Safety in Numbers

When developing strategies to improve safety, whether on the roads or in the workplace, complete and impartial information is essential. This is a challenge when people's lives are involved, as the general public and those making decisions about strategies interpret increases in incidents and accidents differently. The challenge is to objectively analyse safety statistics to inform strategies that aim to reduce both the rate and total number of accidents.

The statistician's role is also to provide guidance on what can be reliably learnt from year-to-year patterns and separate out the real changes in rates and numbers from those that happen just by chance. For example, the trends in road fatalities across Australia have shown real decreases in both rates and totals. In 1990 there were 2,331 road deaths, which equates to 13.7 per 100,000 population. In 2005 this had reduced to 1,636 deaths or 8.0 per 100,000 population. If the rate had remained constant the number of people killed in 2005 would have increased to almost 2,800. The extent of the reduction would not be fully appreciated by considering the total numbers alone.

Our involvement in road safety analysis began in 1989, when we conducted a statistical review of the first year of Random Breath Testing (RBT) in Western Australia. That and subsequent studies across several states with Professor Ross Homel of Griffith University focused on the critical outcome – numbers of serious crashes – and showed that RBT was not automatically effective in reducing crashes. However, when RBT was implemented with high visibility and genuine randomness of people tested, it did contribute to a substantial reduction in road trauma. The statistical analysis and qualitative work provided guidelines for more effective RBT. 

Since then, we have conducted assessments of the effectiveness of school speed zones, reported on Australian heavy truck crash statistics and produced detailed reports on Western Australian road crashes over three ten-year periods.

We are currently assisting the Bureau of Transport and Regional Economics to develop the statistical methodology for an evaluation of the National Black Spot Program and preparing annual road crash reports for the Western Australian Office of Road Safety. 

In workplace safety, we have evaluated the effectiveness of WorkSafe WA's ThinkSafe SAM advertising campaign in 1999, and are currently assisting the National Offshore Petroleum Safety Authority in the assembly and analysis of their safety data to monitor leading indicators of unsafe environments.

As organisations are expected to be more proactive in their approach to safety, proper analysis and objective measures of performance are becoming more critical. 

Further information can be obtained from Data Analysis Australia at daa(at) or phone 08 9468 2533

December 2006