Dendreon Corp fka DNDNQ

[nerdseeksblonde]

my point of course,
is that some other group claimed to know it
was wrong and, while you always get naysayers
and after the fact selection of prophets
has known problems ( much likse regression models
can be questioned ), they seemed to have some
reasons to be suspicious. I'm just advocating
spending time reconiling results with other known
information. In the DNDN case, with limited clinical
data, there were plenty of opportunities for this
and the best they did was find another immune assay
and ignore the earlier concerns about prognositc
and drug effect components in an observed correlation.

Any derived result- a computed structure or a p value-
needs to be harmonized with other "known" results.

And, the simple approach you suggest is inline with my
comments on IV - simple scatter plots and other baby steps for simple minds
( the ignorant and stupid people ) could make a result
a lot more credible.