A CASE-HISTORY ANALYSIS OF SOFTWARE ERROR CAUSE EFFECT RELATIONSHIPS

Software errors have been studied from various perspectives; however, most investigations have been limited to an individual section or a partial path of the cause-effect relationships of these errors. The present study analyzes approximately 700 errors in 4 commercial measuring-control software products, and then identifies the cause-effect relationships of errors occurring during software development. The analysis method used was: (i) defining appropriate observation points along the path leading from cause to effect of a software error, followed by gathering the corresponding data by analyzing each error using Fault Tree Analysis, and (ii) categorizing each observation point's data, and then summarizing the relationships between two adjoining points using a cross-indexing table. This paper presents four major cause-effect relationships and discusses the effects of the Structured Analysis and Structured Design methods on these relationships.