Investigating outliers

One of the most infamous stories of scientific oversight is the late discovery of the Ozone Hole. In 1985, scientists from the British Antarctic Survey published findings of abnormally low ozone levels above Antarctica. This led to the creation of the Montreal Protocol, in which 197 countries agreed to reduce emissions of substances that are harmful to the atmosphere.

This discovery came as a huge shock, because NASA had been monitoring ozone levels since the early 1970s and had not detected these findings earlier. It turns out, NASA’s satellite sensors had been calibrated according to predefined standards, so large deviations were flagged as errors and not properly investigated. Had these outliers been analysed, NASA would have ‘discovered’ the ozone hole several years earlier. Although NASA has since clarified that the situation was more complex than that, I think this story still teaches a valuable lesson.

As individuals, we bring our own lenses, biases, and perspectives to work. Perhaps we notice things that others don’t, or understand data in a different way. It can be easy to flag our perceptions as 'large deviations’ because they don’t fit a predefined norm. But while it’s important to be aware of how our lenses shape us, they aren’t necessarily a bad thing. Just because something is an outlier, doesn’t make it less worthy of investigation. On the contrary, outliers are valid, valuable, and maybe even world-altering.