In psychosocial field numbers come as counts, scores and ranks (raw score more usually) are often mistaken for measures and then misused in statistical analyses for which they are quite unsuited. Some of this kind of number are merely labels. When we use numbers to compare outcomes, we do arithmetic with them: arithmetic done with numerical labels makes nonsense. For corrects data analysis, it must be done with equal interval, constant unit and linear measures. This paper report how we can do (in psychosocial field) to transform raw score in fundamental measurement.