10th October 2011, 01:18 PM
P Prentice Wrote:except that you have to take into consideration the sigmaC13 relative to VPDB .
and use the correct calibration curve.
P Prentice Wrote:and the fact that you might not be within the same 95% - given there is another 5% not reflected in the range
................
No that doesn't follow. Statistically there is a 95% chance that the 'true' date falls within the range in both cases.
NB (hint)....I'm not saying that the 'true' dates are the same......I'm saying the two results are the same. But that's the rub with measuring something scientifically..........its all in the errors and levels of confidence