Quantitative analysis: about right or precisely wrong
Dr Dale Cooper spoke at RISK 2016, a conference organised by the Risk Engineering Society and Engineers Australia in Sydney on 18-20 May 2016.
Outline of the talk
Quantitative analysis of uncertainty has a long history. Early work was based on relatively simple approaches that could be resolved manually or with hand calculators. Modern computing allowed calculations to be undertaken more efficiently, but the methods of analysis evolved more slowly. This presentation discussed the reasons why some quantitative models still generate gross errors and suggested ways of improving the implementation of quantitative risk analyses.
The main sections of the talk were:
- Introductory health warnings!
- Why do we build models?
- Getting the numbers right
- How can we do better?
Quantitative modelling is not as easy as consultants and software vendors pretend. The modelling process must be managed carefully.
The key is generating insights that will support and add value to decisions:
- Model complexity and detail may not always be needed
- Keep the models as simple as possible
- Validate the process carefully.
About right is far better than precisely wrong!
Would you like a copy?
If you would like a copy of the presentation, please contact us.