Although most interesting problems are multivariate (MV) and students and citizens need to be able to reason using MV data, appropriate challenges are rarely encountered in class. In this paper we argue that the curriculum (and ideas about statistical literacy) should encompass reasoning with MV data. Statistics education can occur in a range of disciplines and in informal setting –notably on the web. Strategically, there is a need for dialogue with educators in other disciplines. We also argue for greater collaboration with data providers, who are engaging increasingly in ‘People Net’ (PN) activities–in short, a reconceptualisation of the education community. The SMART Centre has developed generic software shells that facilitate the import of MV data into interactive displays. We also describe their successful use in Citizenship classrooms where students interpret large-scale survey data on topics such as sexually transmitted infections and drug use.