@device(postscript) @libraryfile(Mathematics10) @libraryfile(Accents) @style(fontfamily=timesroman,fontscale=11) @pagefooting(immediate, left "@c", center "@c", right "@c") @heading(Software Metrics for Distributed Development) @heading(CMU-CS-96-190) @center(@b(Bernd Bruegge, Allen H. Dutoit@foot)) @center(November 1996) @center(FTP: CMU-CS-96-190.ps) @blankspace(1) @begin(text) In this paper, we present empirical evidence that metrics based on communication artifacts generated and recorded by groupware tools can be used to gain significant insight into the distributed development process that generated them. Moreover, we demonstrate that such communication metrics can provide better insights in the quality of processes and products than metrics based on outcome (e.g., function points, lines of code, words of documents). Most important, they can be uniformly applied throughout the process, independently of implementation technology, development infrastructure, or the existence of a product. We discuss our approach in developing and testing the use of communication metrics for distributed developments. We present a testbed for distributed development, a senior level software engineering project course at Carnegie Mellon University, in which we conducted several studies and experiments from 1991 to 1996 with more than 400 participants. Such a testbed represents an ideal environment for empirical software engineering, providing sufficient realism while allowing for controlled observation of important project parameters. We present three proof-of-concept experiments to illustrate the value of communication metrics in distributed and evolving development projects. Finally, we propose a statistical framework based on structural equations for testing hypotheses in an environment characterized by constant change. @blankspace(2line) @begin(transparent,size=10) @b(Keywords:@ )@c @end(transparent) @blankspace(1line) @end(text) @flushright(@b[(22 pages)])