Because we need to define "dangerous" - the authors and you are making an assumption that what they call "inattention" is the same as "dangerous" whereas the co

Because we need to define "dangerous" - the authors and you are making an assumption that what they call "inattention" is the same as "dangerous" whereas the comment you are replying to is questioning that assumption. If the link between "inattention" and accident rates is strong, then measuring that should be equivalent. In medicine there is a concept of measuring disease-oriented outcomes instead of patient-oriented outcomes, because it's easier to show an intervention affects the disease-oriented outcomes, and much harder to show a change in patient-oriented outcomes. So for example, a disease-oriented ourcome for hip-replacement surgery would be "range of movement after 2 weeks", patient-oriented might be "can go up a flight of stairs" The analogy here seems relevant.