ABSTRACT
This chapter explores several early experimental situations for testing a theory of group structures. It shows that group structures ought to change and that how they changed would depend in part on the nature of the problem being solved. The strong inference tree has its base in the Faucheux and Mackenzie experiment. The two Berkeley experiments provide additional branches and twigs on this tree. Despite the shortcomings of the original design, the main results and the secondary analyses of the data using tools developed after 1966 are theoretically relevant. They constitute a core of basic results about structural change. The closer reasoning about why these results occurred and the more detailed analyses and implications of structural change came much later.
