Thursday, November 27, 2014

learn social preferences from Taylor Swift

Yeah so I love Taylor Swift... and not just because her lyrics are great for learning about social preferences!

Social image:
Don't look at me,
You've got a girl at home,
And everybody knows that,
Everybody knows that.
 Pure altruism:
I don't even know her,
But I feel a responsibility,
To do what's upstanding and right,
Social norms:
It's kinda like a code, yeah,
And you've been getting closer and closer,
And crossing so many lines.
Guilt-aversion:
And it would be a fine proposition,
If I was a stupid girl,
Self-image:
But honey I am no-one's exception,
This I have previously learned.
Empathy or indirect reciprocity:
And yeah I might go with it,
If I hadn't once been just like her.
And a bonus lesson on commitment devices!
Call a cab,
Lose my number,
You're about to lose your girl,
Call a cab,
Lose my number,
Let's consider this lesson learned.

Saturday, November 22, 2014

publishing in different fields

This is simultaneously hilarious and awful. My immediate reaction is "Wow, this would never ever happen in economics." But I don't know - does it really happen so much in other fields? I certainly still get submission solicitations for journals or special issues of questionable quality, but I don't know anyone who doesn't ignore them across the board. This news article implies that people in other fields don't always ignore these solicitations, but how common is it really?

Maybe I'm googling the wrong phrases, but I can't find much solid information about publication practices by field. There are a couple papers specific to economics but I don't know of any good cross-field studies.

So I have to rely on anecdote for my opinions. In economics, we're obsessed with journal quality. Everyone is dying for a top-five article, and publishing in a crap journal is as bad or worse than not publishing at all. This might be intertwined with the fact that we're very far to quality end of the quality-quantity tradeoff (or at least, even crappy papers are very long and attempt to make a substantial contribution), and the other fact that there are very few authors per paper. I imagine this makes it much easier to scrutinize each paper when it comes to, e.g., tenure review, since there are only a few to dig into. Also, publication lags are extremely long for several reasons including ridiculously long delays in getting referee reports back after each submission, having to go through several cycles of revisions before final acceptance, and having papers rejected at a couple of journals before you even get started with that cycle of revisions. I read awhile back (sorry I don't remember the source) that the average time from project onset to publication is 6 years, and the first link above says that the average time a paper spends in the revisions cycle at the journal it will ultimately be published by (so presumably this doesn't include delays from journals that previously rejected it) is 2 years. Altogether, publishing in economics is downright nuts.

In the sciences and engineering, based on what I've gleaned from conversations with many friends in all kinds of fields from biology to mechanical engineering to astronomy, papers are much shorter and come in large numbers and published in so many different journals it would be impossible to keep track of their quality. Even conference proceedings are considered real publications. Sure there are the holy grail destinations like Nature, but in the meantime it's entirely acceptable to push out 20 papers in miscellaneous venues, each with 15 authors. It's correspondingly much easier to get negative results and replications and similarly individually-minor-but-very-important-in-aggregate results published.

Is there a magical field that is somewhere between these two extremes? Where papers are consistently significant works and held to high standard and throwaway publications are held in disdain to the extent that authors don't even bother publishing them, but in which the referee process is quick and requested revisions more reasonable (i.e. solely about ensuring rigorous results, not about catering to the reviewers opinions on how to frame the paper or what extensions/new treatments they'd be personally interested in seeing or whatever)? And in which there exists outlets for minor-but-sound results?