Re: When is a Year Not a Year?

[Dann Pigdon <dannj@alphalink.com.au>]

philidor11 wrote:
> 
> So, if I'm following correctly, he means that:
> - dates have a substantial range of error or
> - dates are not expressed in calendar years or
> - both.
> I'm inclined to think that the correct answer is that when mya dates are
> given for the Maastrichtian, for example, that these are referring to
> calendar years, but that they could be off.
> If so, is there a good rule of thumb possible error percentage to be
> applied?

All dates are provided with an error range. When you see "125 MYA" this
is usually an average. Lets say the date is 125 +/- 5 MYA. This means
there is a 65% chance of the actual date being within one standard
deviation (that is, 120-130 MYA). There is a 95% chance of it lying with
two standard deviations (115-135 MYA). The chance of the "actual" date
being precisely 125 MYA is unlikely, however it IS likely to be
somewhere around that figure. Of course, just because something is
statistically "likely" doesn't mean the stats are reflected by reality.
The unlikely happens - otherwise it would be "impossible". A 125 +/- 5
MYA date could actually be 140 MYA in calendar years (it's not likely,
but still possible).

-- 
________________________________________________________________

Dann Pigdon                   Australian Dinosaurs:
GIS / Archaeologist         http://www.geocities.com/dannsdinosaurs
Melbourne, Australia        http://www.alphalink.com.au/~dannj/
________________________________________________________________