you're going to run into a couple problems -- 1) Intoxication produces more than one change (thus you have inextricably linked dependent variables): alcohol impairs motor control (particularly coordination) and affects the person's approach / avoidance in social situations. Your task needs to clearly separate these (I know of no way) 2) We already know that stuttering severity is influenced by "social anxiety." By successively increasing audience size stuttering severity increases. 3) Similarly we know that the propositionality of the text can also influence stuttering severity (e.g., when asked to talk about a complex, unfamiliar topic stuttering severity is increased compared to a simple or well-known topic). Also, repeated readings of the same text results in a decrease in stuttering severity (the adaptation affect) 4) It's unclear to me how you're defining stuttering "aside from the general alcohol effects" Stuttering is a speech-motor disorder. Any coincident change in speech motor planning / execution produced by your task will make your measure invalid. 5) what do you mean "different kind of stutters"? Different types of disfluencies? Are you looking for subgroups within the stuttering population? 6) What do you mean "language changes"? 7) Experience drinking alcohol and sensitivity to alcohol are going to be lurking variables. You'll need to find someway to account for this. 8) Temperament is also going to be a variable to consider--not all people who stutter are "socially anxious" 9) The situation you describe includes more than just alcohol and social anxiety. You've got background noise (stuttering severity can decrease when speaking in the presence of noise), time (is it alcohol or experience in that environment that produced the change). So, all in all, you need to more clearly define your question. (and figure out how to convince an IRB to let you give adults alcohol. I'm sure that can be done, but it may be an up hill battle).