Re: Critical Values of the Pearson´s Coefficient of Skewness
- From: "Luis A. Afonso" <licas_@xxxxxxxxxxx>
- Date: Mon, 06 Aug 2007 04:24:55 EDT
Afonso misleadingly described his methods as being "exact". Yet everytime he runs his simulations, he gets a different result. In contrast, a mathematical result in the form of an equation is exact.
My response
The Jack Tomsky growing dyslexic illness prevent him to find the difference between a model exact (Monte Carlo) and a numerically exact model. I had the opportunity to stress that the conventional way to present mathematical statistics in what concerns test hypotheses is , generally speaking, a full mess of solving it, substituting the proposed problem by other, different, that has analytical solution and furthermore this solution is only correct asymptotically (infinite data). We use a twice wrong algorithm.
This is known from all time, teachers in Math. Statistics do not fail to note it.
Monte Carlo is, by its own construction, model correct, however they could not provide values free of random fluctuations. All Statisticians agree that Monte Carlo is an accurate tool to provide solutions free from the pointing out errors. Thy have being used thoroughly and they will forever.
______________
Luis
.
- References:
- Critical Values of the Pearson´s Coefficient of Skewness
- From: Luis A. Afonso
- Critical Values of the Pearson´s Coefficient of Skewness
- Prev by Date: APPEAL: Sci. Stat. Math. MODERATED
- Next by Date: Re: Density Functions
- Previous by thread: Re: Critical Values of the Pearson´s Coefficient of Skewness
- Next by thread: Re: Name of a distribution.
- Index(es):
Relevant Pages
|
