![]() When the UK statistician Ronald Fisher introduced the P value in the 1920s, he did not mean it to be a definite decision basis. Statistical measures like ‘significant’ test results and P values always need interpretation, when one considers what they really mean: the chance of observing data under the assumption of a null hypothesis (of no correlation or no effect) therefore, they only reflect the likelihood that the null hypothesis is true. This element is the setting of limits for relevance, called ‘equivalence limits’ (European Commission 2013) or alternatively ‘limits of concern’ (EFSA Panel on Genetically Modified Organisms 2010). Therefore, another element has to enter the discussion if biological relevance is of prime importance, as it is for decision-making in risk management. Statistical significance is determined by the precision of the measurements, and as such is not connected to the biological relevance of observed differences. ![]() Denoting something as statistically significant does not mean it is biologically relevant. Statistical analysis is a (undoubtedly very useful) tool for extracting information from data and helping scientists blend data and background knowledge to derive scientific conclusions-no more and no less. Most importantly, biological relevance should always be preferred over statistical significance in any evidence-based decision-making. hypothesis testing, P value), which simply asks ‘Is there an effect?’, while other more recently published papers promote the reporting of effect sizes and confidence intervals and to ask ‘How much of an effect is there?’ (Ellis 2010 Nuzzo 2014). Most of the guidelines favour a traditional approach (i.e. 116 mentions that there is no single approach to the statistical analysis of data and that statistical methods continue to develop so that new and modified approaches may continue to be proposed (OECD Environment, Health and Safety Publications 2012). Anses 2011 EFSA Scientific Committee 2011 Festing and Altman 2002 OECD Environment, Health and Safety Publications 2012). There are several guidelines and publications dealing with the statistical treatment of toxicity study data (e.g. Statistical significance and biological relevance We compare the traditional ANOVA approach with a more modern LMM approach, and we investigate the use of standardized effect sizes as proposed by EFSA ( 2011). In this paper, we describe the statistical methods used for analysing the data from the GRACE 90-day studies (Zeljenková et al. Several observation and examination data are recorded and compared between the treatment and control groups. ![]() The idea is to administer diets containing the plant under study as a component: in treatment groups, this component consists of GM plant material (high and low doses), and in a control group this component consists of conventional plant material. Although there is a fundamental difference (dosing range) between testing chemicals and whole food/feed, repeated-dose 90-day oral toxicity studies nevertheless have been included in the integrated approach of assessing the potential toxicity of GM plants (EFSA Scientific Committee 2011). Toxicity studies are now a mandatory part of the risk assessment of genetically modified (GM) food and feed in Europe. This general OECD test approach has been applied to the testing of whole food/feed derived from genetically modified organisms (GMOs) in order to consider toxic effects holistically rather than for a single compound. At least three dose levels of a test substance and a concurrent control are administered daily per os for a period of 90 days to groups of animals (OECD/OCDE 2014). In this context, repeated-dose 90-day oral (subchronic) toxicity studies are usually carried out to evaluate the toxic potential of a chemical in more detail after initial information on its toxicity has been obtained from acute or repeated-dose 28-day toxicity tests. OECD has developed standard procedures employing animal models to assess the toxicity of chemical compounds to humans. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |