The Galileo Group at Michigan State

In 1972-73, Woelfel and Barnett moved to the Department of Communication at Michigan State University, Woelfel as an Assistant Professor and Barnett as a graduate student. Of first priority was finding a replacement for SOUPAC, the University of Illinois statistical package which they used to make Galileo spaces. Kim Serota, a graduate student in Barnett’s cohort at MSU, began the task of combining the fragments of FORTRAN code written by Frank Belllinger and Woelfel at UI into a coherent program that could implement the Young-Householder-Torgerson solution and carry out the many other operations expected of the Galileo after the spaces were formed.
He soon found Richard A. Holmes, Jr., an undergraduate computer science major who quickly took over the task on a relatively full time basis, with funding entirely from private consulting, and remained the main programmer for the Galileo group until his untimely demise in 1991.



The development of a free standing FORTRAN program was of utmost importance to the Galileo enterprise, since none of the existing psychometric or statistical packages could do most of the things needed by the Galileo user community.
Holmes modified the Galileo software as required by its users. Much of this work was done for commercial employers, such as Market Dynamics, Inc, Opinion Dynamics, Inc., Galileo Opinion Dynamics International Inc.(GODI), The Galileo Company and others, but a good deal of programming was done pro bono for graduate student research projects.
John Marlier constructed an elaborate design to test hypotheses about Muzafer Sherif’s Social Judgment Theory. Until Marlier’s work, ratio scaled pair comparisons were considered to be unreliable for the individual case, and were used only as means of many cases. Marlier showed that they were, in contrast, more precise than traditional scales even for individual measurements.
George Barnett, Kim Serota and James Taylor investigated the use of Galileo for developing political campaign strategies. The mathematical analysis of multidimensional Riemannian vectors was done by Woelfel, presented for the general case in Woelfel, Holmes, Cody and Fink, and the code (called the Automatic Message Generator) was written by Holmes.