Mortality Survey Suggestion - More Input

[Guest guest]

Regarding the sample size needed to get meaningful results out of a

mortality survey of CRONies, here are some data, using an example of

the age group from 45 to 54:

For the **non-CRON** population (males + females): out of every 100

in the survey only 0.432 would be expected to die in a given year.

If CRON is as effective in this group as it is in fruit flies then

the number of CRON participants expected to die (after at least four

years on CRON) would be 0.086. That difference would not be

noticeable, let alone statistically significant.

To notice a difference that small would require a large sample. If

we had 10,000 people in that age group then we would expect to see 43

deaths among the general population and only 9 in the CRON group.

THAT would be a useful number.

But I doubt we could find 10,000 participants for each of the six ten-

year age groups from 25 to 84.

And if we could, the administration of a survey of 60,000 people

would need a full-time employee, I think.

Oh well ...........

Rodney.

[Guest guest]

Rodney: that's the reason I never followed up on this subject

after initally getting all fired up about it. Out of the almost 700

people in our group, about 70 responded to the recent poll we took. That's

one out of 10 and that's considered high. Even counting other CRONIES

around and about, it would only be about double that number or 1500 or so.

I'm sure there are others out there in the world, but not everyone is

interested in keeping up with an online group. In fact most in the group

(as with most groups) are lurkers and wouldn't be interested in taking part

in any survey.

on 11/5/2003 7:01 PM, Rodney at perspect1111@... wrote:

> To notice a difference that small would require a large sample. If

> we had 10,000 people in that age group then we would expect to see 43

> deaths among the general population and only 9 in the CRON group.

> THAT would be a useful number.