In late 2006 I had the pleasure to attend a most inspiring lecture on "The Myths of Risk" by Terje Aven during the annual Synergi Network Meeting in Stavanger. This was reason enough for me to check out the man's work (which numbers an impressive numbers of articles and books, growing by the year - more reviews to follow on this site in due time) and over the last couple of years I also had the pleasure to meet him on several occasions. 

Flipping briefly through the pages (only 190 in total, of which the last 45 are dedicated to appendices and reference lists) of this book, it appears to be pretty technical at times. And Aven goes a long way in explaining the Bayesian approach to uncertainty and stuff like that…

I have to admit that occasionally I found myself lost in the formulas - and that despite the very good first appendix that explains/refreshes the basics of statistics. Frankly: I skipped some of these more technical parts, and in the end I didn't miss all that much, because the book is not so much about building risk assessment models (an impression one might get from briefly flipping through the pages), but rather on discussing how to approach this, the general thinking processes to risk and uncertainty, and (maybe most importantly) how to use it in the decision making processes.

Let's stress this right now. According to Aven, and I wholeheartedly agree with him, risk analysis is not the same as decision making, but merely one tool in the process. This process involves also judgement, personal bias and management review. A quote says: "Risk is primarily a judgement, not a fact".

In the first chapter (actually Chapter 2, as Chapter 1 is only a short introduction), Aven discusses traditional approaches to risk (like the 'best estimate') and the shortcomings of these methods, stating why there is a need for new, or rather renewed, thinking on the matter.

Here one of the positive aspects of the book already shows: Aven uses different views to the phenomenon of risk; from the (most commonly used) accident risk to reliability to finance and project risk. This adds extra flavour and cross learning from various disciplines. Another positive aspect from the book: each chapter is concluded by a paragraph that discusses relevant used literature and sources with critical and useful comments.

Chapters 3 and 4 then tackle the problem and lay down structures of how to think about risk and risk analysis, and how to handle uncertainties and specify probabilities with a paragraph on modelling principles (the big picture, mind you, no detailed 'how to' here). Again this is illustrated by examples from diverse disciplines. Interesting - and most valuable in discussions - is the stress that Aven puts on so-called observable quantities that tell us something about 'the world'.

According to Aven: "probability is an expression by a person based on some knowledge about an observable quantity" (i.e. it depends upon background knowledge which forms the basis of an evaluation), and that "a probability is a judgement, and there is no strict mechanical procedure producing one correct value". This should be an important principle to recall when entering discussions about risks!

Chapter 5, then, is probably the most important - and most widely applicable - chapter of this book. Discussed here is how Risk Assessment should be used to support (NOT: perform!) decision making. This part of the book discusses what good decisions are, how to make them, and how to handle conflicts and trade-offs (again a diverse number of examples passes by). Then follows a discussion of risk problem classification schemes/strategies (i.e. risk-based, precautionary and discursive) and their applications. Also a discussion is enclosed how 'closeness to hazard' and 'level of authority' influence the decision making and risk treatment strategies.

The final (short) chapter provides a high level summary and conclusions with directions to relevant pages, providing a good bookmarker.

Overall this is recommended reading for those dealing with risk assessment and risk management on a regular basis. It might provide a good eye-opener (or refresher) or two…

For further information, let me include here the book's description from the back cover:

Everyday we face decisions that carry an element of risk and uncertainty. The ability to analyse, communicate and control the level of risk entailed by these decisions remains one of the most pressing challenges to the analyst, scientist and manager. This book presents the foundational issues in risk analysis - expressing risk, understanding what risk means, building risk models, addressing uncertainty, and applying probability models to real problems. The principal aim of the book is to give the reader the knowledge and basic thinking they require to approach risk and uncertainty to support decision making.

• Presents a statistical framework for dealing with risk and uncertainty.

• Includes detailed coverage of building and applying risk models and methods.

• Offers new perspectives on risk, risk assessment and the use of parametric probability models.

• Highlights a number of applications from business and industry.

• Adopts a conceptual approach based on elementary probability calculus and statistical theory.

Foundations of Risk Analysis provides a framework for understanding, conducting and using risk analysis suitable for advanced undergraduates, graduates, analysts and researchers from statistics, engineering, finance, medicine and the physical sciences, as well as for managers facing decision making problems involving risk and uncertainty.

ISBN 0471495484