I'm worried about this, not because of 'skewing' the results- no survey in which the participants select themselves, so cannot be controlled in any way can be regarded as remotely scientific- but because it is bound to have been set up to support a dubious policy agenda of the AA and by filling it in we risk giving it false credibility by bumping up the sample size. Expect to see a press release attacking a presumed-liability straw man, claiming lots of support from cyclists as well as drivers...