Eliminating human error
“It’s time… we admit we’ve been managing safety wrong.”
Kurt noted the focus on human performance initiatives gained attention after the March 28, 1979, partial meltdown of the Three-Mile Island Nuclear Reactor in Dauphin County, Pa. Coincidentally, Thomas Gilbert, known as the father of human performance technology, published Human Competence: Engineering Worthy Performance in 1978, in which he describes the Behavior Engineering Model (BEM) for worthy performance analysis.1 The following is a visual representation of Gilbert’s BEM from a Crossman, Crossman, and Lovely article.2
Crossman, et al. describe Gilbert’s theory that “the root cause of most performance problems lies not with faulty worker behavior, but rather in the defective context of the workplace (e.g., Information, Resources, Incentives).” (Gilbert, 1978)
Kurt listed several of the seminal contemporary authors in the human performance arena, namely Dr. Todd Conklin, Dr. Sidney Dekker3, Dr. Karl Weick and Dr. Kathleen Sutcliffe4. My focus here is on Dr. Todd Conklin’s work.
Errors almost every hour
As Dr. Conklin points out in his Pre-Accident Investigations – An Introduction to Organizational Safety, humans make errors; even the best of us make errors almost every hour every day.5 Our focus on improving safety needs to be on the system not the person. For example, Conklin notes the number of highway fatalities over the past 50 years has remained essentially level at 38,000 people per year, even though the number of cars on the road has increased. Rather than trying to change the driver, the auto industry has increasingly made cars safer.
Conklin notes we have done just the opposite when it comes to safety at our job sites. “We have tried to get safety performance by ‘leaving everything the same except fixing the worker.’ It is about time to admit to ourselves we have been managing safety wrong.”6
Conklin describes three layers present in our organizations: 1) the individual worker, 2) the organizational system, and 3) the performance expectations established by management. Management tends to only focus on the first layer when failure occurs and attempts to “fix” the worker.7
We need to start thinking differently. Conklin draws upon Eric Hollnagel’s definition of failure, which he defines as: 1) failure happens because the worker believes that what is about to happen to them is simply not possible; 2) failure happens because failure often has nothing to do with the tasks and processes that the worker is currently doing; and 3) failure happens when the worker feels the possibility of getting the intended outcome is worth whatever risk is present in the work environment.8
Two types of failure
Next, Conklin describes two types of failure — Individual Failure and Organizational Failure/System Failure. For the Individual Failure, the worker is the one hurt (e.g., slip, trip, fall, exposure, strain). Not protecting the worker from getting hurt is viewed as an Individual Failure. The Organizational/System Failure occurs when systems allow a threat resulting in adverse consequences. Conklin notes in a System Failure, someone or something has broken through the layers of defenses thought to be in place to protect the facility, people and reputation. Ask yourself, when a failure occurs do you conclude the worker failed the organization or the organization failed the worker? I would agree with Conklin that the latter is more prevalent than the former.9
Three parts of a failure
Conklin presents three parts of failure: 1) Context, 2) Consequence, and 3) Retrospective Understanding.10
Part 1 – Context is everything happening at the time of the work being performed leading up to the failure. Attention must be given to the interacting relationships of all the moving elements of the circumstances.
Part 2 – Consequence is an unexpected outcome that occurs which the organization views as bad. Conklin notes that failures have immediate consequences, likely to trigger immediate reactions.
Part 3 – Retrospective Understanding is everything that happens after the failure. If those workers in Part 1 above would have known what they know in Part 3, there is a good chance the workers would have done something – anything – differently in order to prevent Part 2 from occurring.
How do we change organizations’ reaction to failure? Clearly, everyone pays close attention to how management reacts.
Conklin’s approach moves the organization from a “crime and punishment” reaction to a “diagnose and treat” response to failure.11 He offers the following tools to affect this shift.
Better questions, better answers
Too often safety professionals assume managers know the kinds of questions they should be asking when a failure occurs. Instead, Conklin offers the following:12
1. Your response to an event matters!
2. Are the people OK?
3. Is the facility safe, secure and stable?
4. Tell me the story of what happened.
5. What could have happened?
6. What factors led up to this event?
7. What worked well? What did not work well?
8. Where else could this happen?
9. What else do I need to know about this event?
Make the formal less formal
Avoid a “courtroom-like” atmosphere and run post-event meetings in a way that seeks to learn what happened. Also, use them for successful events.
Learning Teams are ad hoc groups of workers from any level in the organization brought together to answer one question – Something has just happened, what should our organization learn from this event or potential event? “How” the event transpired is more important than “why” and “why” is more important than “who.” Another key element is this is a worker-owned process.13
1 Gilbert, T. 1978. Human Competence: Engineering Worthy Performance. New York: McGraw-Hill.
2 Crossman, R.M., D.C. Crossman, and J.E. Lovely. June 2009. Human Performance Improvement – Key to sustainable safety excellence. In Professional Safety. Park Ridge, IL.
3 Dekker, S.W. A. December 28, 2014. The Field Guide to Understanding ‘human error’ 3rd Edition. Ashgate Publishing Company. Aldershot, U.K.
4 Weick, K.E. and K.M. Sutcliffe. 2007. Managing the Unexpected: Resilient Performance in the Age of Uncertainty 2nd Edition. Jossey-Bass. San Francisco, CA.
5 Conklin, T. 2012 (Re-printed 2014). Pre-Accident Investigations – An Introduction to Organizational Safety. Ashgate Publishing Limited, Surrey, England.
6 Ibid. pp. 12.
7 Ibid. pp. 18.
8 Ibid. pp. 20-22.
9 Ibid. pp. 22-23.
10 Ibid. pp. 24-26.
11 Ibid. pp. 39.
12 Ibid. pp. 40.
13 Ibid. pp. 43-44.