A Guide to Analysing Counts and Proportions in Complex Situations.

Variation, correlation, consternation? For nearly 100 years problems from agriculture and biology have presented strong challenges to statisticians, and then biologists have often felt equally challenged in applying their solutions. So even though, statisticians may feel that they have developed a range of very effective methods to analyse counts and proportions from complicated experiments, these methods have yet to take their place in the routine toolkit of many biologists.

This talk will provide guidance to some of these new methods. It will describe the types of biological investigation that have led to the development of methods such as generalized linear mixed models and hierarchical generalized linear models. It will show how these methods extend the more familiar generalized linear models to allow you to take account of additional sources of error variation, such as blocking in a field experiment, or parental effects in animal experiments. It will explain the output that is obtained - what it means and how to use it to assess the fixed and random terms in the model. Finally it will show how to present the results for publication.

Real examples will be used to illustrate each aspect, and the aim will be to provide the insights needed to recognise when the methods are needed, and the confidence then to apply them.