When comparing results from different populations there is an underlying assumption that the response patterns elicited by the instrument used are stabile across the study groups. But it remains an open question whether this assumption is correct or not--especially in the case when the populations are involved to a different degree into the problem which is the object of the investigation. In fact, we are able to show that the dimensional structure of an instrument developed for the evaluation of stigma psychiatric patients are exposed to, varies depending on the degree of personal involvement with psychiatric disorder. This leads us to the conclusion that a quantitative comparison of parameters appears not warranted as long as the dimensional stability of the instrument over the various populations has not been shown.