EHS Today - June 2009
By: Terry Mathis, ProAct Safety
Printable Version
Three years ago, I was asked to help an organization improve their safety performance. The organization was a maintenance group that performed heavy and dangerous work. They had less than 250 employees with 17 recordable incidents from the prior year. The workers were well experienced and their safety manager was organized and caring. After asking them what their current safety focus was, I was told that the safety efforts were currently focused on two items: PPE and housekeeping.
After completing a specialized Pareto analysis of the organization’s accident data, I determined that neither of these two items would have prevented any of the recordable incidents from the previous year. I then interviewed 100 workers and asked them what they perceived to be the most likely way to get injured on the job. I received 85 different answers.
When the results of the analysis were shared, and the safety efforts were refocused to the top four items from the analysis, they reduced their incidents to three recordables the next year, followed by one the year after that.
Most people view perceptions as something to be measured; not managed. But I have found that if perceptions are not managed, they will have variance that can cause lack of correct focus in safety efforts. Workers who do not accurately perceive their greatest risks often waste their limited safety efforts on ineffective strategies. Correcting perceptions can direct safety efforts for maximum effectiveness.
Unmanaged perceptions can be impacted by several sources, including worker experience, common sense, un-tabulated data, and the limitations of memory.
Perception Source 1: Worker Experience. Perceptions that are formulated around a worker’s individual safety experience are very limited. One person’s experience seldom equals a statistically significant sample of workplace safety issues. Few workers base their perceptions solely on their own personal experience, but also rely on accidents they have witnessed or discussed with fellow workers. In interviews with over 17,000 workers, I have rarely found a worker who based his own perception of risk on knowledge of more than 10 accidents. Many workers who have not yet been injured on the job have developed an attitude of “it won’t happen to me” based on all the accidents that haven’t happened to them (yet).
Perception Source 2: Common Sense. Perceptions based on common sense have a potentially broader foundation than those based on personal experience alone; but they, too, are limited. The term “common sense” is often misunderstood. It originally referred to having sense about common things that happen commonly, regularly, or obviously; in other words, high-probability outcomes. The problem is that most industrial accidents are not a result of high-probability risks. In fact, research indicates that most industrial accidents are the result of relatively low-probability risks. Common sense is the very reason a worker won’t stick a piece of metal into an electrical outlet, but will use the wrong tool for a job and feel safe doing it. Any risks that result in an injury once in a hundred or more times, usually escapes the attention of common sense. When a worker has taken the same risk many times accident-free, it is no longer viewed as a risk.
Perception Source 3: Un-tabulated Data. Many safety professionals share every accident report with the workforce. Many even include accidents from other sites or other organizations. However, if the data is not tabulated in a way that helps workers to assess the magnitude of risks, there may be no benefit to viewing it at all. Workers who have been exposed to many accident reports during the year do not accurately perceive their greatest risk, even though such information can be determined from the analyzed data. Too much data with little or no tabulation often causes vastly different perceptions among workers. Each worker tends to base their perceptions on a sub-set of the data.
Tabulation alone does not always align perceptions. Tabulating the data using criteria such as body part injured or accident type is of limited use. The worker knows which body part is at greatest risk or which type of injury is most common, but not the best strategy to prevent the injury.
Perception Source 4: The Limitations of Memory. In our interviews, workers bombarded with accident reports cannot usually recall the details of more than 10 accidents. The record was 31 and the average is 9.7. Neither of these numbers constituted a statistically significant, or representative, sample of workplace injuries. In over 40% of the interviews, workers did not see any discernable pattern within the accidents they could recall, and were not focused on the most common risks represented by that set of accidents.
So, worker experience, common sense, un-tabulated data, and limitations of memory help cause the failure to focus workers on the items that are most likely to cause their next accident. Incorrect focus can cause personal safety efforts to produce sub-optimal results. The problem is not lack of effort, but lack of accurately focused effort.
If these four sources are forming most safety perceptions and those safety perceptions are not accurate, what can we do to change this situation? The answer is to manage the perceptions rather than letting these sources form them. A good safety perception management process has several elements that center around transformational thinking and good communication practices.
Transformational thinking is an alternative to Kaizen. Transformational thinking does not ask how we could make small continuous improvements, but rather asks what one thing, if done differently, would make the greatest improvement? Transformational thinking must begin with the safety professional(s) and the management team before it can effectively reach the workforce. Pareto analysis is a key tool of such thinking, but not the traditional Pareto analysis of body part and accident type. The Pareto of transformational thinking is a Pareto of solutions, not problems. Such analysis requires a different thinking than that of most accident-investigation processes. An accident that has not yet occurred doesn’t have a root cause and may have multiple potential prevention strategies. When using past accident data, you can’t focus on “why it happened.” You have to ask “what could have prevented it” and rate these potential solutions by their potential impact on accidents.
Once the transformational Pareto is completed, it must be 1) communicated to the workforce in a way that promotes understanding and allows for questions and answers, and 2) a part of each accident report from this point on. Workers must accurately perceive what prevention strategy will have the most impact. Each new accident should either reinforce or modify this thinking. As long as accidents continue to fall into the targeted categories, workers should focus on strategies to prevent these types of accidents. Efforts focused this way tend to produce significantly superior results than efforts directed by unmanaged perceptions. A shift in accident trends should result in a refocusing of safety efforts toward the new greatest risks and prevention strategies.
Many sites that are engaged in perception management supplement the efforts with memory tools and activities to promote the focus. Acronyms are utilized to promote ready-remembering of the top 4-6 accident-prevention strategies. Posters are displayed with the acronym strategies, photos, descriptions, examples, Pareto percentages, and other information to help form and maintain the focus. The focus strategies are the topics of tool-box and shift-startup meetings. They are the subject of training sessions. Some are used to form a checklist for supervisor or peer audits. Sites gather data to measure the use of these top strategies and to help identify barriers that make it difficult or impossible to use them. This data can be used to formulate action plans to facilitate the workers’ use of these accident-prevention strategies.
Another aspect of managing perceptions is ongoing communication. Most sites under-communicate safety data and cause another inaccurate perception - that safety is not a high priority at the site. Most workers equate priority with the percent of time expended on it. So, if managers talk about production ten times for every one time they talk about safety, workers perceive that production is a higher priority than safety. Likewise, if accident data does not get communicated frequently and if it does not accurately reflect the proper priorities for accident-prevention strategies, workers will start to rely on the four sources of inaccurate perceptions. It is difficult to over-communicate safety and easy to adjust if you find yourself doing so.
If you are not managing perceptions, doing so could potentially transform your safety results. If you are already trying, further improvements in your techniques and/or level of communication could also be beneficial. The internal and external costs for doing such a project are minimal. Many sites are already putting forth enough effort, they simply need to better direct that effort. Sites utilizing these techniques report not only improved bottom line, but improvements in overall safety culture and trust levels in the workforce. Managing perceptions creates F.O.C.U.S. (Forming One Common Understanding of Safety).
