The most common, serious fault that we see in cause trees is the tendency to blame people rather than push on to find the real root cause. The remedy is to use an understanding of human factors, but most investigators struggle with this. Luckily the US Federal Aviation Administration, the FAA, has shown how to succeed and has provided easy strategies for investigators.
So what's the problem? An example might help. Perhaps you have been involved in problem solving where the conversation goes along these lines;
- Incorrect drugs given to patient.Why?
- Incorrect interpretation of prescription. Why?
- Nurse error.
- Equipment failed at half normal life. Why?
- Incorrect lubricant was used. Why?
- Operator error.
- Improvement program stalled. Why?
- Participants responsible for the work felt they were not recognised. Why?
- Management error.
This type of discussion assigns a blame or guilt to a particular person or group of people. It may be useful for lawyers but tends to be counterproductive for people wishing to prevent problems happening again.The "hindsight bias" that gives perfect vision of the problem when we look back at it can make it hard to think that the circumstances were not obvious to the operator. So we tend to blame the person rather than asking why they made an error.
It may also be that we are also influenced by a "confirmation bias". Perhaps we think that all managers fail to recognise the efforts of their subordinates so circumstances that resemble that scenario help us to jump to the conclusion.
This approach fails to find the root cause. But much worse than that is the fact that this approach gives the message that Root Cause Analysis is about blaming people. This becomes a major barrier for people participating in future investigations. Who would participate if there is a feeling that the purpose is to point the finger at them or their friend?
So how can investigators avoid this problem?
In a nutshell, whenever the investigation is leading toward blaming "Human error" the team must hesitate and turn the attention to understanding the Human Factors that caused the error. It is often the case that a different person would have made exactly the same error. We have to understand the factors that contributed to the undesirable event if we want to eliminate reoccurrence.
The aviation industry was once considered to be hazardous but now passengers on airliners enjoy the safest mode of travel.
This is to a large extent thanks to the FAA and the approach they have taken to understanding Human Factors. Many people and organisations have done research and have developed ideas about Human Factors. The FAA's success has come from creating and deploying very practical approaches to using the concepts.
The FAA recognised that different people need different approaches to applying Human Factor Concepts. Flying instructors who teach pilots and aircraft engineers responsible for maintaining aircraft need easy to remember tools. They apply these tools in the cockpit or in the workshop throughout the day and in feedback and analysis settings to help understand why pilots and mechanics make errors and to help them avoid the errors.
FAA's flying instructor handbook described this technique with the words "slips" and "mistakes". We'll come back and describe this technique later in this video.
On the other hand investigators into significant issues need the full Human Factor process that incorporates key concepts that form the foundation of this science. The process is called Human Factors Analysis and Classification System or HFACS. In essence investigators use a library of human factors that have been previously developed to flesh out the cause tree rather than simply blaming human error. Not many people memorise all of the 180 strings of factors on the standard FAA listing. Most people need to keep a paper listing to refer to during investigations or use software to provide the factors.
An example will illustrate how software makes this an easy task.
Take the case of an aircraft that has been lost due to controlled flight into terrain.
- The aircraft crashed into a mountain. Why?
- The pilot could not see the mountain. Why?
- The pilot was in cloud and not qualified for instrument flying. Why?
Perhaps we are now tempted to write "Pilot Error"?
But instead of that, let's select Human Factors. The Analysis and Classification System is different in different activities. The system administrator can create and customise the system to suit a particular organisation or activity. Let's select the Aviation database.
Investigators can review the many different combinations to identify likely scenarios. Perhaps it was a skill based error. Perhaps a breakdown in the visual scan?
Or perhaps it was a case of an adverse mental state that created a precondition for the accident. Possibly a condition called 'Get home ites' where the pilot feels compelled to press on passed the point when it is safe.
Or perhaps the pilot was a student and the instructor failed to provide guidance.
Let's say for this example that it was a case where rules were often bent and it is suspected that flight was continued at low altitude in instrument meteorological conditions.
Clicking on the insert button adds this analysis string to the chart.
Now the investigation team would modify the words in each box to more accurately reflect what the evidence shows.
This process has helped the investigators to think through all of the different possibilities in an objective way. However it must be stressed that the analysis is purely theoretical unless evidence is gathered to support or deny the ideas. Gathering evidence is critical to success.
In this case it was found that the weather patterns near the airport frequently have a low cloud base and that experienced pilots familiar with particular land marks frequently bent the minimum visibility rules. Unfortunately the pilot who was flying at the time of the accident was relatively new to the field.
The fact that the investigators went beyond a bald statement of Pilot error meant that remedial action could be taken to stop others falling into the same trap.
You can see that this technique is very powerful and is ideal for thorough investigations. But the investigation team will need access to a Human Factors Analysis for the activity.
Slips and Mistakes
I mentioned earlier that there is a simplified technique based on the idea of slips and mistakes. Instead of the four initial categories described above the simplified technique has two categories: "slip" or a "mistake". Rather than saying "operator error" the investigators would ask whether the behaviour was a "slip" or a "mistake". The FAA flying instructor's manual says;
"A slip occurs when a person plans to do one thing, but then inadvertently does something else."
Slips are errors of action. Slips can take on a variety of different forms.
- One of the most common forms of slips is to simply neglect to do something.
- Other forms of slips occur when people confuse two things that are similar.
- Other forms of slips happen when someone is asked to perform a routine procedure in a slightly different way.
- Time pressure is another common source of slips. Studies of people performing a variety of tasks demonstrated a phenomenon called the speed-accuracy trade-off. The more hurried one's work becomes the more slips one is likely to make.
A mistake occurs when a person plans to do the wrong thing and is successful. Mistakes are errors of thought. Mistakes are sometimes the result of gaps or misconceptions in the student's understanding. Gaps in understanding can happen in many ways. Perhaps an operator "learns on the job" and picks up the short cuts and errors of the previous operator. Or perhaps false assumptions are made. A person uses a fix that worked on other occasions in similar but different circumstances
The process of classifying an error as a slip or mistake helps the investigator to become more objective about the error. The investigator's mind turns to understanding why there was a slip or a lack of understanding. This moves the discussion towards useful countermeasures rather than just blaming an individual.
The key message that I hope you take out of this is to avoid simply blaming a person in your investigations.
People rarely want to fail. It may well be the case that you or I would have made the same error in the situation. It is much more useful to take an objective approach to understanding the factors that led to the error. The full Human Factors Analysis and Classification System is best for thorough investigations but will need some learning, or a paper copy or software support. For day to day problem solving is it helpful to classify errors as slips or mistakes to zero in on what is really causing the problem.