Lots of items kicking around in the blogosphere that deserve more attention than I have time to give them right now. (I'm off to start taking soccer classes in about an hour -- hold a good thought for my knees, please!) But I wanted to share.
- At Log base 2, Nick Barrowman considers the epistemic (and perhaps ethical) consequences of missing values in statistics. He writes:
Missing values are a bit of a dirty secret in science. Because they are rarely mentioned in science education, it's not surprising that they are often overlooked in practice. This is terribly damaging--regardless of whether it's due to ignorance, dishonesty, or wishful thinking.
Read the whole thing and learn how statistical and experimental methods work together in generating what we know -- or don't.
- At nanopolitan, Abi looks at how worked up the editors of Nature are getting at the idea that researchers in South Korea could be offered cash incentives for academic publications. Abi points out that there are already extrinsic rewards for such publications:
[I]t's important to keep in mind that publications in high impact journals lead to many other rewards: better chance of funding for your future projects, awards and recognition (some of which come with cash prizes), and early promotion (with its built-in cash incentive). Thus, an explicit award for each publication just happens to extend this trend.
As well, he detects a bit of sanctimony toward the scientific culture of South Korea in the wake of the stem cell scandal. While I think I may have more worries about what cash incentives do to the ethical environment of a scientific community than Abi here, I very much agree with his suggestion that if there's a problem, it's not just a South Korean problem -- it's one in which even the editors of Nature are up to their necks.
- Via Pandagon: If a man passes a "domestic violence propensity test" with "flying colors", and then he goes on to stab his wife to death, is it reasonable to conclude there might be something wrong with the test?
- From three years ago on Kieran Healy's weblog (people, I have a lot to read), this beautiful post on the continuing battle over whether (and which) social sciences are actually scientific. A taste:
Informed answers to this question are rare. Instead, you tend to get half-baked ideas about predicability and falsifiability as the criteria for science being put into service shoring up one's allegiance to a chosen tribe. If predictability looks like a shaky foundation (What? You're telling me bright young economists don't get hired on the basis of successful predictions?) the ground can be shifted to the existence of "basic shared premises within the field." If these shared premises begin to look a bit too metaphysical, then we'll move to a different criterion. Whatever it takes to preserve the phenomena. It's easy and fun. All you have to do sacrifice is your consistency.
If you don't read the rest of the post, you're living a lie.
More content soon, provided I don't end up in one of those tragic head-on collisions with a soccer ball!