Causally Appropriate Graphical Modelling of Time Series

This work provides an accessible introduction to the use of causally appropriate graphical modelling of time series data. This includes a discussion of what role, if any, causality has in the field of statistics with specific reference given to Granger-causality. The graphical modelling for time series approach, GMTS, proposed by Reale (1998) is presented and critically assessed with reference to the analysis of datasets from economics, ecology and the environmental sciences. This book is recommended for anyone who is interested in, or would like an introduction to, the application of graphical modelling for time series analysis and/or issues surrounding causation in statistics. In addition, many prospects and challenges are identified for future research in the wide range of fields given treatment in this thesis.

Carla S. Meurk, MSc (1st class honours): Studied Philosophy, Mathematics and Statistics at the University of Canterbury. Currently studying towards a PhD, University of Queensland, Australia.