In reply to Ericrd:
I would imagine that most people that have been involved in accidents have been involved in more than one. So you need to specify "your first accident" or "your worst accident" depending on what you want to study.
You also do not make any distinctions about what really classifys as an "accident". Is an injury necessary for you to consider it an accident? Different respondents are going to be reporting their reactions to totally different incidents as it is set up currently; one person might be talking about how they fell off on the lead, sprained their ankle and then limped off home for some r&r, someone else might be takling about how they pulled of a giant boulder, killed their partner and were left disabled by the ground fall. The before-and-after for these situations is completely different and yet you will be considering them all together with no info differentiating them from each other.
There is a HUGE difference between an accident where you personally were injured and one where your partner was injured and one where the dude climbing next to you on the cliff was injured. I strongly suspect that the effect of the accident on the thought processes of a climber has a huge amount to do with who the accident happens to.
I tried to complete your questionaire but I'm not even sure when the first thing that would classify as an "accident" happened and I would be wildly guessing about how confident I felt about different aspects of climbing either before or after an accident that happened 5+ years ago.
There are obviously many many more variables to take into account but I'll stop listing them. My point would be that while I don't know exactly what you are hoping to show with your survey, I can't imagine how the diverse set of situations you are analysing, coupled with the completely subjective, and in most cases retrospectively guessed, answers can lead to any valid conclusion.
I think you would get more useful information by disecting climbing essays about specific accidents (and god knows, there is no shortage of them) and then contacting the authors and asking if they would mind replying to you with an analysis of how their personal feelings about risk changed as a result of the accident (assuming they can remember). This approach still has plenty of flaws but would surely yield more useful information.
Though it would require quite a bit more time and effort than making an online survey