PROCEDURES FOR OFFICE ANALYSIS: A CRITICAL REVIEW

Abstract

Because office automation has not fulfilled its promise of making work more productive and satisfying, researchers have developed techniques for specifying better requirements for office automation and support. Four such office analysis techniques have been publicly proposed, differing in how much of the complete analysis-to-prescription cycle they cover, what aspects they analyze, and how they bound the "office." Review of these analysis processes points to three key issues: 1. Office analysis is weak on prescribing specific support/automation products; 2. we do not know how to evaluate different analysis techniques; and 3. we have not yet specified the criteria by which we would decide which technique is good. In answer to these issues, we suggest that extensions of some promising schemes for prescribing specific products be explored; that techniques be compared using an efficient "transcript experiment" approach; and that the criteria for acceptability for an analysis method be that its descriptions be reliable and valid, and that its prescriptions be valuable to the workers in the reorganized, computer-supported office.