@flutherother Or, to be even more fair, we could include Seralini’s perspective:
“inconclusiveness of findings is not a valid ground for retraction. Numerous published scientific papers contain inconclusive findings, which are often mixed in with findings that can be presented with more certainty. It is for future researchers to build on the findings and refine scientific understanding of any uncertainties.
It is important that scientists do not overstate their findings or draw conclusions that are not justified by the data, but Prof Séralini’s paper does not do this. Because Prof Séralini’s study was a chronic toxicity study and not a full-scale carcinogenicity study, which normally requires larger numbers of rats, he conservatively did not do a statistical analysis of the tumours and mortality findings. Instead he simply reported them, without drawing definitive conclusions. This is in line with the OECD chronic toxicity protocol, which requires that any “lesions” (including tumours) observed are recorded.[4]
The criticisms of the low number of rats and choice of rat strain have been addressed by Prof Séralini’s team in a comprehensive response to critics that was published in FCT,[5] as well as by independent scientists writing in support of the study.[6]
Experts in statistics writing in support of the study have pointed out that large numbers of animals are only required in safety studies to avoid false negative error, where a toxic effect exists but is missed because too few animals are used. In the case of Séralini’s study, this was not an issue. The toxic effects of the test substances were so pronounced (there was a “large effect size”) that smaller numbers of animals were sufficient for statistical significance.[7,8,9]”
http://www.gmoseralini.org/journal-retraction-of-seralini-study-is-illicit-unscientific-and-unethical/