Adrian W. Bowman

In studying any subject for the first time, motivation is important. For probability, some might be intrigued by the logical framework it provides but for many it is the relevance to contextual problems, and sometimes the appearance of surprising answers, which stirs up interest. Examples of motivating problems, some well known and some less so, will be discussed.

The task of identifying probability models, even in the very simple form of appropriate distributions, from a problem context is a skill which requires practice. Technology can help by providing a supportive environment in which students can be prompted with simple guidance, to help in developing the skill of identifying appropriate distributions throught the appropriate characteristics. Some 'Model Choice' software (developed by John McColl and others in Glasgow) will be described.

In addition, animated graphics has great potential in assisting the intuition of those to whom these ideas are new. Some different forms of graphics will be described and illustrated, on problems which range from elementary to non-mathematical explorations of more sophisticated concepts such as spatial processes. Tools for creating material of this types, such as the rpanel package for R and some similar web-based tools, will be outlined.

**Keywords:** Animation; Technology; Teaching

**Biography:** Adrian Bowman is a Professor of Statistics in the University of Glasgow. His research interests include flexible regression models, currently being applied to spatiotemporal environmental data, and the analysis of three-dimensional shape data, currently being applied to human faces. He has played a variety of roles in the Royal Statistical Society, including a term of office as Secretary of the Research Section, as editor of Applied Statistics and a recent spell as an Honorary Officer. However, he also has a longstanding interest in the use of technology in teaching, dating from the 1980's when he first began to write dynamic graphics software with his colleague Derek Robinson. More recently he has been a key member of the development team of the rpanel package for R. He has also played a leading role in the Higher education Academy Mathematics, Statistics and OR network in the UK.