In the last years, and more and more frequently, people are going over the graphic analysis as a favourable vehicle for obtaining statistical studies in situations in wich measure objects make difficult a direct appreciation. The graphic analysis, with a great tradition in the social science ambit, allows, on one hand, the consideration from the information less inflexible than with the application of the classical statistical technics and, on the other hand, the incorporation of new performance forms that are not close to inflexible statistical judgment. In that way the classical evaluation instruments in the ambit of the Environmental Psychology (P01, 1989) are build for incorporating different factors entailed with the measure object. In that writting will be analized experimentally a questionary referred to the quality of life in the city, defined and based on the inclusion of different topics connected with it (apartments, services,...). That parciality forced to analize each one of those appreciation realities for getting the best measure instrument with wich it could be established the interval structure of the questionary. (Pol et Als., 1991). The different technics that allow that reduction are majority searched, in the consideration of the information as interval scale, being, really, common to work with nominal and ordinal scale.So, here, it is proposed to explain the use of technics that look for that situation, preferentialy the Multiple Correspondence Analysis (Lebart, Fenelon, Morineau, 1985) and the Facet's theory (Canter, 1985), for the purifying of -the cited questionary and reexpound it down from the results.