Also this week: Andrew Gelman vs. the Wilcoxon signed ranks test, PLUTO, and more.
The Pluto Flyby happened this week, which made it an extra fun week to be on twitter. This BBC piece helps put things in perspective, as does this video from the NY Times. It’s mind-boggling that it went from a massively pixelated blob to this gorgeous image. One of my favorite images of the week has been this one, showing the reaction of NASA scientists to that image linked to in the previous sentence. The expressions on their faces are amazing! We’re also learning lots more about Pluto’s moons, including its biggest moon Charon. Overall, it’s been amazing to follow along. My daughter has been totally captivated, too. This article also has a good summary and a great slide show at the bottom showing images of Pluto through time.
The city of Melbourne, Australia, gave email addresses to trees so that people could report problems such as hanging branches. Instead, people are using the email addresses for all sorts of other purposes, and sometimes the trees reply. Wonderful!🙂 (ht: @catherineq)
Andrew Gelman says you shouldn’t do the Wilcoxon test. Interestingly, he also doesn’t like what I’d think would be the most obvious alternative, a randomization test. The post and comments have an interesting discussion of the pluses and minuses of different alternatives here.
Sticking with Gelman, he passes on the news that there really is a “hot hand” in sports. Of interest to me because I use the example of the hot hand in introductory biostatistics. I may need to revise my lectures in light of the latest results. But first I need to make sure I’ve fully grasped the very interesting-sounding little probability paradox at the core of the latest research on this.
Gelman also passes on news that economists, like psychologists, can now bet real money on whether various experimental studies published in leading economics journals will replicate. Sounds fun and potentially informative. I’d think it would be reassuring if, on the whole, people working in a field are able to predict which results will replicate. In psychology, a commenter on Gelman’s blog reports making money by betting against replicability of small sample studies, counterintuitive studies, and social psychology. A betting market also seems like an improvement over one-on-one bets between opponents, since it can be hard for opponents to agree terms.
Princeton economist Uwe Reinhardt with the greatest syllabus ever. It begins as follows:
After the near‐collapse of the world’s financial system has shown that we economists really do not know how the world works, I am much too embarrassed to teach economics anymore, which I have done for many years. I will teach Modern Korean Drama instead.
Although I have never been to Korea, I have watched Korean drama on a daily basis for over six years now. Therefore I can justly consider myself an expert in that subject.
Click through for the notes from the first lecture. I’m still trying to decide whether it’s completely a joke, or whether he’s going to use Korean drama as a fun way to teach economics.
And finally, an impending increase in realized fitness.🙂