> It relied on self-report of the biker (and if I remember correctly, it was just a single one, so not enough data).
Unless I'm misunderstanding what's meant by "self-reporting", it sounds like regardless of how many riders they got data from, it would say that there are 0 fatalities to bikers because people who are dead don't respond to surveys.
The study was performed by Ian Walker. Self reporting means he was both the researcher and the cyclist - he rode his bike, reported the data, then analyzed it. There were no surveys or other riders. Unless you count "female Walker", when he wore a wig.
This is not a study of biker fatalities and my intuitive guess is they would be significantly higher for bikers without helmets for a number of reasons.
I agree that they would likely be higher for bikers without helments; I guess I misunderstood what the study was about, but my point was that if you relied on self-reporting for the number of fatalities, nobody would be able to report that they died for obvious reasons.
Unless I'm misunderstanding what's meant by "self-reporting", it sounds like regardless of how many riders they got data from, it would say that there are 0 fatalities to bikers because people who are dead don't respond to surveys.