Measuring Peer Effects in a Network Perspective. Survey Design and Privacy Issues

The present contribution discusses a strategy to address methodological and ethical challenges in network survey designs. The approach seeks to balance statistical rigor with the protection of partic-ipants’ privacy. Its applicability is demonstrated through a case study involving high school students in Southern Italy aiming at exploring the factors related to their post-diploma educational choices. Recognising the role of peers’ influence, the study adopts a statistical model that includes the students’ relationships within a class, i.e. the class adjacency matrix, considering different kinds of interactions among classmates. All students in the classes selected through a quota sampling procedure are included in a whole network study, raising important ethical and privacy con-cerns. To address these issues, anonymisation protocols are implemented to ensure data protection and management.