The Longitudinal Analysis of Organizations: A Methodological Perspective

The case is made for performing more longitudinal research into organizations and their adaptive processes. Previous literature on the empirical longitudinal analysts of organizations is classified into five categories on the basis of the breadth of scope of the research, the number and diversity of organizations being studied, and the degree to which mathematical and statistical procedures are used to develop conclusions. The strengths and limitations of each of the five types of longitudinal research are discussed. It is argued that there should be a more concerted attempt to undertake research that focuses on a broad array of organizational variables in many organizations and which uses quantitative methods to derive results which are replicable and whose range of generality has been carefully established. Several proposals are made toward this end. Methods are suggested for gathering valid and reliable longitudinal multivariate data from numerous organizations in an economical fashion. Also, techniques of statistical and mathematical analysis are proposed for handling the complications caused by multivariate nonlinearities in longitudinal data. Of particular interest in this regard is the use of techniques of organizational taxonomy generation and testing and the use of algebraic heuristic procedures for identifying and differentiating among common scenarios of change in organizations. These suggestions should help to combat the non-cumulative nature of longitudinal research by avoiding the very common problems of specification error, nongenerality, and non-replicability.

• 저널Management Science · 28(9) · 1013–1034
• 토픽Cognitive Science and Mapping · Artificial Intelligence
• DOI10.1287/mnsc.28.9.1013
• 저자Danny Miller, Peter H. Friesen