Harel Cain, 18/08/2011 07:43:
If the partial scholarships could somehow be defined, for example, to be 75% of the average flight cost from one's home to the conference, then those extra 25% (which could be USD 50 or USD 500....) will show one's seriousness.
The big advantage of fixed sums is that you don't have to bother about people choosing the cheapest/most sensible flight, accommodation and so on, which is almost impossible to do; but then you have to choose a sensible amount making some calculations before. The requirement to do so is having serious statistics on the costs faced by the attendees; this implies very complex surveys in university systems, but should be easier for us, we don't need to be as efficient as the Studentenwerk (although I don't get how the linked questionnaire can help for the next Wikimania[1]). Then you have to decide what share of the estimated cost you want to cover: if it's too low the incentive in null (only people who would have attended anyway get the scholarship) and the money is wasted, if it's too high the effectiveness decreases because you have less recipients and because we don't have a way to measure the commitment and results (so an unknown amount of money will be wasted). So far the amount of partial scholarships has been chosen to be good for most attendees to maximize participation, but we could also decide to keep the share more or less the same for everyone and give different (fixed) amounts to different areas of the world, if we think that everyone should have the same incentive no matter the cost. I'm not so sure of that: perhaps it's more effective to maximize the participation of "locals" and for Wikiania 2012 it's better to give 2-3 (partial) scholarships to wikimedians from Mexico (or a far USA state) than 1 to a wikimedian from Europe (random examples).
Nemo
[1] And for Wikimania 2011, it would be more interesting to know an objective piece of information such as the number of partial scholarship recipients who refused it because it wasn't enough.