Articles to 2016-04-06

Zum Seitenende      Übersicht Artikel      Home & Impressum

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


My short evaluation of Berkowitz 2015 (list of 2015-11-06) was based on the scant information from their article itself. Frank has generously taken on the task of reevaluating their raw data and given us his figure 2, which should have been in there from the start. What we see is a totally amorphous and widely spread data cloud with no discernible trend whatsoever. Of course mathematically, calculating a regression will always yield a result and the slope will never be exactly zero. And of course the more parameters you have, the better your curve will fit any random set of data. In their reply Berkowitz et al. brazenly employ a third order polynomial and base their whole argument on its extreme left quarter. In any polynomial fit the slope at the extremes outside and at the end of the data range is meaningless. If their fit were at all valid, how do they interpret the maximum at around one and the hiatus and slight fall up to three? That part at least is inside and (as much as possible) supported by the data.

They also completely ignore Frank’s most relevant comment. The number of uses per week is not random. If parents choose to use an intervention, that is expected to be helpful, often or nearly not at all, there must be a reason for that choice, and that reason and attitude alone is easily able to explain far more than their claimed near-zero effect.


Male students are judged 0.07 standard deviations more competent by their peers than female ones are while their attained grade average is 0.14 standard deviations higher. Contrary to what the title and the abstract claim Grunspan et al.’s figure 2 shows, that males overestimate males and females overestimate females and by nearly similar margins. So what is it in there, that makes this result worth reporting?

To put this into perspective (and assuming the result was as claimed): If I were in a group with a 56--58 % majority of engineers and the 43 % minority of, say, law students considered themselves smarter than us, would I go crying discrimination and be discouraged and tempted to give up and leave the course? Would I?


I’ll freely admit it, I too sometimes read trash. And so I happen to have read all that I came across concerning the Amanda Knox case in Italy. Having read Starr just now has opened my eyes. Anyone who still believes, he can get any more reliable information out of the yellow press – and when I use the term it includes The New York Times, Time Magazine, Die Welt, and FAZ – than out of Prekariatsfernsehen, i.e. the goggle box, is sadly mistaken. The term Lügenpresse is wrong too – to be able to lie one must make a meaningful statement in the first place – aimless blather devoid of all content does not qualify.

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!