Sergio Mauceri

Per una survey integrata e multilivello. Le lezioni dimenticate della Columbia School

The main idea expressed in this article refers to - and elaborates on - the contributions of the Bureau of Applied Social Research and of its mentor and founder Paul Lazarsfeld. It underlines the importance of how, in social research, it is necessary to develop and maintain a multilevel and integrated approach to surveys. Using sociometry and contextual analysis in the design of surveys, enables us to connect three levels of observation/study - individual, relational and contextual - which are often kept isolated and separate in social research. A standard approach to surveys creates casual samples of individuals, as if they were isolated units living in social emptiness, and limits itself to conduct data analysis that creates relations between individual variables. This integrated multilevel approach is instead the solution proposed to overcome the atomism and micro-sociological reductionism of this standard approach to surveys.

Ri-scoprire l’analisi dei casi devianti. Una strategia metodologica di supporto dei processi teorico-interpretativi nella ricerca sociale di tipo standard

Re-discovering deviant case analysis. A methodological strategy supporting theoretical-interpretative processes in standard social research - This article contains a reflection that intends to assert the heuristic fertility of integrating common procedural models with analysis of deviant cases. This analysis was originally proposed by Paul F. Lazarsfeld and promoted between the ’40s and the ’60s at Columbia School, without much impact on quantitative social research circles. Deviant cases are cases that do not display expected behaviour or attitudes. Instead of considering deviant cases irrelevant because they are residual, this procedural strategy accentuates their importance and promotes an in-depth study to maximize data quality and support interpretation processes of research results.