# Re: different priors (flat, uniform, etc)

*From*: "Reef Fish" <large_nassua_grouper@xxxxxxxxx>*Date*: 28 Oct 2006 04:48:44 -0700

John Uebersax wrote:

Hi Ruth,

Sorry about Bob (Reef Fish). He's an embarassment to the

newsgroup--and, as is probably obvious should you look at his history

of posts, has the opposite of constructive motives.

What I said in my previous post is correct.

Cheers,

--

John Uebersax PhD

I was very explicit about what John Uebersax said was wrong about

Bayesian statistics:

JU> So I think we should just dispense with the word "prior"

JU> here.

RF> That is a completely NON-Bayesian attitude.

JU> Usually one looks to theory for this. Often one just assumes

JU> a normal distribution. That follows when, for example,

JU> the parameter reflects the joint influence of many different

factors,

JU> some positive and some negative, such that a bell-shape curve

results.

RF> That is not even a good NON-Bayesian approach. You don't

RF> simply ASSUME anything has a normal distribution or any other

RF> disribution. You use data to VALIDATE whether that is even

RF> a reasonable assumption.

RF> I could be wrong, but I believe a "flat distribution" and a

"uniform

RF> > distribution" would mean the same thing.

You WERE wrong! This is one of the most basic ideas in Bayesian

prior distribution that John Uebersax did not even know!!

RF> You are wrong in your entire understanding of what Bayesian

RF> Statistics is about. A flat distribution as used by pseudo-

RF> Bayesians is NOT necessary a uniform distribution. A uniform

RF> distribution cannot have infinite endpoints, for one thing.

JU> What I said in my previous post is correct.

What you said was completely WRONG, and you gave the OP the

worse statistical advice anyone could possibly have given.

Even Anon Bob O'Hara pointed out your error, while he made

his usual error himself which I corrected and provided the OP

the INFO about Bayesian priors relative to the UNIFORM prior:

RF> You are ALMOST correct. That's why I said Uebersax is NOT a

RF> Bayesian. We already know Bob O'Hara isn't one. :-)

RF> The posterior distribution is the likelihood function if the prior

is

RF> "diffuse" (which is NOT the same as a "uniform" or "flat" prior).

which pointed out BOTH John Uebersax and Anon O'Hara were wrong.

Then I proceeded to explain some related concepts and info about

the use the UNIFORM and other prior distributions in that post:

==== begin excerpt

For Bayesian Inference on the parameter p of a Binomial distribution

or a Bernoulli Process, the beta distribution is a member of the

conjugate prior family -- meaning both the prior AND posterior

belongs to the same distribution family -- Beta.

The uniform distribution on (0,1) is a Beta distribution with

parameters (1,1) and is an INFORMATIVE prior.

Beta(1/2, 3) is reverse J-shaped.

Beta(1/2, 1/2) is U-shaped, symmetric around 1/2.

Beta(2, 2) is symmetric unimodal, so is Beta(2,3).

Beta(2,1) is the triangular distribution on (0,1)

Beta(3,2) is unimodal, skewed to the left.

Beta(3,1) is J-shaped, so is Beta(2. 1/3).

As you can see, the Beta family CAN represent a wide

variety of opinion about p and hence is a reasonably

good APPROXIMATE prior distribution for one to choose

that best-reflects one's opinion, without having to do any

work on integration of the product of the prior and likelihood,

because the posterior distribution form can be written

IMMEDIATELY given the sample information.

That's the usefulness of a CONJUGATE prior on certain

problems (for Bayesians). However, even the family of

conjugate priors are grossly inadequate for a true Bayesian

for expressing his opinion about a particular p of a

Binomial. That's why Robert Schlaiffer had spent a large

amount of time providing numerical assessment software

and numerical integration software for just that ONE

problem (and other uniariate parameter problems) and

had written a book about it.

==== end excerpt

You, John Uebersax, who hardly EVER had anything correct

or useful to post in sci.stat.math had the GALL to make this

unsubstantiated ad hominem statement about me, in this

particular instance of my correction of your ERRORS and

ill advice? :

Sorry about Bob (Reef Fish). He's an embarassment to the

newsgroup--and, as is probably obvious should you look at his history

of posts, has the opposite of constructive motives.

Look at your OWN (John Uebersax') posting history:

Leaving The Catholic Church Is Not A Solution alt.atheism 3 hours

ago

Leaving The Catholic Church Is Not A Solution alt.atheism 3 hours

ago

different priors (flat, uniform, etc) sci.stat.math 3 hours ago

different priors (flat, uniform, etc) sci.stat.math 23 hours ago

Leaving The Catholic Church Is Not A Solution alt.atheism 23 hours

ago

Orthodoxy, postmodernity and the Emerging Church

soc.culture.south-africa 23 hours ago

Origen and reincarnation (followup) soc.history.ancient 25 hours

ago

The (infamous) regression and correlation discussion -- Summary

sci.stat.edu 3 days ago

{MEDSTATS} Correlation from disturbed data MedStats Oct 14

The current post was the one 3 hours ago. (nothing but ad hominem)

The post 23 hours ago was the one with the ERRORS and bad advice

The (infamous) regression 3 days ago was one of vacuous content

You should stick to your atheism and soc. culture and soc history

newsgroups, and leave out your POLLUTION of statistics groups!

People who live in glass houses shouldn't throw stones, John Uebersax!

JU> look at his history of posts,

This is the MOST RECENT history of MY posts (counted by Google)

511 messages sci.stat.math

61 messages sci.stat.edu

26 messages sci.math

14 messages sci.stat.consult

14 messages alt.sci.math.probability

Principal Component Analysis using R sci.stat.edu 27 minutes ago

My 300th post for the month of October sci.stat.math 7 hours ago

Testing for normality sci.stat.math 9 hours ago

Assessing credibility of a q-q plot by presence of outliers

sci.stat.math 9 hours ago

Principal Component Analysis using R sci.stat.edu 10 hours ago

Experienced Statistician to help decide whether a regression is legitim

sci.stat.math 10 hours ago

Experienced Statistician to help decide whether a regression is legitim

sci.stat.math 11 hours ago

(1 Typo correction) Re: Testing for normality sci.stat.math 15

hours ago

different priors (flat, uniform, etc) sci.stat.math 19 hours ago

Read those for STATISTICAL content and substance.

John Uebersax, you are just ANOTHER one of those who posts in

sci.stat.math who are ignorant in statistics and have nothing but

NOISE to post.

-- Reef Fish Bob.

.

**References**:**different priors (flat, uniform, etc)***From:*wtplasar@xxxxxxxxx

**Re: different priors (flat, uniform, etc)***From:*John Uebersax

- Prev by Date:
**Re: Testing for normality** - Next by Date:
**EasyFit - Distribution Fitting Software** - Previous by thread:
**Re: different priors (flat, uniform, etc)** - Next by thread:
**Re: different priors (flat, uniform, etc)** - Index(es):