A grasp of statistics is necessary for undergraduate and post-graduate students in many subject areas. Statistics teaching is thus a generic issue and not something that is only the concern of mathematicians. Unfortunately, statistics is often seen as a daunting subject and ‘mathematics anxiety’ occurs commonly among students in all subject areas except Mathematics itself.

A statistics teaching method (lectures plus handouts) explaining major statistical concepts, approaches and methods using almost no mathematics has been developed and piloted. Despite the simple approach, the statistical methods covered are not restricted to ‘introductory level’ techniques. For example so-called ‘advanced’ techniques such as logistic and log-linear analyses are explained using easy-to-understand words and diagrams. This teaching approach revolves around illustrating and emphasising that the methods covered (parametric analyses) are all part of the same family of techniques, called Generalized Linear Models, GLMs, and can thus be viewed in essentially the same way. Different GLM acquire different names according to the particular combination of categorical or continuous explanatory variables explored and the assumed distribution of residuals (e.g. normal, Poisson or binomial). An understanding of one leads readily towards understanding the others and the subject of statistics can take on a less disjointed, feel. This is regarded as an important goal in statistics teaching.

The method is intended to be applicable to any subject area that utilises statistics. It has been piloted mainly on post-graduate students in Biosciences, and particularly in an MSc module that PhD students may also attend on a voluntary basis. Module participants have evaluated the course as good and recommend that it should be kept the same. Responses to relevant parts of ‘Student Evaluation of Teaching’ questionnaires were generally very positive. One lecture was formally evaluated by a PGCHE assessor and scored highly. It thus appears that this method of explaining statistics is useful and appreciated.

Statistics courses usually, and ideally, have some practical component but this teaching method is mainly applicable to the lecture setting and has not been developed to fit with any particular analysis package. The intention is that these lectures equip students with an understanding of what they are trying to do with statistics and of how the tests work, and that with this they will readily grasp the specific terminology and format of most analysis packages when they use these themselves. Future developments of this method include to create a short, publishable, textbook from the lecture handouts and to expand the method to cover more non-parametric tests. Draft copies of the textbook will be distributed during the presentation for discussion.