ABSTRACT

This article begins by examining the recommendations of Uhl-Bien and Marion that the LAMPE theory of organizational leadership could be enhanced and improved if it were to incorporate elements of complexity and complexity leadership theory. Their advice should be reversed: complexity leadership theory should incorporate the theory, methods, and models already tested in the construct of the LAMPE theory.

The reasons for this conclusion is based on a general discussion of the conditions under which a processual theory can be tested and the testing procedure be made rigorous. A new idea called a process skeleton is introduced. Process skeletons can be enhanced into theoretical analytic models and theoretical analytic models can be reduced to their process skeletons. The theoretical analytic model is often not the model actually tested. Reducing the model actually tested and the theoretical analytic model ideally should yield the same process skeleton. This discussion then delves into the issue of aggregation of data. It is shown that because different groups often have different processes and the same groups can have different processes at different times, etc., the practice of aggregating data to fitinto a statistical model is extremely dubious. Finally, the requirements of testing a multi-level processual model are discussed. It turns out that processes and multi-level processes have a common mathematical representation. According to this, complexity leadership theory cannot be most rigorous and the LAMPE organizational leadership theory might be.