Friday, September 01, 2006

Brain melt

I promised to write more about the social responsibility of scientists. I'm doing it at the moment (can you imagine a better way to spend a Friday evening?) but two current stories in HS made my brain go to irritational overdrive.

I was writing about how academics can and should do more than just their share as individual citizens to help humanity combat climat change, social injustice etc., because they have the credibility and the platform to influence others, and thus they can carry more weight than most people (will post it tomorrow). This is another reason why all the zero/pseudo-science is "harmful", as it not only tarnishes the reputation of the individual wacky hacks making formulae for dunking biscuits, but it eats our common capital of credibility as well. Also, every self-evident research result that gets published during the silly season will have an effect on how people feel about their tax money being spent on research.

Journalists, working at the interface of science and society, have their share of responsibility as well. There is often real research behind the bleeding obvious results, and there are critical questions they could ask to separate hacks from the real things. Too often the journalists are either too bored to do their job properly or they just like the silly soundbite too much to interfere with the entertainment with high-brow questions. Some might even feel they are doing a service to the society by showing how tax money goes into waste when you give it to humanists, social scientists or in general, to academics, and showcasing inanity or simplifying findings beyond triviality is their version of "critical science journalism".

The two stories illustrate the point. First, the longevity of your marriage depends on how much you want to stay together. We knew that already, before Tuula Pukkala's PhD thesis about this was published, or the story about the research was released. But, this tautologous triviality is not the main point of the thesis (haven't read it, I'm guessing based on the article and the press release), as it actually deals with a whole host of factors and points of view to marriage. The idiotic headline is made by the journalist who wrote the story about it, or perhaps it was cooked up by the editor. Unfortunately, it will be attributed to the researcher who indeed identified motivation as one of three factors that contribute to long marriages.

The second story: Professor Andrew Oswald of Warwick University has recently published research on how the political opinions of parents seem to shift to left or right based on the sex of their children: having a girl will shift them towards labour, while having a baby boy will make the parents more likely to vote tories. We are talking about correlations between 2 % changes in reported likelihoods of voting behaviour and gender of babies. Not exactly solid as a rock as findings...

In the story, the journalist admits that since professor Oswald is friendly and humble, and doesn't seem at all like a chauvinist or provocateur, it is easy to believe that his results have a firm scientific basis. Ahh... give me a break. Is that really the analysis? Since when was the relevance of research determined on ad hominem -criteria? Did it ever occur to him to ask questions about how prof. Oswald defends his interpretations of causality when the results are simply correlational? Or any interest in discussing how significant (not just statistically) the 2% effect is, in actual terms? Any comments about the reliability of the self-report method used to gauge the political opinions? Or, whether the results are just specific for the UK or could they be generalised to other European countries or internationally?

I don't want to challenge prof. Oswald's numbers or his dexterity in quantitative methods. And, his explanation for the findings (in short/bluntly: girls need more "protection" and so parents think more about how the society will be like, while boys are more competitive) sounds relatively plausible, especially since the research was conducted in the UK, which is notoriously conservative in gender roles. And he clearly is a serious researcher who's done important work on for instance happiness (so I'm told). But this study reeks of soundbite science and should be shot down for closer inspection by the journalists, as they have done in Britain. And not only by the feminist ones who don't like Oswald's news, but also those who care about the quality of social science and responsible use of statistics in it, and those who want to put things in perspective.

HS, the biggest daily in Finland, the one with the best resources and most credibility, has failed again, twice in one day, to write anything sensible, critical or vaguely interesting about research when given good opportunities. I'm purposefully being cruel and stiff in the case of the Oswald-story, as it is perhaps not supposed to be about his work, but about the person, and as such it is decent (difficult to tell from the on-line version, as you don't see the context in which it is going to be published on paper). But, the readership would have deserved better in both occasions. And so would the researchers themselves. And while these stories will not hurt the credibility of academia and university research in general, or even the credibility of the paper, they don't really manage to serve as the interface between science and society either.

Rant over, I'm going home to watch some silly DVD, to complete the process of turning grey matter into fondue.

(Pic: "Brain Melt" by George Curington;

No comments: