Data obtained in a 1977 survey of 10 percent of the secondary school principals in the United States were subjected to statistical regression analysis to determine the relative effects of two types of school centralization. "Fragmentary" centralization involves the imposition of central control over specified facets of education in a number of settings, as is the case, for example, in federal categorical programs. "Integrative" centralization, implemented only at the state level, places all aspects of schools' and districts' educational functions under one central hierarchy. The analysis revealed that, as hypothesized on the basis of organizational theory, fragmentary centralization is accompanied by expansion of the administrative and macrotechnical (curricular program) structures of local schools. Integrative centralization, on the other hand, encourages simplification of these structures. This report discusses the theoretical background, hypothesis development, research strategy, data analysis, results, and conclusions of the study. The limitations of the original data are discussed in the report. An appendix presents the items included in the original questionnaire. (Author/PGD)