In Psycho-Social studies, the semantic differential is one of the most used measurement instruments. It is used in order to collect data from persons answering the items of tests and questionnaires. There are many advantage in the use of this kind of tool. On the other hand the problem of the indefensible equivalence among the semantic differential intervals arises. This paper suggests a solution: through a random sample experiment (Montecarlo method) it is possible to estimate how the achieved results change in relation to variations of its intervals, by using a standard semantic differential. If the difference among intervals is not more that ± 10%, the semantic differential can be considered a good appraisal instrument.