Please join us for a two-hour online course on
P-values and the normal assumption in scientific research
With presenter Professor Adrian Barnett.
About the course:
P-values are widely used in scientific papers, but are frequently misused and misunderstood. A p-value is not the probability that the null or alternative hypothesis is true, nor are they a useful decision rule for was is “significant” or not. We will cover what p-values mean and what they can be used for, and look at confidence intervals as an alternative to p-values. We will consider practical significance, also known as clinical significance or public health significance.
Many scientists wrongly believe that their data need to be normally distributed in order to use parametric tests like the t-test or linear regression. This leads scientists to use non-parametric tests (such as the Mann-Whitney test), which are useful but are limited in their inference and flexibility. We will learn why “lumpy” and not normally distributed data should be celebrated, and how the central limit theorem allows us to use parametric tests for most experiments with a sample size over 100.
P-values and the normal assumption are both widely misunderstood in scientific research. This short-course will improve your understanding of their use, which will help you produce better statistical analyses.
The course is aimed at researchers who use statistics, but who may not have formal training in statistics or completed their training some time ago. The course will not use any statistical software but will instead give practical examples for discussion. The aim of the course is not to show you how to calculate p-values, but how to understand them.
About the Presenter
Professor Adrian Barnett graduated from University College London with a BSc in Statistics in 1994. He worked for SmithKline Beecham and the Medical Research Council as a statistician before coming to Australia to do a PhD. He completed his PhD in Mathematics in 2002. His current research interests centre on improving the quality of health and medical research. He is the current president of the Statistical Society of Australia.
SSA Members: $50
General fee: $150
Membership is available from $240/pa for full membership. Find out more here.
Please register before 26 June 2020.
For more information please contact Marie-Louise Rankin.
Cancellations received prior to Friday, 19 June 2020 will be refunded, minus a $20 administration fee. From then onwards no part of the registration fee will be refunded.