more and more popular and used in applied economics. Unemployment duration, time needed to find a new job, enterprise survival or waiting for a commodity to be sold are good examples. There are a number of tests to compare survival distributions proposed in statistical literature. The aim of this research was to analyze, by the means of computer simulations, the effectiveness of survival tests as implemented in STATISTICA software. The following tests have been outlined and compared: Wilcoxon, Geehan, Cox-Mantel, Peto & Peto and log rank. Random samples were generated from Weibull distribution.
Journal: Przegląd Statystyczny
- Issue Year: 60/2013
- Issue No: 3
- Page Range: 305-323
- Page Count: 19
- Language: Polish