Assorted links

Comments

Re: No.1 It's the first rational thing EU regulators have done in years!

From the article about traffic fatalities:

"The Michigan study found a nearly 20% decline in deaths among young drivers, age 16 to 25. Among the possible reasons: the increasing number of states that use graduated licensing programs that delay granting full driving privileges until teens have more experience, and rising teen joblessness."

Umm.. It seems to me like there's a big difference a 25 year old and a 16 year old in terms of driver safety. Since the population as a whole is aging it seems like the biggest contributor to this is that the composition of this age group is getting older.

Great points, Adam.

Adam--

I took part in a class taught by one of the researchers earlier this year. Throughout the semester she had us perform various puzzles/projects with our groups. During this time, relative group performance (groups were the exact same the entire class) remained relatively constant throughout the semester, which suggests that empathy and communication skills were not just valuable in the short-term, but beyond as well, even during times when group members were quite familiar with each other (and the "emotional iq" was a very good predictor, which for us was measured by guessing the emotions of a person by looking at a picture of their eyes).

I doubt this was confirmational bias either--this was an MBA course at a analyticaly focused school--not an audience sympathetic to pc claims/ideas.

Peer Review: As in all things, it depends. Some referees are knowledgeable and diligent. Others are not. Some give fabulous feedback while others criticize omissions which were directly addressed in the paper.

Blind peer review removes reputational effects. Case-Quigley-Shiller would get a C in an Econometrics class if the authors weren't so well know and respected.

Many Ivy League grad students get published in top journals edited by their advisors when those same papers would be rejected if published by anyone from a second or third tier school.

Many papers are accepted only because they jibe with the viewpoint of the referees or editor while contrarian views are rejected. Research opposing anthropogenic global warming or denying a connection between dietary cholesterol and blood cholesterol have been systematically and deliberately suppressed in peer review and grant committees for years.

On the other hand, some papers (e.g. Card and Krueger) were published because they had novel results contradicting theory, despite being crappy research.

Blind review would probably be the best improvement, but any qualified referee will probably figure out who wrote it anyway.

I don't believe the peer review process is thoroughly broken as much as the "publish or perish" attitude is broken. It leads to loads of crappy and useless research and poor teaching. Many of the greatest scholars produced one magnum opus in an entire career.

Then you have scholarly lightweights like Yellen and Kagan rising to the top of their profession mainly through administrative roles garnered through connections. Some older, well known academics wouldn't get tenure if their publication record went before the committee today. Some people get tenure only because of their political activism, an abundance of melanin, or the lack of a penis.

Some academic disciplines are far worse than others when it comes to determining what "good" research is. The harder the science, the more difficult to pass the trash.

With the internet reducing the marginal cost of publication and searching to near zero, why not just publish EVERYTHING and let the market decide what is useful and important by citations, comments, and peer accolades from the entire community. It would make every paper just like an invited conference piece with discussants.

4. I wonder what would happen with groups versus individuals on many different tasks, particularly the individuals that did poorly on the tasks that the groups did very well in, perhaps the tasks with the largest variation in group performance.

2. "Studies that report a high level of IRR are to be considered less credible than those with a low level of IRR."

So, if reviewers agree, they are wrong. If they disagree, that means they are right. Okay, I understand now.

"this was an MBA course at a analyticaly focused school--not an audience sympathetic to pc claims/ideas."

Ha! Yeah, no way status climbing elites would want to distinguish themselves from the bigotry of the unwashed. Ha!

Higher gas prices might disproportionately affect 16-year-olds over 25-year-olds.

As Six Ounces said, it seemed odd to me, especially for the WSJ, that they talked about total numbers of deaths rather than deaths per mile traveled.

Rates are what are important, not totals, for something like that...

I would give some credit to urbanization. Rural drivers cover long distances at high speed. The death rate among teenage, rural drivers is truly terrifying. Even with the same skills and habits, young drivers in suburbs and cities are moving more slowly. Low-speed crashes are rarely fatal.

Comments for this post are closed