Optimising experimental design needs a sound understanding of basic principles and a good appreciation of a range of designs and when to use them.  More than ten years ago it was clear that among scientists using laboratory animals there was a widespread lack of recognition of the importance of valid comparisons and avoidance of bias, and a reluctance to use efficient designs. This has been emphasised by several publications over the last decade.  Education was seen as a key element in addressing this, but teaching experimental design is different from teaching statistics as a subject, and where animals are used design involves more than data collection and valid comparison of experimental groups. Interactive workshops have evolved as a good way of giving researchers the knowledge, understanding and skills to optimise their design for the experimental question being addressed and the constraints within which they have to work. 

An optimal design would be one which will produce valid results, be efficient in use of experimental material, and involve least severity for the animals used.  The talk will consider what is needed for this, as that determines the content of workshops or other types of teaching on experimental design using animals.  It will also consider the range of designs and their usage, and how researchers might acquire the skill to select an appropriate and optimal design.  To assist this, tutors could think of an educational 3Rs – repetition, recapitulation and reinforcement.   How this approach has been incorporated into an effective mix of presentations and problem solving will be illustrated.