Thursday, January 10, 2013

Bad Data; an addendum

In my last post on Bad Data I mentioned that exactly one published paper by Hauser (the 2002 Cognition paper) was retracted and that this struck me as pretty slim pickings and not worth all the fuss generated. A kind correspondent pointed out to me that the results of this paper were recently replicated and is coming out. The paper - Chasing Sounds-  is by Julie Neiworth and using a slightly different methodology replicates the earlier findings.  By my count this means that for all of the sturm and drang the forward march of science was impeded not one whit. Indeed, given the last replication, one might conclude that withdrawing the paper has been a greater impediment than publishing it was.

Let me end with one personal remark. There are many scientific vices. Among these are virtuous self righteousness and envy.  Few things are as gratifying as bringing down a high flyer for malfeasance. See, s/he got there by fraud! We can ignore what s/he said (thank god as it gores my ox!).  Of the scientific vices these have often done more harm than deliberate fraud, let alone sloppiness. Moreover, they are much harder to police given how easy they are to dress up as virtue. At any rate, it seems that Hauser is three for three and that fuss in that teapot was more like a mild summer shower rather than a tempest.  Personally I hope that Hauser soon comes back from exile and starts psychologizing.


  1. This is an interesting take on the Hauser affair. One question. You write "Of the scientific vices [irtuous self righteousness and envy] have often done more harm than deliberate fraud, let alone sloppiness." Can you give a couple of examples where that has happened? Thanks

  2. Good question. I have personally been involved in some cases, but mainly at the reviewing process. Some papers seem subject to far more stringent requirements than others. This can often be attributed, in my view, to the views expressed and by the people expressing them. A decent example where very good work was nibbled to death is discussed in the Roediger and Arnold paper cited in 'Does anyone ever learn anything.' These subtle forms often block excellent ideas from emerging and being reasonably considered.

    Unpopular ideas face a very uphill battle. But it is impossible to simply say that there ideas are disliked. Rather, they fail to meet the requisite standards is the theme of choice. That's what I intended.