I agree that philanthropy and helping is really interesting, because I think in here we'll find many examples of informal learning occurring, through people mentoring, giving ad-hoc advice, acting as templates, etc. I'm so excited about this!

Using the categories provided by the respondents, and using multiple catgories would give a richer picture of the guild dynamics and purpose. Setting this in context with your research into what motivates people, you could then look at whether those multifaceted guilds had more motivated players. Also what types of guild foster motivation maybe?

As to order, when you list something there is an inherent priority – i.e. you usually list the most prominant things first. Would you want to weight categories according to order, unless the respondent suggests that it is an equal balance? And if you weighted it would certain categories come out as dominant, and how then does that link to motivation and learning?

Are you doing follow up interviews? If so one of the things might be interesting is that "philanthropy" or "helping" category. When people mention this does this mean contributing some gold to buy armour, crafting items for guild use, or is it the mentor/mentee relationship? Certainly a lot of initial help I recieved in game play, and still get is from more experienced "hard core" players who can advise on tactics, advise on best places to farm materials or even take you around instances/areas for first time in a hand holding manner whilst you become familiar with it.

I could add a category for "multiple purposes", but that wouldn't then say very much, would it, other than to differentiate between, as you say, multifaceted and single-purposed guilds. I'm not set on identifying a single purpose. The only reason I took that approach originally when developing the coding was because the question specifically asked for a primary purpose. As you can see from the responses, though, that's not what I got. That's partially my own fault because of the sample answer given. Mea culpa!

I think I can accomplish your distinction if I allow multiple codes for a given respondent. If there is only one code, then, in the mind of that respondent, their guild for them was primarily for one thing. If there is more than one, then it's more multi-faceted. That would still leave the questions of how many activities to code for (all of them?) and how to decide the order or importance (order they appear)? It has a more practical issue, too: how do I code and analyze multiple ones in SPSS? SPSS for this sort of thing isn't exactly my forté, so I'm not sure how to go about that or if it's possible.

Could you not add a coding category for 'Multiple Purposes'? This would avoid shoehorning lots of Guilds into categories where they don't quite fit – and it might help you to identify significant differences between single-minded Guilds and those that are more multifaceted.