Re: Gam Address & Healthy survivor effect

[Martin Fraser <mfraser@channel1.com>]

I'm REALLY surprised at this thread, as it has morphed into statistics talk. I'd have expected such response from mathphobes and anti-intellectuals, but not from a group who deal with facts and figures for a living. 

Statistics do NOT lie, they're just numbers! 

If correct, they can at worst be misinterpreted, the fault of the uneducated or noncritical reader, not the presenter. To resist, abhor or poo-pooh statistics makes just as much sense as distrusting all numbers larger than 7 or dismissing the taxonomic code, or refusing to read sentences with more than 12 words (like this one). You'll have denied yourself vital tools for understanding the world around us.

I'm no statistician, but I have proper respect for this vital field of mathematics.

Mark Twain is a great and colorful writer but his words cannot be taken too literally. As for the Biostatistician who ask "what do you want to prove", this sounds like a perfectly reasonable question which HAS to be asked of (before) any study. a BAD statistician might just grab data, and sift out some correlation and then allow you to formulate a theory to explain - BAD Science. You must have a question, or theory which is to be proved (supported) or not by the given data. Any incidental finding is not properly reportable.

just my $.02

Best Wishes
Martin Fraser



- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
> Martin W. Fraser, MS, DABR
> Assistand Clinical Professor of Radiation Oncology
> CHEM Center for Radiation Oncology
> Stoneham, MA  
> (617)279-2289
mfraser@channel1.com