In reply to Alan James - Rockfax:
I just think you need to be cautious about making inferences, as Coel suggested, about a wider population (2/3 of UKC readers are BMC members) from data that probably has quite high systematic biases resulting from the way the data was collected.
Don't get me wrong, it depends what you want to get out of the survey, if you are conducting it as a market research exercise, and want to know the priorities of your core user base for the purposes of developing the site and UKC as a business, then basing it on the people who use your site the most and can be bothered to fill in a survey seems an entirely reasonable thing to do.
My point is about the usage of statistics - the people who are coming to the stats purely for interest, should be cautious about drawing conclusions that the data does not support - without knowing more about whether there were any biases in the people who completed the survey (and it would be astonishing if there were not, ie. 86% of your responders were male) it is not particularly valid to draw firm conclusions about 'UKC readers' as a whole, or worse, 'UK climbers' despite the temptation to do so.