First the link to this week’s complete list as HTML and as PDF.
***
I’m not sure Frenda et al. has any meaning at all. Let’s look at the context. All of the subjects were voluntary participants in a study and well aware of the fact. In all probability they were all WEIRD, i.e. well off, educated, aware of their rights, and not easily intimidated by authority. Even if one of them were to damage expensive equipment it would take evidence of malicious intent for the university to do anything about it. All of them hadn’t slept all night, were tired, and this was the last step before they would be able to go home and lie down. Now is there any plausible reason, why the more impulsive among them should not just say “what the heck” and be done with it?
As the authors admit themselves (if well hidden right in the middle of the least read part of the text), this scenario is very different from that of an accused in a crime case. Or it would be, if the accused had any reason to have confidence in due process. If he believed, they were out to get him anyway and signing or not would not make any difference, then the cases would indeed become comparable. But in that case there would be much more wrong with the system than could be fixed by a simple change in interrogation times.
***
Charbonneau et al., Devkota, Daugé et al., Villarino et al. and Zeevi et al. all make another strong case against the prevalent antibiotics abuse. But there’s more.
Traditional medicine has always set great store by dietary advice. Of course in the European case that was derived from fanciful, untested, and wrong Aristotelian hypotheses, but in Arabic and Oriental medicine it may well have been experiment- and evidence-based. These new results make a strong case for a second look.
***
Johansson and Mervis can be read as an appeal for the return to the old Ordinarienuniversität, where the senate of tenured department heads was the arbiter of al things in university life. Of course most scientists are not tenured professors and used to be completely subject to the whims of their department heads, who were not always the revered paternal figures of legend. But is it really better to be at the mercy of an anonymous, indifferent and largely unaccountable bureaucracy?
***
There is a deep and fundamental difference between measuring what is and modelling what might be. The latter may be quite valuable in hypothesis building but can’t add anything much to the testing of hypotheses. Unfortunately as McKinley and Ilyina demonstrate again, this distinction is increasingly getting lost.
***
Isn’t it funny how a headline can say the total opposite of an article’s content, if it is politically correct and follows the current reigning ideology as in the case of Hvidberg?
***
Perez et al. compare population development derived from mitochondrial DNA and from summed radiocarbon dates and again confirm the principal usefulness of the latter, albeit with large error ranges. Interestingly the best fit between the two is achieved by the simplest and oldest model for taphonomic loss. Their two favoured and newer, empirically fitted models come out noticeably worse. This should not come as a surprise. When a stochastic, variable and little constrained process is modeled with three parameters given with up to nine and at least four significant digits each, the result should and can be dismissed out of hand. (All three models cited are in my archive. For the newest, Williams 2012, see the list of 2012-02-24. The others predate the start of this blog.)