Re: Question on Ratios
From: Richard Ulrich (Rich.Ulrich_at_comcast.net)
Date: 06/30/04
- Next message: Richard Ulrich: "Re: "The Ratio of Two Gaussian Fuzzy Numbers""
- Previous message: Osher Doctorow: "The Cross-Influence Partial Derivative in PI"
- In reply to: Ross Clement: "Question on Ratios"
- Messages sorted by: [ date ] [ thread ]
Date: Wed, 30 Jun 2004 10:47:38 -0400
On 30 Jun 2004 05:14:56 -0700, clemenr@wmin.ac.uk (Ross Clement)
wrote:
> Hi. I'm curious about ratios that add up to 1.0. Unexpected results
> can be found from these due to the additivity. E.g. everybody
> eventually dies, and hence if one group of people is likely to die due
> to one particular cause, then the proportions of some other causes
> must rise. E.g. I have heard (may be an urban legend) that vegetarians
> are more likely to die due to violence than the average across the
> entire population. Vegetarians claim that this is due to the lower
> chance of them dying through certain common diet-related causes.
>
> Are there statistical tools or methods that can be used to identify
> which ratios in a set of ratios that add up to 1.0 are 'unusual', and
> which are atypical due to the additivity constraint and the
> atypicality of other ratios?
If this example is what concerns you, set your mind at rest.
This is why and where we use 'survivor curves' and competing
risks; and "expected life span" as a criterion for comparing
life styles.
Is this what you mean by having a tool or a method? For life
span, there is no additivity constraint. I don't know about the
advantages of eating Vegan, but smoking and alcohol have
one-directional effects. Non-smokers live longer, but the
older smokers continue to have higher risks of heart disease
and cancer, and no particular advantage, anywhere, that I recall.
Could you state a different example, where timing does not
provide a solution?
-- Rich Ulrich, wpilib@pitt.edu http://www.pitt.edu/~wpilib/index.html
- Next message: Richard Ulrich: "Re: "The Ratio of Two Gaussian Fuzzy Numbers""
- Previous message: Osher Doctorow: "The Cross-Influence Partial Derivative in PI"
- In reply to: Ross Clement: "Question on Ratios"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
|
