Posted by JRM on September 11, 2000 at 09:44:42:
In Reply to: Re: Explain that to me, please posted by Randy on September 09, 2000 at 18:40:50:
An in depth study won't help you here. You can count deaths, but that really won't help you. The primary missing data is the "near miss" How many things went bad when the buddy did help, vs. how many went bad when the buddy didn't.
I think that would be much more interesting a read. Most times when things go really wrong, yet we somehow manage to scrape through, it isn't logged anywhere except in our own personal list of "epics".
And statistical significance cannot be reached with this topic. There are too many biases in the data. Besides, 99.7% of all statistics are made up on the spot. The only way you could possible analyse this in any meaningfull significant (mathmatically speaking) way is to greatly narrow the scope, and then what use is that?
There is one proper mathmatical application to this problem, and that is infinity. Because it seems that will be how long this topic will be beaten. Call the rendering plant, this horse is dead!
just some thoughts,
JRM
-- I took lots of statistics, I only retained about 46.3% though.