Why is the chi-square value appropriate in this case.
Hi AlNg,
the Chi-Squared Test does not concern itself with environmental conditions. It is a merely mathematical way of distinguishing (if you don’t already know) whether two groups of measurements are actually measuring the same thing or not. For instance, as an athletics teacher, I was once involved in some research to find out whether boys are faster than girls. (It is recognised that they are from aged 12 or so, but we wanted to find out if the difference was apparent earlier). We timed about fifty six year olds, and of course there were some fast girls and some slow boys, and the measurements overlapped somewhat. However, following the procedure of the Chi-Squared Test, we ended with a value that demonstrated that there was only a 5% chance that the two groups of measurements came from the same population, concluding that even at aged six, on average boys are faster than girls, so girls should be given separate races on sports days!
The statisticians of the Nature paper were confused by the Chi-Squared Test that they carried out on the Shroud, as it showed that there was only a 4% chance that the Oxford sample and the Arizona sample come from the same cloth. As they had seen the two samples being cut, they knew that they were from the same cloth, so the Chi-Square needed to be explained. The explanation they came up with was that the errors had been under-reported, and were greater than quoted. This seems a little paradoxical, but curiously, the more imprecise a couple of measurements, the more likely they are to be measurements of the same thing. After all, if I say a piece of string is 15cm long and you say it is 20cm, then one or both of us must be wrong, but if I say 15cm give-or-take 3cm, and you say 20cm give-or-take 3cm, then we could both be right, as 17cm and 18cm lie within both our measurements.
Since then, alternative explanations have been suggested, such as minor contamination, substantial reweaving, or radiocarbon enrichment.
What the 1989 NATURE report did not bother to tell us is that the C-14 data obtained from the medieval control samples was not disparate and does pass the robust statistical analysis tests.
Not at all. “An initial inspection of Table 2 shows that the agreement among the three laboratories for samples 2, 3 and 4 is exceptionally good,” says the Nature paper.
Prior to the publication of the NATURE article Prof. Phillips of Harvard University wrote a letter to the British Museum…
I have no evidence for this. He did write to Nature after the results had been made public, and his letter appeared in the same edition as the paper on the tests. I have no evidence that the British Museum responded either, although Robert Hedges of the Oxford laboratory did, and explained that miraculous explanations are not susceptible to scientific study. “If a supernatural explanation is to be proposed, it seems pointless to make any scientific measurement on the shroud at all.” I agree with him.
