Posted By:Agatha
|
Quote:
Originally Posted by Vixen (Post 10754646)
Not analogous at all. The correct analogy is that 200 labs up to ISO standard be randomly allocated samples from various different areas, with a further number of labs having a dummy sample, with none of the labs knowing which type they received, together with a control sample which looks identical to the original.
Nowhere did I say one lab should test all of it x 200 times. (200 is a suggested figure to make statistical analysis robust.) But then you knew what I meant. |
What you are asking is equivalent of taking 200 blood probe to test for diabetes.
But nobody does that because the sample is uniform. You take one sample, at most 2.
The same was done here : the cloth construction is as uniform as it can be so the sample area is enough and 3 labs is overkill.
via International Skeptics Forum http://ift.tt/1NVajx3
Aucun commentaire:
Enregistrer un commentaire