First the link to this week’s complete list as HTML and as PDF.
Halberda is another example, of how a statistically significant result is not at odds with it being completely meaningless. In his animated GIFs the ranges for the different mathematical aptitudes overlap nearly totally, the predictive value of his result is essentially nil.
Zorzi on the other hand is helpful, relevant, Read the rest of this entry »
First the link to this week’s complete list as HTML and as PDF.
Simonsohn makes a pertinent point about scientific fraud. Journals should admit their responsibility and do something about it instead of falsely framing themselves as the innocent victims.
All secondary reports, Schoeninger included, try to frame Henry’s result in terms of human ancestors. I can’t see why. Read the rest of this entry »
First the link to this week’s complete list as HTML and as PDF.
Eisenberg is a prime example of pure chartmanship. The only diagrams he shows are of the effect he does not discuss (and following the rule “if you caint see it, it aint there” these are meaningless) and his reported R2 is consistently below 0.05, i.e. pure noise. His largest possible effect in figure 2 is barely one fifth of the densely packed inner area of the scatter in figure 1 and neither variance nor standard error are given. Read the rest of this entry »
First the link to this week’s complete list as HTML and as PDF.
Witztum et al. is the primary source behind the Bible Code. As far as I can tell it has since been refuted by reproducing similar results from other works of literature. Rips, the main author, is a distinguished mathematician but no statistician. The referees chosen by Statistical Science were though, and they’d have preferred to be able to reject the manuscript but could not find a reason to do so. Read the rest of this entry »
First the link to this week’s complete list as HTML and as PDF.
There are two new articles on antibiotic abuse. Negishi confirms the rationale behind giving antibiotics for viral infections. Viral infection does indeed suppress the immune reaction against bacteria. By describing the mechanism Negishi et al. also open the way towards a more sensible treatment. Abt et al. demonstrate yet another downside of antibiotics, killing beneficial gut flora also impairs the antiviral immune reaction. Read the rest of this entry »
First the link to this week’s complete list as HTML and as PDF.
While seemingly confirming theory Fujita et al.’s result is decidedly fishy.
As a well-off mother experiences less conflict between offspring benefit and cost to herself and as raising nutritional value beyond the optimum confers no benefit, we expect the sex difference to be low at high SES and high at low SES, contrary to the reported result.
Read the rest of this entry »
First the link to this week’s complete list as HTML and as PDF.
Lager & Torssander report another spurious non-result from the humanities posing as science. In their table 3 they test for 42 outcomes and from the text another 84 (age under 40 and whole group) are untabulated. Out of these 126 tests 8, i.e. 1 in 16 come out as barely significant, meaning with a probability of 1 in 20 for occurring purely by chance. So what exactly is it they are reporting here and what did the referees do in their peer-review?
First the link to this week’s complete list as HTML and as PDF.
Concerning the various papers by McIntyre & McKitrick and McShane & Wyner I note:
– In no way do they invalidate climate reconstruction from proxies as such, their issues are solely with methods used in aggregating a huge number of proxies.
– As the models are trained on a short time span exhibiting two steep rises, Read the rest of this entry »
First the link to this week’s complete list as HTML and as PDF.
Berna together with Carmody 2011 provide new support for Wrangham’s disputed fire hypothesis.
Yong has another take on the decline and fall of the empire of science Read the rest of this entry »
Sehr geehrter Herr Kompa,
ich unterstütze Ihren Kampf um die Meinungsfreiheit und habe mich Ihrem Aufruf folgend an den Kosten beteiligt. Dennoch erscheint mir dieser Prozess irrelevant und an zwei Stellen an den eigentlichen Problemen vorbeigehend.
