This problem is too complex to choose a statistical
modeling without careful thinking.
I agree that we should focus on identifying variables first.
After that, how can we model underlying relationship, because
we don't have any information about the true
experiments; therefore, I am doubt about any statistical
models in some sense.
It is a truly challenge.
On Wed, 17 Jul 2002, Rossi Hassad wrote:
> Many prominent US academics (statistics) have been recommending complex
> statistical modeling to help with decision-making with respect to
> counter-terrorism issues, in particular bio-terrorism. It is my view that at
> this early stage, emphasis should be given to qualitative research so that
> appropriate variables can be identified, which can then be measured and
> modeled if deemed appropriate. What do you think?
>
> Thanks,
> Rossi
>
|