Articles to 2017-07-15

Zum Seitenende      Übersicht Artikel      Home & Impressum

First the link to this week’s complete list as HTML and as PDF.


The main result from Chao seems sound, although the fact that this knowledge will be utilized by commercial fund raisers, who typically syphon off three years worth of contributions before the intended recipient begins to see his first penny, and who abuse their employees’ enthusiasm for the good cause by paying them a pittance leaves a bad taste.

But their supplementary online experiment makes me wonder, what kind of delusional world social scientists live in. A gift voucher for one cent is derisory and insulting and will cancel any goodwill I may have felt beforehand. Also making use of a voucher typically involves typing or pasting a lengthy number code. At one cent you will need to cash one thousand vouchers in a single hour even to reach the minimum wage. Even if this feat were possible, it certainly is very hard work and it comes on top of the supposed free gift. And this is supposed to mean anything?


In Choi et al.’s figure 3 the late adopted subgroup counterintuitively and contrary to their claims do not do any better than the Dutch controls. If age at test is the main driver for the difference between early and late adoptees, as the authors claim, than the Dutch control group according to their table S1 is far too old to provide a meaningful control. In sum their whole claim hinges on the very slight deviation of the late adopted midpoint from the straight line, a deviation well inside the point’s error range. This badly designed non-result should have been rejected even by a mediocre journal.

PNAS are fast losing all the credit I used to be prepared to accord them. With it being the most prestigious scientific journal world wide, what are we left with? Is the age of science and reason finally over now?


I am not sure, whether Margaryan et al. means anything or not. An admixture of, say, five percent of immigrants will leave a five percent signature in each of the full genomes of typical descendants a few generations later and can be detected. Mitochondrial genomes on the other hand follow a single straight line of descent. You need twenty genomes on average to find a single sign of a five percent admixture. Here we have 52 samples for the full 8 ka, typically less than six per relevant time slice. Seeing that even a total 100 % population replacement will still typically retain some common haplogroups, what is this small sample supposed to prove?

Zum Anfang      Übersicht Artikel      Home & Impressum

Creative Commons Attribution-Share Alike 3.0 Unported License Viewable With Any Browser Valid HTML 4.01! Valid CSS!