Some time ago we received a survey asking for our opinions on and data on how we use our behavior based safety observation program. We were assured that the safety department wanted our honest opinions and suggestions. I think I hurt the author’s feelings … he hasn’t talked to me since.
My concerns then, and now: 1.) By setting a quota of the number of cards submitted, it is likely that folks will just make stuff up to meet the quota, thus corrupting the data and invalidating any trends or conclusions that the office folks might draw from the aggregate observations; 2.) For crew members who participate in the JSA and pre-job planning process and have good attitudes and want to work safely, the sorts of things we are invited to identify as “at-risk” behaviors just don’t happen that much; 3.) The observation cards are primarily geared toward licensed officers observing unlicensed deck or engine room personnel and identifying behaviors that may result in hand or eye injuries, slips or falls, that sort of thing; nowhere is there space/context/invitation to observe officers in the decision-making, planning, execution of voyages or vessel moves that may result in incidents costliest to the company. 4.) What I said then was that a much greater emphasis on no-comeback reporting of near misses and better root cause analysis of actual incidents would better serve safety.
Questions: 1.) Why are we still using mid-20th century technology to identify and correct risky behaviors? Isn’t there a better way by now? 2.) Is this driven by ISM, the customers, insurance … what? 3.) Does anyone have experience of better methodologies?