Guest
In article <63c23$453cf3f2$49ecff9$26858@DIALUPUSA.NET>,
unsettled <unsettled@nonsense.com> wrote:
/BAH
unsettled <unsettled@nonsense.com> wrote:
Lloyd Parker wrote:
In article <ehfndt$8qk_013@s799.apx1.sbo.ma.dialup.rcn.com>,
jmfbahciv@aol.com wrote:
In article <453A5164.754CBC24@hotmail.com>,
Eeyore <rabbitsfriendsandrelations@hotmail.com> wrote:
unsettled wrote:
Eeyore wrote:
The survey was to determine death rates from all causes pre and post
war. Quite simple really.
Pre-war of course there weren't any deaths from either US killings or
any insurgents.
And of course you faithfully believe Saddam's historical
records as being accurate and true! Bwahahahahahaha
The figures for the pre-war era encountered by the group tally with CIA
figures !
What era? And there aren't death certificates for those
in hidden mass graves. So any person asked about people
they know who died couldn't have shown a certificate.
This person who disappeared could have been reported by
10 households. Do you not see a problem in collected
unique datums?
/BAH
So if anything, the prewar deaths are over-reported, since you're relying
on
people to tell you, and for post-war deaths, you have death certificates.
No bannana for this one either. When you have bad data
you are not in any position to decide it has some other
value that makes it useful.
Some deaths are over reported. Others are unreported.
Still others are correctly reported. How can you
draw any valid conclusions out of such data. You might
hazzard a guess, but hazzard is the operative word
whenever you try that.
Wouldn't the standard deviation of hazzard guesses be +-50%?
/BAH