Re: web based health studies

Dennis Roberts (
Sat, 26 Sep 1998 16:50:45 -0400

Spike Jones wrote:

> > spike wrote:>complete medical history, form the proverbial info-nudist
> > colony, in
> > >order to
> > >determine if vitamins prevent breast cancer? spike
> >
> > Ian Goddard wrote: Extracting the signal from numerous anecdotal
> > claims might be almost impossible, particularly if you
> > factor in the possibility of intentional disinformation
> > that could be inserted freely by various interested
> > parties for one reason or another. I think we have
> > to rely ultimately on controlled laboratory research.
> >
> > A project your talking about could be done side-by-
> > side and see if it comes of similar conclusions. I'm
> > not sure I see what the privacy concerns are, assuming
> > people submit such personal information voluntarily.
> i am looking for a way to make a database to correllate vitamin
> e usage with breast cancer for instance, one not influenced by
> those who manufacture and sell vitamin e. spike

From having designed many survey forms to collect medical info, a good way to prevent individuals from submitting multiple forms is to make them tedious to complete.(Its hard to get them to fill out even the simple ones, much less do them correctly.) You may lose some potential contributors to your database who bail-out before they are finished completing your survey but its likely to deter your average data vandal and the total amount of useful info gained should be greater. Shoot for having a good approximation of the truth.

Dennis Roberts
 Boost your website with our start-up page!

"Do not be deceived by...some false secondary power, by which, in weakness,
 we create distinctions, then deem that our puny boundaries
 are things which we percieve, and not which we have made." --- Wordsworth