Context Dependence and Across Group Variability: Their Relevance for the Choice of Indicators and of their Scalability

The problem of context dependence and across group variability is considered within the Rash model of items scaling. The model is theoretically intended to establish scaling results to be consistent across groups; but the feasibility o such a kind of across context and across group validity has to be implemented through methodological strategies which have to cope with: 1) software limits and reliability conditions; 2) with sociological conditions of across groups and context variability of items'meaning.