R help - May 2013 - Multinomial-Dirichlet using R

Hi:

I have asked this question on Cross-Validated. So it might be a cross
posting but havent received any responses to it.

I am trying to see which distribution will best fit the data I am working
on. The dataset is as following:

    Site          Nausea      headache        Abdominal Distension
    1                   17                   5                   10
    2                    12                   8                   7
    .....

So each site has  total # adverse events for each type/category and have
equal # patients per site, say 60 and there are 63 sites. If I were to
analyze the data for multiple outcomes per site, the number of events per
category given the category response rates can be assumed to be
independently distributed. They can be modeled by a multinomial
distribution with parameters n=60 and category response rates pi1, . . . ,
piC for site i. The individual variation in category response rates can be
modeled by a Dirichlet distribution.

Just wondering if I am thinking through this correctly.
If so, could someone share some thoughts on how this could be done in R?

Really Appreciate any help.
Thanks
Anamika

	[[alternative HTML version deleted]]

On May 12, 2013, at 4:44 PM, Anamika Chaudhuri wrote:
> Hi:
> 
> I have asked this question on Cross-Validated. So it might be a cross
> posting but havent received any responses to it.
> 
> I am trying to see which distribution will best fit the data I am working
> on. The dataset is as following:
> 
>    Site          Nausea      headache        Abdominal Distension
>    1                   17                   5                   10
>    2                    12                   8                   7
>    .....
> 
> So each site has  total # adverse events for each type/category and have
> equal # patients per site, say 60 and there are 63 sites. If I were to
> analyze the data for multiple outcomes per site, the number of events per
> category given the category response rates can be assumed to be
> independently distributed. They can be modeled by a multinomial
> distribution with parameters n=60 and category response rates pi1, . . . ,
> piC for site i. The individual variation in category response rates can be
> modeled by a Dirichlet distribution.
> 
> Just wondering if I am thinking through this correctly.
> If so, could someone share some thoughts on how this could be done in R?
This is more of a statistical question, and I'm surprised no advice was
offered on CrossValidated. I responded there. You should find that this sort of
question is handled in "S-PLUS (and R) Manual to Accompany Agresti?s
Categorical Data Analysis" (2002) 2nd edition by Laura A. Thompson, 2006?,
which is very easy to find with a Google search.

-- 
David Winsemius
Alameda, CA, USA