I started a survey a while ago to find out what factors people consider important when they are giving their opinion on requests for adminship, it's had some good responses so far and I'm announcing it again here to involve a wider audience:
http://en.wikipedia.org/wiki/Wikipedia:Requests_for_adminship/Factors
The responses so far have been interesting. There are some trends developing:
* some factors are broadly considered important by most people (eg, candidate's history of interaction with other users),
* some factors are not considered important on their own, but can be good indicators of other important factors (eg, involvement with collaborative efforts - not important as a factor but a useful indicator of interaction with other users), and
* some factors are not considered important at all (eg, whether candidate has demonstrated a "need" for the tools).