This is meaningless without comparing to non-developers, or the general population. For all we know from this data, developers may like beer less than non-developers. Combine this with the problems cited by jmduke, and I'm not sure it is enhancing the reputation of the "developer focused industry analyst firm" that published it. Yes, I have no sense of humor.