= 27d567bc2d48d9f40e9ccc20ea370b15)
paper-machine comments on Open Thread, October 1-15, 2012 - Less Wrong Discussion
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (477)
The drama caused in other sciences due to not sorting authors alphabetically is truly depressing.
Economics also has the tradition of ordering authors alphabetically. And economists with earlier-letter surnames end up having more successful careers, quite possibly as a result of that tradition.
And yet the effect isn't strong enough for economists named Zweibel to change their names to Aardman?
If the tradeoff is between either some acknowledged bias or a great deal of drama, I choose the former.
There are possibilities for less biased orderings which are also less drama-prone. For example, choose a day 1, then use alphabetical order, but advance the first letter of the alphabet for each a day after day 1. Today is the alphabet for author ordering starts with A, tomorrow with B, and so on. If that still introduces a bias, then perhaps the alphabet should start with the same letter on consecutive days, but alternate between going forwards and backwards.
If you want an more-or-less unbiased but deterministic way to do this, you could sort the authors by whose birthday is closest (in either direction) at time of publishing. This additionally makes it so the precise date doesn't matter too much. Making it closest upcoming birthday would be simpler, but if a colleague's birthday is one day before you it kind of sucks.
But probably the random-order idea, as suggested by the article, would be even easier.
The reason I wanted something deterministic is that I wasn't convinced that scientists would generally trust something that looked random with the stakes being somewhat high. When I think about the amount of scientific fraud, I'm not even sure that they should trust each other.
My alphabetical scheme isn't ideally random-- it gives an advantage to authors whose names begin with unusual letters.
Thanks for the information that date of publication can be somewhat foggy.
In the article Hanson proposes a simpler method.
In some sciences there are established rules for the role of the principal investigator (first author) and the last author. In the life sciences at least it's often enough to be the head of a laboratory whose equipment was used to get on a publication, no additional input required.
Why would you want to hide that information, or does your proposal encompass only all the middle authors / "contributors" (often well over a dozen)?
Mu; I don't want to hide that information. The cultural traditions (calling them "established rules" is a stretch when there's so little agreement about what they are) surrounding authorship hides them in a lossy format while causing at least thousands of scientist-hours of drama a year. If authors want to explicitly flag down who did what, they can do it somewhere in the article proper.
A possible counter-argument would be that employers want to see, e.g., how many papers a given research has been the PI on, without reading the article proper. The current system already doesn't work for that use case; the status fighting over first authorship has created the ridiculous notion of "co-first-authorship" and false declarations that "Dr. X and Dr. Y contributed equally to this project."
The best part?