Comparaisons diachroniques et substancialisation des variables : exemple de l’évolution des inégalités scolaires

摘要

The evolution of educational inequalities is difficult to measure. It is necessary to build an indicator of inequality but also to take into account the social changes which affect the social classes and the value of the diplomas. Whichever the statistical tool used, there is always “quantitative” democratization. It is not the same with “qualitative” democratization. We propose to analyze it by building a latent variable called “educational potential”. We eliminate “quantitative” democratization by giving a relative value to the diplomas. The study of 19 triennial generations between 1920 and 1976 shows that “qualitative” democratization relates mainly to the children of the agricultural categories. This result is similar to that obtained with the log-multiplicative layer effect model. On the whole, can we speak about a “qualitative” democratization?