Related Pages:

EuSpRIG Amsterdam 5-6 July 2001

http://www.gre.ac.uk/~cd02/EUSPRIG/

The papers included:

1. Harmen Ettema, PricewaterhouseCoopers, NL "Spreadsheet Modelling: Practice, audit and future"
2. J. Paine, University of Bristol, UK "Safer spreadsheets with model master"
3. Raymond Butler, H.M. Customs & Excise, UK "Applying the Cobit framework to spreadsheet development"
4. Brian Knight, University of Greenwich, UK "An evaluation of the quality of a structured spreadsheet development methodology"
5. Patrick O'Beirne, Systems Modelling Ltd., Ireland "Euro conversion in spreadsheets"
6. J. Raffensperger, University of Canterbury, New Zealand "New Guidelines for writing spreadsheets"
7. Grenville Croll, Andersen Business Modelling Group, UK "A history of spreadsheets"
8. D. Nixon, University of Salford, UK "Spreadsheet auditing"
9. David Chadwick, University of Greenwich, UK "Peer audit and self-audit for reducing errors"
10. Paul Janssen, PricewaterhouseCoopers, Netherlands "Spreadsheet assurance by ´control around´ is a viable alternative to the traditional approach"
11. D. Knight, Brixx, UK "A real alternative to spreadsheets"

Reflections on the conference, by Patrick O'Beirne.

In common with many quality conferences, the ever-present fight is against unjustified optimism. Ray Panko of Hawaii University pointed out at the 2000 symposium that, just like everyone else, experts only learn from experience if they receive and review systematic feedback. His classes include an exercise where only 14% of students get a spreadsheet correct; when asked to put up a hand if they think they are in that 14%, more than 50% in the class do so.

Brian Pettifor of PwC commented that "We never fail to find an error, but the world is not tumbling down around us." Their mandatory practice is a risk assessment by management of the criticality of the spreadsheet. Tools can make it easier for people to adhere to standards to allow others to audit the spreadsheet more easily. To mitigate higher-level errors that will slip through a technical review, they recommend a shadow (parallel high-level) model. The question is what to do if problems are found - fixup, re-implement, re-engineer…?

David Chadwick of Greenwich University emphasised that good teaching is not just "click-bites" but how to recognise and avoid errors. Students need a consciousness of the business costs and consequences of errors. They are surprised that others can have different versions of the same specification.A common teaser that works well is "Can you find all the errors in this?"

Proceedings are for sale at GBP 25.