Analyzing big data sets holds the promise of big insights. But the axiom "garbage in, garbage out" is particularly apt, since conclusions can be only as good as the raw data itself.
Even if we could develop a test or a screening process to find a pilot who would intentionally crash a plane, and that system was very, very good, virtually all positives would be false positives.
Doctor sick notes could be disclosed to employers without endangering confidentiality.
Cost cutting by commercial airlines has led to a worrying rise in inexperienced pilots.
It takes two people to launch a nuclear missile - it's time the same level of safety returned to the cockpit.
Tabloid assumptions that depression lay behind the Germanwings tragedy are damagingly premature.