Some references. PM me and we can discuss how you might obtain a copy. 'Reducing error and influencing behaviour' HSG 48 HSE Books HFRG (Human Factor Reliability Group) document "Improving compliance with safety procedures - reducing industrial violations". HSE had at one time a section of their website devoted to 'Leadership and Worker Involvement'. Don't know if it's still there though.

I would suggest the HSE website has some of the best resources you will find anywhere in the world. It is primarily aimed at major hazard industries, but the principles remain the same. Have a good look around http://www.hse.gov.uk/humanfactors/index.htm Behavioural safety has many uses, but I would argue the topic of human error is quite different. For example, behavioural safety is often aimed mostly at people working at the sharp end whilst human factors, which is understanding why people fail and how those failures can be prevented, looks at the errors of managers and weaknesses in systems and organisations. You may find the Baker report into the BP Texas City accident of interest. The summary of findings at the front gives a good account of underlying failures that can allow errors to occur and/or for them to have significant consequences. You can download it at http://www.bp.com/liveas...s/Baker_panel_report.pdf

Hi Kieran, does that mean that managers can get it wrong too ? Gosh Merv

The only way you will eliminate the potential for human error is to eliminate the potential for the human to influence what is happening. One of the fundamental principles I was taught many, many years ago is that we should seek to "control the process not the person". In othe words we should seek to structure the workplace, equipment, etc., so that it is intrinsically safe and so that the human cannot by his mistakes be exposed to the hazard. There are several benefits of this: 1. If we make the process intrinsically safe, then we remain in control. The moment we rely upon the human to carry out a task in a certain way we lose the ultimate control 2. If we control the process then we can often find ways to make this fail-to-safe. In other words, any failure will not put the human at risk. I would love to know how you make a human fail-to-safe! 3. The cost of making the process intrinsically safe may appear considerable, but it is almost always less, particularly in the medium or long term, than that of attempting to control the person. One approach should, in my view, be that the health and safety team are involved at the very start, and then throughout the development and implementation of any new process, design of the workplace, specification of equipment, etc. The intrinsic safety can then be designed in from the outset. Making the process intrinsically safe is often achievable, but it generally requires team work, i.e. the involvement of not just health and safety personnel but designers, engineers, the maintenance team and those who will have to carry out the work - and, of course, a comitment by senior management. Chris

I would suggest as others have to look at 'Attitude' and 'Behaviour' however because it gets into the realms of the human element it is in my opinion a minefield. If you can influence the attitude of 'it will never happen to me brigade' you would be onto a winner. I would agree with Chris Packham as this is approach I very much endorse and adhere to. Without the factor of 'control' related to workplace ways of working can result in errors and higher accident rates.

Merv One of the themse of Kahneman's research and current teaching is how profoundly endemic error is in all human behaviour. He emphasises very strongly how error has been inherent in some of his own work, starting with his first professional job, as the first psychologist in the Israeli army. The only technique he very strongly affirms is the statistical technique of multiple regression, which he used to standardise recruitment processes in the Israeli army, based on research by Paul Meehl. I believe this technique also supports the paper he and Tversky wrote which sparked the development of 'behavioural economics' and the award of the Nobel Prize to him, after Tversky's death. The same statistical technique underpins the best research about behavioural safety and ergonomic applications e.g. the NIOSH equation on manual handling, simply by quantifying the relationship between the key variables and safe performance of well-specified tasks. In chapter six of his report, Loftstedt makes the same argument when he strongly emphasises the power of and need for scientific argument about safe behaviour and environments at work.

Many thanks to everyone who took the time out to reply to my post - some very useful thoughts and suggestions for further research. Regards, Neil.

Heavens, you learn a lot in these forums!