Think of Challenger IV, the Bhopal chemical plant in India, Chernobyl, the Herald of Free Enterprise ferry, the Hatfield rail crash, and Three Mile Island, amongst many other disasters, and you get some idea of the consequences of large-scale error. These events are thankfully rare, but small-scale errors occur every day in organizational areas as diverse as production quality, procedures, protocols, equipment use and task design, as well as human functioning. And, while it is our preference to work with issues of error before they turn into major incidents, we are available for accident investigation should the need arise.

To determine the best way to intervene and minimize the chance of small- or large-scale error reccuring, Psychonomics takes its cue from psychological research, which defines error, simply, as a deviation from an intended outcome. From this perspective, the cause of error can originate in any part of an organizational system. Normal cognitive functioning, for example, is seen as one major source of error; where any error is a natural consequence of knowledge extraction or use. Highly complex systems - with humans as a distinct part - are, therefore, guaranteed to fail at some stage, if, quite simply, the capacity of human cognition is exceeded or changed in some way. The solution, which we seek to implement, is to make the system and all its elements fit the person, not the other way round. And, as will be apparent from the information on this page, our brief as industrial psychogists is much wider than simply reworking the human factors component.

Another source of organizational error stems from innate neurological functioning, This can cause human behaviors, or indeed biases, to be maintained. And this is why procedures and protocols that should have been implemented are often not, or indeed, where these system constituents exist but it has become common practice to disregard parts of them. A common situation that we have also dealt with is the manager who ploughs on using the same limited approach to employee concerns about, say, production quality or scheduling pressures. This individual may do so, research suggests, because of inherent, evolutionary programming. It all harks back to our caveman ancestry where the approach may have been beneficial in the dogged persistence of following prey till it was caught. But in these times and situations where open-mindedness and flexibility are often key, it is a handicap, and to change, as well as address the current problem, outside assistance is needed.

In many circumstances error isn't a difficulty, the system - made up of task structures, procedures, people, and so forth - copes; or the system is, in fact, non-critical, so that any errors do not matter. When the error is severe, though, it may be the cause of a serious accident. Yet, despite what you may have heard in the media when scapegoats are sought following terrible incidents, blame can rarely be solely attributed to an individual employee, and, more importantly, errors cannot be eradicated totally, only their antecedent factors dealt with to minimize their effect. The goal, so fondly trotted out in news interviews by senior executives following crashes, sinkings, derailments, equipment failures, and the like, of a zero error profile is unnatainable. The certainty is that errors will happen again in some form, and steps need to be taken to allow for this so that their severity and impact are minimized. For any good organization this should be an ongoing activity. Defining what these steps are and then implementing methods to take them is what Psychonomics will help you work towards. Where should the process of error management begin? Error, we find, in most circumstances, is a function of the system as a whole and the way it has been constituted, as well as the interactions within it including those of its human components or operators. Deal with the whole system - rather than waiting for a serious event to occur and then aportioning blame - and you will also deal with error.

We consider the wider context too. Why have violations or the conditions for error occurred? Sometimes, like a series or tremors before an earthquake, we find that minor errors prior to major ones - such as with respect to quality, procedures, or safety - have been overlooked, even when they were known by staff and reported to management. And, in many instances, we find that there is reinforcement for emphasizing performance, money or promotion, and this has overridden safety concerns. Indeed, the reinforcement for emphasizing safety is a non-event - ie the accident hasn't happened. This asymmetry often leads to conditions in which the seeds for future accidents can thrive.

The systemic basis of error aside, most companies know something about the practical steps needed to reduce error, and these are often a good place to start. For example, making displays, dials, meters, and indicators more user-friendly, and not prone to predictable perceptual errors. Further, sharp changes, such as a move from electromagnetic to electronic instrumentation are to be avoided. This was, in fact, the cause of a plane crash at Kegworth in the UK, rather than the simply attributed 'pilot error'. However, Psychonomics also considers the effects of memory, decision-making, reaction time, attention, and arousal, on employee functioning.

In terms of more far-reaching solutions, the implementation of a safety culture is obviously primary. However, other interventions we may apply include:

	A change in the safety-efficiency trade off
	The application of training, including simulations
	Improvements to procedures and protocols
	The implementation of state-change feedback systems
	Computer support systems for setting safety parameters
	Reductions in multi-modal complexity
	Changes to shift patterns
	Enhancements to management commitment
	Changes to task design where, amongst other things, ways are sought to minimize the load on human functioning under high stress conditions.

In addition, together with our client, we seek to make the organizational system as error tolerant as possible. As we've noted, some types of error are inevitable, and may even be acceptable in a situation that requires a trade off between speed and accuracy. Further, in any complex system it is impossible to specify every condition that might lead to an error or accident - and it might also be impossible to specify all the many ways to reinstate the system. The error tolerant system that we would propose attempts to avoid irreversible actions, where the actions of an employee would have a disastrous and irrevocable effect on total organizational functioning. When it comes to reducing error, recovery, in our view, must always remain an option.