Re: Question about the usage of Binomial dist
- From: "\"Luis A. Afonso\"" <licas_@xxxxxxxxxxx>
- Date: Thu, 17 Aug 2006 11:31:28 EDT
Response to
Bob Ling post on Aug 17, 2006 9 :32 AM.
1___I used the Z test for to decide if I must reject or not H0: pA=pB., which is a *shorthand* for the exact meaning the parameters are sufficiently apart such that we must accept that they are different (H1 true)?
This is a thing that I am repeating since several months but does not enter to the Ling´s brains.
2___Instead Ling uses the same test, but wrongly, by pooling the variances which is formally forbidden unless the samples are issued from the same Population. Concerning this error Ling does not say anything. He doesn´t because every word he says in this particular his unlearning is more and more evident.
When he wrote *used the wrong test statistics* directing to me it is paranoia in its more pure state: he is speaking to himself.
3___I suspect that Bob´s idea is that one can use the so-called Student´s test (but I restate that he chose Z with a wrong variance estimate of the difference on proportions) in order to be secure that we are not bounded to the Normality condition of the Populations. Nothing more wrong. The more general test in this area is those relative to the Behrens- Fisher problem could not get rid of this assumption!
4__A mystery that remains uncovered: WHAT IS really the area of the Yurra expertise? Who knows?
_____licas (Luis A. Afonso)
.
- Follow-Ups:
- Re: Question about the usage of Binomial dist
- From: Jack Tomsky
- Re: Question about the usage of Binomial dist
- From: Reef Fish
- Re: Question about the usage of Binomial dist
- References:
- Question about the usage of Binomial dist
- From: mmkhajah
- Question about the usage of Binomial dist
- Prev by Date: Re: Do you want it?
- Next by Date: Re: HELP ME, PLS!! - Error analysis question
- Previous by thread: Re: Question about the usage of Binomial dist
- Next by thread: Re: Question about the usage of Binomial dist
- Index(es):
Relevant Pages
|