ABSTRACT
The word ‘statistics’ has two general meanings: the discipline of the study of data, and as the plural of ‘statistic’ – a measurement derived from data. Virtually all numerical results published in journals come with some estimate of variability attached. Notwithstanding the idea that ‘honest’ measurement might determine that some of such accuracy is spurious, this can only be a good thing. One of the mysteries of reporting measurement accuracy is what it means. If postgraduates take a large’ sample from a population, and calculate the mean, this sample comes from a normal distribution. This is true regardless of the population. Any population, any sample and the mean comes from a normal distribution. In terms of analysis of data, transformations and scaling have to be considered more carefully. In particular, the transformation has to match the requirements of the analysis.
