Scientific fraud scandal in animal psychology

[LeftishBrit]

http://arstechnica.com/science/news/2010/08/harvard-professor-found-guilty-of-scientific-misconduct.ars

Bit of a shock; though some of Hauser's findings did seem too good to be true. A bit ironic that this is the author of 'Moral Minds' - there seem to have been one or two kinks in his moral mind.

[TZ]

1. Gee you don't think he had a political agenda do you?

:sarcasm: I've encountered this crap before unfortunately. The arrogance of being so sure you are right about something you will do anything, even commit fraud to show it. Science and political agendas (which it seems like this guy was up to his neck in) do NOT mix.

[Odin2005]

2. I'm not familiar with Hauser, what's his political agenda?

[TZ]

3. I picked it up from the article

He is trying to prove that animals think in human like patterns. I suspect some sort of animal rights agenda here. Specifically this..
" has been an outspoken advocate of the idea that animals possess many of the abilities that we think of as uniquely human."
I did a lot of study in animal behavior in college (ethology) and to me his research sounds very hinky---like he's starting not from the null hypothesis but from a presumption of his own that he is trying to prove. In simple terms.."I'm going to prove that animals are just like people but with fur". But thats just my feeling on this. Evolutionary psychology is all kind of...questionable to me.

[LeftishBrit]

4. I don't think there's a political agenda in the usual sense...

And in fact I think many of his viewpoints would go down quite well with most on this forum (morality doesn't depend on religion, but is part of the way that our brain evolved).

My guess is that it's more a matter of wanting to get 'exciting' results - both for the sake of fame and funding, and perhaps also just because he would *like* the exciting results to be true. People do get very excited at finding that babies or animals can do something previously assumed to be specific to human adults or older children.

I have come across cases of academic fraud, though not as severe or long-lasting as this; and usually the motivation seems to be more one of getting exciting results and proving one's own theories and advancing one's career and 'after all this is the result we *should* have got, so let's just say that we did!' rather than a political cause in the usual sense.

I do agree about being rather distrustful of evolutionary psychology as a whole: much of it seems to be Just So Stories, with little evidence base except for 'we do this, so it must have an evolutionary purpose!' And some of the more gung-ho evolutionary psychologists can have a lot in common with creationists: traditional human behaviour patterns and social roles are ordained, whether by God or evolution; and any attempt to change them through social changes or education is 'denying human nature' and likely to have bad consequences. Evolutionary psychologist Kanazawa even used the word 'evil' to describe feminism - just as a Christian Rightie creationist might.

But I think that Hauser's work was much more evidence-based than most evolutionary psychology. Just one small unfortunate matter: the evidence on which it was based turns out to be faked!

[TZ]

5. Maybe "political agenda" isn't quite the right word

Biased maybe? You know how people are supposed to design experiments with no particular outcome expected? I think that he has gone and tried to design experiments, that will get him the best results for his ideas- not necessarily the best test of a hypothesis, and when THAT didn't happen he played with the data to get what he wanted. Confirmation bias, I think its called.
When I've found dishonest research in the past, its almost always because the expected/wanted outcomes were not what the researcher wanted

[enki23]

6. that's not quite how one generally goes about designing experiments

You don't generally start from a "gosh, i have no idea what will happen" standpoint. If you'd ever tried to put together something like an NIH grant, even as an exercise, you'd realize that you sure as *hell* won't get much grant money that way.

Most often, from my minor perspective on the really real world, a typical scientist has to have a pretty good idea of what she expects to see, or at least wants to see, and then has to design the best possible experiment for being able to prove that right. In frequentist statistics terms, you don't just want a low alpha, you also want a low beta (high power). In english, you don't just want to minimize the false positives, you also have to minimize the false negatives (in most circumstances, anyway). If you don't, you don't really know whether you were wrong, or whether your design just wasn't good enough to tell you whether you were right or not. And if that's going to happen to you, you may as well not have done the work in the first place. And to design for that sort of power, you have to have a pretty good idea of what you'd like to see proved right.

[LeftishBrit]

7. True- I think that what TZ meant...

was not that you start with no expectations, but that you should not bias either your procedures or your interpretation in the direction of getting the results that you want. Indeed, you should ensure that your experiment is capable of falsifying your hypothesis.

This is one of my problems with the more gung-ho sorts of evolutionary psychology: the hypotheses are often as unfalsifiable as those of the creationists.

There is also a difference between *expecting* a result and *wanting* it. You may make predictions - you generally do - but this does not necessarily mean a strong desire to have those particular predictions confirmed.

There are two major reasons why people might have an excessively strong investment in a particular outcome. One is of course self-interest. A researcher might want results that confirm that a treatment or intervention, with which they have a connection, is working. (As someone who myself works in areas related to educational interventions, and is very interested in the development of medical treatments, I am highly aware of this!) More generally, 'exciting' results are more likely to be published, and in these days of publish-or-perish, researchers may wish to achieve such results! And younger researchers may have an interest in achieving results that confirm the theories of senior people who may employ or fund them.

At least, there is public and ideally individual awareness of such temptations, and this frequently, if sadly not always, results in safeguards being put into place. The second reason is that some researchers, like some people in any area of life, have a messianic sense that their ideas *must* be right. I have known at least one senior researcher who regarded anyone whose theories or interpretations were different from his own as an 'enemy' and seemed desperate to destroy their careers and credibility. Fortunately, science - unlike e.g. the altie-med industry - *does* have safeguards to reduce these dangers; but some people do get past them.