The Generic Error Modeling System (GEMS) has developed a framework for understanding error types and designing error prevention strategies.
During just about every conference at which I speak at least one person asks, “Okay, what is this human performance stuff?” I typically answer by making an analogy to behavior-based safety programs that are directed at observing and changing the behaviors of workers to produce a safer work environment and to reduce injuries. I like to say that Human Performance is behavior-based safety on steroids, because it looks not only at the individual’s behaviors and actions, but also at the organization’s.
The study and practice of Human Performance is relatively new. The study of errors from a psychological standpoint began long ago, but those studies were conducted from an academic standpoint designed to classify errors into cognitive and non-cognitive categories. They had little practical application.
In the early 1970s, Dr. James Reason began looking at errors both from an academic and practical standpoint. He utilized the research conducted by his peers and predecessors to develop a framework to understand error types and, once understood, to design error prevention strategies. The framework that he developed is named the Generic Error Modeling System (GEMS).
In a speech to a group from the Institute of Nuclear Power Operators, Jeff Lyash, President and CEO of Progress Energy Florida, bridged the gap from the application of human performance principles at nuclear power plants to the transmission and distribution organizations within those companies. He asked the group if line and substation work was hazardous and contained risks, and followed by asking if an error could result in significant adverse consequences that were unforgiving in nature. The crowd responded affirmatively.
Lyash was following Dr. Reason’s practical application of the principles to industry. Dr. Reason ascertained that the framework was applicable to industries in which there was a large degree of risk and in which the consequences resulting from error were unforgiving.
The framework of Human Performance is captured in five guiding principles:
• People are fallible, and even the best make mistakes.
• Error-likely situations are predictable, manageable and preventable.
• Individual behavior is influenced by organizational processes and values.
• People achieve high levels of performance based largely on the encouragement and reinforcement received from� leaders, peers and subordinates.
• Events can be avoided by an understanding of the reasons mistakes occur and application of the lessons learned from past events.
Within the framework for Human Performance it was determined that errors occur primarily through two modes: slips and mistakes. Slips are errors that occur during routine activities that do not require thought. These activities are so repetitive that we often think of them as things that we could do with our eyes closed. Mistakes are errors that occur during activities that require thought. They may be activities that we have some familiarity with or new activities that we have not performed before, but both require us to engage our brain and think through the situation and our actions.
The Generic Error Modeling System has three performance modes in which errors occur. The performance mode is determined by the individual’s familiarity with the task. From this familiarity with task an individual naturally pays a certain level of attention. If an individual is very familiar with a task, his attention level is naturally low; conversely, if his familiarity with a task is low, he naturally pays more attention to the performance of the task.
Progressing from most familiar to least, the three performance modes are skill-based, rule-based, and knowledge-based.
The skill-based performance mode is characterized by routine actions in a familiar setting; these are activities that we take for granted, like driving a car. Most of us who have been� driving for several years do not think about how many degrees we will rotate the steering wheel to make a turn, or at what precise moment we must return the steering wheel to a neutral position to complete the turn. We just do it.
In the skill-based performance mode we do not consciously think about the actions we are performing, we are acting from memory, as though we are on auto pilot. Errors in this performance mode result from slips or lapses in execution due primarily to lack of attention. Because these activities are routine, our natural level of attention is already low and small distractions can divert our attention away from the task and introduce a higher probability of an error.�
The rule-based performance mode is characterized by the performance of prepackaged actions taken because of the recognition of a familiar situation.� These recognized situations are not as familiar as skill-based activities, but they have been previously experienced and we utilize rules developed from experience to negotiate through the task. We apply an if-then logic to the situation; if this happens, then I do this to complete the task.
Using the driving analogy again, when driving and a siren is heard, we assess the situation and apply preprogrammed actions to return to a skill-based performance mode. We determine which way the emergency vehicle is traveling and whether we need to pull over, or just be aware. Although we have some level of familiarity with rule-based activities, they are not routine and require us to engage our brain and think. Errors in the rule-based performance mode result from misinterpretation; we fail to recognize the changes in the routine task and therefore do not apply the correct rule to complete the task successfully.
Knowledge-based activities are those with which we have the least familiarity. This performance mode gets its name from what is gained in these activities—knowledge. These are unfamiliar situations requiring an individual to use analytical skills and judgment to complete. Returning to the driving analogy, suppose you’re traveling down the road, approaching a stop sign, and the brakes fail to respond. For most of us this is a completely unfamiliar situation requiring us to think through the actions necessary to safely stop the vehicle.
Errors in the knowledge-based performance mode are a result of misdiagnosis. In situations that are unfamiliar we do not have, or recognize, all of the information needed to make an informed decision. We rely on assumptions to guide the decision-making process. The chance for error with missing information and assumptions is very high.
The error modes provide terminology that we can use to describe an individual’s actions and how they are influenced by human nature. Another component of Human Performance relates to the organization in which groups of individuals perform activities. The organization’s strategy, goals and processes can also impact the rate at which individuals are involved in errors. If the organization puts in place competing priorities, or has ill-defined expectations or work practices, the chance for errors increase.
Defined work practices, documented processes and business controls are all examples of barriers to errors. Organizations that are effective at applying layers of barriers, referred to as defense in depth, reduce the chance for errors for individuals. Barriers can sometimes be error enablers, if they are not well thought out, too cumbersome or designed with complimentary flaws. Effective barriers are developed by determining the root cause of errors and addressing those causes without introducing new error opportunities.
The guiding principles of Human Performance tell us that no matter how proficient we are, we can all make mistakes; that situations in which errors occur are often predictable and therefore preventable; that we are impacted by the culture of the organization in which we perform activities; that we respond to positive reinforcement from our coworkers; and that if we tie all of this together and learn from past mistakes, we can avoid errors. Having knowledge of the performance modes and the associated error modes, we can recognize the pitfalls we face in a task and take appropriate measures to minimize the chance for error. iP
© 2004 - 2021 Incident Prevention™. All Rights Reserved.