Applied Statistics
International Conference, 19 - 22 September 2004
Ljubljana, Slovenia


A Modest Agenda of Immodest Goals
for Network Analysts

Patrick Doreian
University of Pittsburgh
USA

Over the years, network analysts have accomplished many things with creative and inventive solutions to many technical problems. The field has progressed in an impressive fashion, so much so that some have suggested it is a scientific revolution for the social sciences. This seems a debatable claim. Until we can solve problems that other scholars cannot solve, the claim will be hollow. A suggested list of such problems is this:

  1. test statistically the fit of blockmodels in large networks;
  2. model the evolution of such fundamental network structures;
  3. incorporate network autocorrelation to a wide variety of statistical methods in a dynamic context to predict actor behavior;
  4. incorporate massive shocks to network systems analytically; and
  5. build multiple-equation methods for the study of whole networks.
At a minimum, solving these problems would create the necessary conditions for claiming a revolutionary status for social network analysis - if we dare to make predictions ahead of phenomena to create the conditions for a genuine test of our methods. Adopting this mind-set might be the hardest task of all!