Algorithm to obtain critical values of ratio of maximized likelihoods in Weibull

[td2mafea@uco.es]

[Originally Posted: 1/10/00-- Transferred by ReliaSoft Moderator]

I'm using Weibull 5.0 to obtain statistical distributions of wind speeds. Statistica models used with this data are Weibull or log-normal. I'm interested to perform a test of ratio of maximized likelihoods to decide between the two models; for this, I need to obtain critical values for this test. Where can I obtain an algorithm to obtain these critical values? Anyone can hellp me? Thanks.

[Crawford Smith]

[Originally Posted: 2/3/00-- Transferred by ReliaSoft Moderator]

The criteria for rejecting the null hypothesis H_0:Theta=Theta_0 at a level of significance alpha is: -2log(L(Theta_0)/L(Theta))>ChiSquared_(1-alpha;2)