Marsh et al. are a particularly strong example for the meaninglessness of statistical significance. While their figure 2b clearly shows a significant difference at the group level, the two groups become completely identical by just taking away two individuals from each, and the predictive value Read the rest of this entry »
Successful teachers have always scattered witty or engaging anecdotes throughout their more serious stuff. Gruber et al. now show why this is so effective.
Currently the press is full about stretching before sport being not only not beneficial but downright harmful. Read the rest of this entry »
The result by Franco et al. is unexpected and perhaps disturbing. Publication bias is not a result of journals rejecting null results but of them not being written up and offered in the first place. Understandably having come up with a seemingly plausible hypothesis one will always feel better having it confirmed than Read the rest of this entry »
Dear Sirs, dear Dr. Leshner,
due to the way Science Magazine is organized, a way which until now I have fully approved with, I am not just a paying subscriber but also a full voting member of the AAAS. Now while a customer may be free to ignore his supplier’s interior affairs Read the rest of this entry »
Looking at Boëda et al. I see several caveats. All the supposed tools were found near the tops of gravel layers of similar grain sizes. During excavation the fine grained layers are carefully scraped away, so it is the top of any gravel layer that emerges prominently and catches the eye. Lower down the need to remove stones leaves an irregular surface, where Read the rest of this entry »
By Moshe Feiglin
On Friday I went to pay a shiva call (comfort the mourners) to the parents, brothers, sister and especially – the pregnant widow and orphans of Netanel Arami, may G-d avenge his blood. Netanel plunged to his death on Tuesday when somebody cut the rope from which he was rappelling as part of his work at a construction site. Read the rest of this entry »
Sher et al. use LOWESS to draw regression lines through extremely noisy data. They then go on to interpret those in far more detail than the data warrant. Their single limited sample makes it impossible to distinguish between true age-related changes and random, individual idiosyncrasies. I accept all their first-order results (none of which are new or surprising), but Read the rest of this entry »
Als sarkastisch-rhetorische Fragen, die nicht wirklich auf eine sachliche Antwort abzielten, wurden diese drei von der Moderation bei abgeordnetenwatch.de zu recht abgelehnt. In einer Zeit, in der Parteienverbote die Unfähigkeit zu inhaltlicher Auseinandersetzung kaschieren sollen, muß aber wenigstens hier auf Verfassungsfeinde und das Recht brechende Wiederholungstäter in hohen Staatsämtern hingewiesen werden.