What’s the standard?The first lesson was revealed in all the arguing about the definition of a vote. We added “butterfly ballot,” “undervote,” and “chad” to our vocabulary, along with “hanging chad,” “pregnant chad,” and “dimpled chad.”
Lesson 1: Begin with a clear operational definition of the behavior you are evaluating. Develop your definition through group consensus-building among those involved in the process.
Safety leaders realize the importance of defining safe operating procedures and making sure everyone knows the definition. Government agencies such as OSHA develop safety standards, but that’s not enough. Top-down rules and regulations need to be defined to fit particular work situations. When employees participate in consensus-building exercises to customize safety standards, you maximize their understanding of safe work procedures — and gain the group ownership and commitment needed to support the behavior required by the standards.
This lesson is one of the key factors contributing to the success of behavior-based safety. When work teams develop a critical behavior checklist (CBC) to use in an observation and feedback process, they define safety guidelines in specific operational terms so everyone knows exactly what to do to meet the standard. They accept the standard because they’ve tailored it to their own work practices.
Is the standard adequately defined?The Florida recount was problematic not only because standards weren’t defined by the participants, but also because reliability was not assessed. I’m sure you recall observers holding punch-card ballots up to the light and talking among themselves about whether enough light shined through to count as a vote. It appeared the vote counters were discussing among themselves what the data meant. This obviously slowed down the process, and prevented an estimate of what is called “inter-observer reliability.”
Lesson 2: Estimate the reliability of a standard by having two or more observers examine a situation independently using the same operational definition. If they don’t agree 90 percent of the time or better over a series of test trials, the criteria for reaching a standard must be refined with input from those who performed the assessment.
An inter-observer agreement rate below 90 percent indicates unclear or misunderstood guidelines, and the need for more work at clarifying the standard. This is in fact the process used to develop a CBC for behavioral safety.
The CBC contains a list of behaviors defined by a work team as most relevant to keeping each other injury free. Safe behaviors are defined to increase and maintain through environmental change and/or interpersonal support, and at-risk behaviors are specified to decrease through corrective action.
The CBC definitions are derived initially through group discussion, but subsequently each definition is applied and tested for inter-observer reliability. This can be done in role-playing exercises or in real work situations. A test consists of two or more observers watching for a particular behavior and defining whether it is safe or at-risk.
If 90 percent agreement or better is not found after a series of tests, the behavioral definition (or standard) must be reviewed and refined by the work team, including the observers who used the definition to assess whether a certain standard was reached.
How am I doing?Media coverage of the U.S. Supreme Court’s involvement in the Florida election predicament emphasized that one Justice (Sandra Day O’Connor) questioned the need for a particular vote recount because the voting instructions were clear and straightforward. In other words, since the procedures for using the punch-card ballot were simple to follow, a partially-punched ballot should not be counted. If the voter can’t follow elementary directions for punching a ballot card, his or her vote shouldn’t count.
I heard various lawyers and political strategists for both Democrats and Republicans comment on this issue, and the conversation inevitably drifted back to standards. What is the standard for a valid vote? I never heard anyone mention the most important factor in reaching and maintaining a standard — feedback.
Lesson 3: People don’t know if they’ve met a certain standard unless they receive specific behavioral feedback. Feedback that a standard has been reached is rewarding and supports continued similar performance. Feedback indicating sub-standard performance sets the stage for corrective intervention and improvement.
Did voters know if they voted correctly? Did machines give participants inherent feedback regarding their compliance with the voting standards? Did voters know when they followed the instructions? If not, there is no way to know if a ballot count or recount reflects the voters’ intentions. Interpersonal feedback is an essential feature of behavior-based safety, and it’s critical to the prevention of injuries. After a CBC is used to reliably assess compliance with certain safety standards, behavior-based feedback is given to determine whether the standards are reached, and to discuss what can be done for continuous improvement. Individual feedback is delivered in one-to-one coaching sessions, and group feedback is presented at group meetings.
The importance of standards, reliable assessments, and feedback extend beyond safety, as the presidential election demonstrated. Perhaps you’ll teach these three lessons with a renewed sense of importance. And you might even be able to use your workplace experience to help reform our election practices. After all, those best suited to do this are those who experience these principles profoundly and repeatedly.