Articles to 2014-04-26

Zum Seitenende      Übersicht Artikel      Home & Impressum

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

While there certainly are isolated cases of the kind of problem Macilwain highlights, I fail to agree they are the norm. What he calls environmentalists advocat[ing] caution is all too often just the opposite, like the wholesale demolition of a secure energy supply in favour of unproven, unreliable and expensive alternatives, the kind of thing best represented by Lysenko, the great leap forward, and collective farming. In sum I can see nothing wrong with insisting on science being “experimental, empirical, quantifiable, and reproducible” with a little cautious conservatism thrown in.

While sounding quite reasonable at first, I fear Allen et al.’s suggestion may turn out as just one more wheeze whereby authors can wiggle out of responsibility whenever problems with an article come. I suggest every author has to take full responsibility for what is published under their name. Too many of them admit, when queried about glaringly obvious mistakes, to not even having read the final draft.

Simpson et al. is one more of those plausible-sounding hypotheses not supported by the data but published anyway. First off there is no indication whatever given of the relative sizes of the five groups. However looking at the overall standard errors of 1 % and 9 % for width and number of partners and the distance of outliers at 33 % and 520 % it becomes obvious the two extreme groups can each contain three subjects at most, more probably just one or two, with the rest making up the broad middle. In the results the standard errors for nearly all groups overlap completely. That’s the standard errors, mind, the estimated error for the group average, not the standard deviation, the only sensible quantity when you’re talking about the predictive value of a parameter for the individual. So what we have, again, is a result wholly dependant on one or two extreme outliers alone, with no correlation (let alone causation) for the 95 % of all subjects making up the bulk of the population.

Rozzi is an old article but I found it quoted second hand – from the Observer tabloid no less – in Fagan’s Cro-Magnon (list of 2014-04-18). Of course in the original the Observer’s conclusion, that Neanderthals were hunted and eaten by newly arriving AMHs, is just one possible interpretation among several more probable ones and later research (Bailey et al., J. Human Evolution 57 (2009), 11–26) even disputes the assignment of the mandible as Neanderthal. As Feynman said, never trust the experts and always check the primary sources yourself.

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!