Zum Seitenende Übersicht Artikel Home & Impressum
First the link to this week’s complete list as HTML and as PDF.
Jun et al. make excessive use of the word fact checking in their report and employ it exclusively to describe one of their three outcomes. In fact they do not test for anything of the sort. In all eight experiments their participants can give one of three answers, true, false, and don’t know. The fact, that they will be told the correct answers for their don’t knows afterwards, is irrelevant, as they will not be able to correct their answers then. These participants all take part in a boring task for money and probably could not care less about the truth of some fictitious and concocted statements. They are far too old to be college students, but as members of MTurk they probably all belong in the WEIRD category. As such they will be aware that on average
“don’t know” is equivalent to throwing a coin and that giving a random answer will save them the tedium of receiving and having to pretend to read additional material at the end. So what’s probably really happening here is, they set out being conscientious but, being made aware of the fact of only being one in a hundred, prefer the easier way out in the knowledge it probably won’t skew the result anyway.
So what would an experiment truly being about fact checking look like? I should give one point for correct and deduct one for false as above, but provide some means of real fact checking, like consulting other sources, and award only half a point for a correct answer given after that. This, after all, is what real fact checking is all about.
Only one question remains: What on earth possesses the National Academy of Science to publish stuff like that?
As far as I can see there still is no emerging consensus about where and whom Homo floresiensis may have come from. Until there is I shall continue to take all published results, like this new one by Argue et al., with a sizeable pinch of salt. The purpose of any theory is to make testable predictions. In the case of human genealogies this means describing in advance, what the next fossil found will look like given its age and provenance. So far every new find has come as a total surprise and yielded something entirely new and unexpected. As long as this continues to be the case, the suggested human family tree is still very far from being able to be called a theory.
Changes in the spine and vertebrae are intimately related to our upright gait, making Ward et al.’s results particularly relevant.
Dangendorf et al. is another example of tweaking and correcting measured data in such a way, that they a) better conform to published models and are b) moulded towards a hockey-stick-like doomsday shape. Of course neither point is proof against the validity of these corrections, but the consistency, with which all recent changes of older measurements point in a single direction, certainly raises a question.
Styring et al.’s is a very worth-while study on early agriculture and seems to confirm what has long been assumed about Neolithic subsistence systems. That said their results look particularly weak. They strongly depend on assumptions about contemporary precipitation levels of which we have crude estimates only for the relevant times. On top of that their figure 4 shows only a weak trend, possibly caused by confounders and outliers alone, that seems to explain something like a tenth of the variation if that. This seems a tiny peg to hang a hefty theory on.
Zum Anfang Übersicht Artikel Home & Impressum