Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

Applying Double Standards to ‘‘Divisive’’ Ideas

3 Post author: PlaidX 19 October 2009 12:36PM

This is a commentary by Linda Gottfredson on a paper by Hunt and Carlson about a paper by Richard Nisbett regarding studies done by Arthur Jensen. It's ultimately about race and intelligence, but it seemed meta enough to link to here.

Warning: PDF

Applying Double Standards to ‘‘Divisive’’ Ideas

Comments (5)

Comment author: Johnicholas 20 October 2009 05:49:22PM *  4 points [-]

The question addressed by this paper is: Is it appropriate to hold politically-sensitive research to a higher standard than other research? Despite appearances, it is a political paper, endorsing Jensen's research, condemning Jensen's critics, and its conclusion is that it is not appropriate to hold politically-sensitive research to a higher standard.

Inspired by the crypto-political nature of this paper, another way to phrase the general question addressed by the paper might be: How do we prevent scientific research from being contaminated by politics?

Suppose that we required all scientific papers to disclaim any political agendas (perhaps a standard disclaimer at the top). Then someone with a political agenda could gain the support of a scientific paper, simply by lying on the disclaimer.

Suppose that we required all papers to pass a "political conclusion" filter - editors systematically reject all papers that reach a conclusion that favors one political agenda over another. Then much valuable research would be suppressed - and someone with a political agenda could gain support of a political agenda by corrupting the decision of whether a given paper reaches a political conclusion.

Suppose that we allowed all papers, so long as the data passes certain statistical tests (e.g. p-value of 0.05). Then someone with a political agenda could gain the support of a scientific paper, simply by only submitting papers which support their conclusion.

This might be a signalling problem. Can we make it more costly to create political hay from fucking with the scientific process than any benefit the political actors might get from it?

Comment author: Emile 19 October 2009 03:21:06PM 1 point [-]

It may be "meta" in the sense that it's a paper about a paper about a paper, but a lot of it assumes more familiarity with the papers involved than you'll find here.

Comment author: SilasBarta 19 October 2009 03:37:44PM 5 points [-]

"Oh, hey Bob, I just saw a post on Less Wrong about a commentary on a paper about a paper regarding studies on race and intelli..."

universe implodes

Comment author: Technologos 20 October 2009 07:40:47AM 1 point [-]

And I just read a comment on a post on Less Wrong about...

Comment author: billswift 20 October 2009 01:00:09AM *  1 point [-]

I've only read a few of the specific papers mentioned, but I have read enough in the area to follow the article easily. And it is meta in its excellent discussion of using double standards to suppress research you don't want to think about.