The Stage Hypothesis and Data Administration: Some Contradictory Evidence

Nolan's State Hypothesis on the assimilation of computer technology by organizations provides one of the most popular frameworks for describing and managing the growth in corporate data processing. The model has achieved a high level of acceptance despite little formal evidence of its reliability or robustness. One previously published test was unable to con firm the predicted S-shared growth curve for EDP budgets. Our study of 273 large Canadian organizations tested another prediction of the Stage Hypothesis--that more "mature" DP groups would be more likely to have a formal Data Administration function than less "mature ones--and fai led to find the hypothesized relationship. Further analysis of the results revealed that the var iables used to assess DP maturity do not exhibit a sufficient degree of intercorrelation for them to be considered aspects of a common construct. These results cast considerable doubt on a basic premise of the Stage Hypothesis. INTRODUCTION In this environment, the introduction of the Stage Hypothesis served several useful Nolan's Stage Hypothesis ( 1973) on the purposes. First, it claimed to explain, and assimilation of computer technology by therefore to a certain extent, to legitiorganizations has attracted much interest mize, the rapid growth in DP expenditures from both practitioners and academics that had been experienced. In addition, it since its introduction in 1973. At the time held out the welcome promise that the rate of its initial publication, data processing of growth could be expected to decl ine in departments had been experiencing rapidly the future. Perhaps even more importantgrowing budgets for a decade or more, ly, the hypothesis provided a model of the while the great promised benefi ts of comprocess of corporate computer ization as puterization had proven disappointingly proceeding from Initiation through Maturelusive. This often led to a disenchantity via a sequence of four, and later six, ment with computers on the part of specific stages. Each stage is charactermanagers outside the data processing area, ized by the types of computer applications and a retaliatory claim from within the in use, the organization and management computer department that they were of the data processing department, and its somehow "different" from other corporate relationship with the rest of the corporactivities and could not be properly ation. Both DP and non-DP managers can managed or evaluated using traditional use the model to reassure themselves, and methods. to demonstrate to others that what they