-
Yep, agree with all that - that's why I said it'd probably be more useful in the forum rider down thread than in the general forum itself, as forum rider down tends to be self-reporting, not people making reports for other people.
bothwell-
Actually that's exactly what I thought you said.
But the Rider Down forum isn't self-reported, and the thread contained therein has far too small a sample of actual reports (a minority of the forum users use it) to represent any meaningful interpretation of the data.
Both fail, for different reasons.
If there was a way to have all UK cycling fora talk to each other, along with the major cycling publications, to support a large survey... then I'd support it 100%. But a statistically meaningless sample of biased data from a biased group is worse than useless. If any data scientist or statisician argues otherwise, they're basically bad at their job.
Velocio
The big problem with that proposal is that we seldom get those involved in the incidents coming forward to share their story. We rely predominantly on third party citations and verifiable reports of road traffic incidents to record incidents within Rider Down.
Given that... how would we manage to get those involved to record their injuries anywhere?
Those involved in near misses never even bother visiting A&E, let alone a forum to record their minor scrape, and those involved in more serious incidents go to A&E first and very few come to any internet forum to share the details.
It's somewhat worrying that the very best source of this information... the hospitals themselves... are not able to record it. But I really don't think that even with the best efforts from LFGSS or other fora that we will be able to get people to record their injuries. Hell, I don't even think people record potholes and that's at the level of single click app nowadays.
I'll sticky something if people really want, but in reality I think the chances of it getting more than a couple of reports is close to non-existant and would make any data submitted act as statistical anomalies that if normalised would harm, rather than help, any eventual case.
Yes, they don't have the data... but they're not going to be able to project that data out of an incomplete and biased sampling either, and if they do they're going to led to conclusions not aligned with reality which could be more damaging than good.