Last year two highly-respected scientific journals were criticised for the poor quality of the statistics in many of their articles. Nature and the British Medical Journal (BMJ) were both found to have published articles in which the authors displayed a worrying lack of knowledge and care when presenting their results. The criticisms were originally made in Incongruence between test statistics and P values in medical papers, by Garcia-Berthou and Alcaraz from the University of Girona in Spain. The authors looked at the statistical results in all 181 papers published in Nature in 2001 and a random sample of 63 results from the BMJ from the same year—and their results were alarming.
展开▼