• Some thoughts on controversial GM toxicity study


    The recent study by Prof. Seralini, which suggested that a genetically modified (GM) maize causes severe disease in rats has generated much discussion. A summary of the situation can be found in this Nature News note.

    The more detailed criticism related to the quality of statistical inferences, the size of the sample, etc., is well exemplified here and here.

    Now, Seralini has published a tribune in Le Monde (in French), which makes a number of fair points, in particular about the asymmetry in the level of exigency applied to his work (sample sizes, availability of data) versus that directed to previous studies used to authorize the maize.

    This certainly does not justify potential statistical or methodological shortcomings of the Seralini study, but one must recognize that studies conducted previously, which led to the authorization of the maize NK603, did not generate such a controversy and mobilization of scientific scrutiny, despite a number of similar shortcomings, not to mention potential conflicts of interest of people involved in them.

    The “politics” of the case gives the impression that we, as a society (as scientists?), seem to be more preoccupied by what economists would call “type I errors” (here, rejecting by mistake the hypothesis that the maize is not toxic given that it indeed is not, i.e. classifying as toxic by mistake, and therefore not authorizing an innocuous product) than by “type II errors” (here, failing by mistake to reject the hypothesis that the maize is not toxic given that it is, i.e. classifying as non toxic by mistake, and therefore authorizing a dangerous product). If that’s true, it certainly says something of the institutional framework (I am thinking broadly of the relationships between businesses, governments, scientists and scientific authorities, and the media) surrounding this issue, which is well worth analyzing.

    Leave a comment | Share on Facebook | Share on Twitter

  • last posts

  • Archives

  • Frontières

  • links

  • Paraguay


    © Design graphique et développement : www.mrcam.fr