- 25/7/08
- 6,358
- 28
- 0
Anne-Marie, I wouldn't mind running more of these and working on refining the method. And to do that I'd like to get everyone's input too. I am also aware that whatever is produced will not make everyone happy, especially when the results don't match personal expectations. But believe it or not, these things are not too inaccurate and do paint a fairly good picture. Think of it as an impressionist's painting, where short and thick strokes are used to quickly capture the essence of the subject rather than its details. Responses to this first poll have been mostly positive, so I'm inclined to think that a rating system is worth the time and effort...
My initial intended audience was more the dealer community rather than forum members, and I wonder whether perhaps, at least to start with, results of such polls should be shared only with the dealers themselves. The system I used was designed after all for sellers to assess the strength of their customer base and act on it, and not necessarily for us to assess them.
Anyway, these things are time consuming, and frankly I couldn't be bothered doing anything like this again if once it's up I have to deal with a bunch of time-wasters who type before reading, don't like what the results are saying, fail to see the strengths and weaknesses of the thing, or act as bloody experts on non response bias, sampling techniques, measurement validation or any other "Stats 101" nonsense. In the end, these are always going to be crude instruments, leading to crude results, on a friggen rep watch forum! Still, if there is enough interest I'd be happy to work on a new and more sophisticated poll-based system.
My initial intended audience was more the dealer community rather than forum members, and I wonder whether perhaps, at least to start with, results of such polls should be shared only with the dealers themselves. The system I used was designed after all for sellers to assess the strength of their customer base and act on it, and not necessarily for us to assess them.
Anyway, these things are time consuming, and frankly I couldn't be bothered doing anything like this again if once it's up I have to deal with a bunch of time-wasters who type before reading, don't like what the results are saying, fail to see the strengths and weaknesses of the thing, or act as bloody experts on non response bias, sampling techniques, measurement validation or any other "Stats 101" nonsense. In the end, these are always going to be crude instruments, leading to crude results, on a friggen rep watch forum! Still, if there is enough interest I'd be happy to work on a new and more sophisticated poll-based system.