We all have various ways of performing - let's call it "social masking" (in this particular context, at least) - and for some of us, some of those ways entail less stutter. Some of them entail more. It's a fickle thing and not something that has a universal pattern across stutterers. Many also report changes in the stutter frequency depending on their habituality with each such way of performing, or role - i.e. more stutter, once getting used to the role - or less, once becoming more settled in the role. To my limited knowledge, it's not something that is well understood in the context of stuttering, and I don't recall seeing any research in the area.