Changes

* Since badges are opt-in, can you combine the data about who chose to accept a badge and who was eligible for one but didn't accept it? Could these two groups act as separate cohorts for a twin study?

==Gear==