Safety Data: Signal and Noise Continued from page 47 L eaders must have conviction. They often know, in their gut, when something is the right move. That said, even the best leaders can’t see the future with certainty; and, having additional information to support decision-making is a valu- able tool. Peter Drucker once wrote: “What gets measured gets managed.” Though that may be generally true, not everything we can measure is worth spending time managing. Data is critical to supplementing “gut” decision-making; but hav- ing the right data at the right time is even more important, especially when it comes to safety. Safety data is an important element of effective safety management, but it must be used in the right context to translate into safety intelligence. In this short article, we’ll review just what con- stitutes data, how it can be used to better inform aviation safety decision- making, and why using safety data on a widespread basis continues to challenge our industry today. What constitutes data? Simply, data is a source of, or even the prod- uct of, compiling information. Data may include metrics, measurements and observations. Data surrounds us, and modern technology enables us to have massive amounts of data at our fingertips at any time. Data is the ingredient that allows statisti- cal information to be constructed, and it falls into two broad catego- ries: quantitative and qualitative. In practice, many people have a prefer- ence for quantitative data because it expresses information in terms that seem more concrete, i.e. count, percent, temperature, cost, rate, and so on. Quantitative data is great, but it often doesn’t have context embedded. Qualitative data—gener- ally categories, descriptions or plain text—may be a bit tougher to wrap our heads around, but the context and fluidity of categorical informa- tion means that qualitative data can provide tremendous insight alone or when combined with quantitative figures. The catch is that analysis can be a bit more time consuming. In practice, most of us are sitting on quite a bit of data already, though we may have to step back from daily operations to identify what it might be trying to tell us. Common quantitative data for aviation ground service providers might include metrics like: ■ ■ ■■ ■■ ■■ ■■ Gallons of fuel sold over a given time frame Aircraft movements Employee tenure Employee turnover rate Revenue and profit (and other P&L metrics) ■■ ■■ Injuries, accidents and incidents Safety reporting rate Data is a source of, or even the product of, compiling information. Aviation Business Journal | 3rd Quarter 2017 Of course, this list isn’t exhaustive, and you’ve probably noticed a few differences between the items listed. For instance, some are related to eco- nomic factors while others focus more on safety-specific areas of interest. Some are outcome-based, but others look at precursors to those outcomes. Those differences are important to good data analysis, because they provide diversity of information— helping to reduce bias. The differ- ence between data sources that are outcome-focused and those that are input-oriented is that outcome mea- sures—often called reactive, lagging, or backward-looking—are important but cannot be changed by the time we have observed them. Naturally, all we can do is report on them and attempt to achieve different future outcomes. Input measures, also called leading, predictive, or forward- looking, are those things that we can understand and modify in real time before some event of interest actually occurs. Balancing those perspectives is also critical to supporting decision- making with data because it ensures we have appropriate navigational information going forward while making certain we inform future decisions with lessons from our past. Continued on page 51 49
