Perhaps we need to look more closely at those expectations. Can we break the paradigm that says lasting improvement and culture change take three to five years? No doubt it requires a long-term commitment. But recent experience with several clients and an examination of current research findings leads us to this suggestion: Safety professionals should examine the opportunities to achieve fairly immediate, measurable improvements.
Over the past 15 years we have worked to reduce the cycle time involved in implementing behavioral safety processes. In the last five years, we reduced the time required to design and implement these initiatives from an average of 18 months to four to six months. Once such procedures are implemented, organizations experience almost immediate reductions in accidents.
In one case, accidents before implementation were in a state of statistical control. The system regularly produced an average of one accident per month. Immediately after initiating behavioral observations, the rate dropped significantly.
This corresponds with most of the existing empirical research on behavioral safety (for an excellent review of such studies, see Sulzer-Azaroff, 1994). All of the existing empirical research studies show that such approaches provide both immediate and statistically significant reductions in injury accidents.
In a detailed reexamination of the original Hawthorne studies, Parson (1974) noted that the changes in performance reported in the original study are easily accounted for by differences in the feedback available to each group. For example, all employees in the study were paid based on the group's production figures which were posted on a bulletin board for all employees to see. The smaller group of employees who participated in the study were removed from the larger group to a separate area, paid on the production rate of their now smaller group, and their production figures posted in their new work area. The experimental group's feedback was more personal.
The point is that the Hawthorne Effect demonstrates the importance of good experimental design, and many well-designed studies suggest that implementing behavioral safety procedures will result in immediate and measurable improvements in safety.
1. Know the elements that have to be addressed and identify a series of steps to focus on each one.
These elements are milestones for each step, and team members can have a clear understanding of what they are trying to achieve. Each milestone should have a clear target that fits into a realistic project schedule.
2. Invest the time to plan the process in day-long off-site meetings. This is better than in two-hour meetings every other week. Our experience is that a well-planned process, facilitated by a knowledgeable expert, can result in the design being completed in four two-day meetings. Using two-hour meetings, implementation efforts often require two years or longer.
3. Benchmark other companies rather than create the process from scratch. The design team certainly should not simply copy what another organization has done, but it should look at several examples of what others are doing.
4. Conduct a kick-off meeting and begin observations when approximately 20 percent of the employees complete observer training.
Then continue training for other employees. Training is the final place where teams can save time. Often organizations do not begin implementation until they have trained all employees in how to conduct behavioral observations. Most existing studies report that results begin as soon as observations begin.
5. Avoid unnecessary training.
Employees participating on the design team need the details, but most employees don't need extensive training in behavioral concepts to complete safety observations. They should have a basic understanding of the behavioral process and its rationale, and they need to know how to complete and discuss observations. Such training should typically require eight hours or less. Much of our traditional thinking draws a dichotomy between the quick-fix and a longer-term perspective. We are seemingly asked to choose between the tortoise and the hare. If the hare hadn't stopped short of the goal in the fable, the choice would be obvious. Most existing research suggests that implementing a behavioral process can produce immediate results.
The first question is: How do we implement such a process effectively? That's what this article has attempted to answer.
The second question is: How do we achieve lasting results? Here the unfortunate reality is fairly simple‹lasting change takes lasting effort. And this is not so easy.