The claim (ascribed to Albert Mehrabian) is much more limited than that. Those are the alleged percentage influences on how much we like a person on the basis o

The claim (ascribed to Albert Mehrabian) is much more limited than that. Those are the alleged percentage influences on how much we like a person on the basis of a conversation with them. See, e.g., http://www.kaaj.com/psych/smorder.html http://www.kaaj.com/psych/smorder.html (which I think is Mehrabian's own site) which says "Total Liking = 7% Verbal Liking + 38% Vocal Liking + 55% Facial Liking" and adds: "Unless a communicator is talking about their feelings or attitudes, these equations are not applicable". I'm frankly doubtful even about that much, but Mehrabian certainly never claimed that any such equation holds for more "contentful" communication.