Abstract: Theoretical developments in software quality could be greatly enhanced if a theoretical definition of software quality existed. Content analysis is used here to objectively identify the dimensions of software quality. The dimensions are derived from the perspectives of both producers and consumers. Results show that the most important software quality dimensions include accuracy, capability, communication, completeness, conformance, features, flexibility, serviceability, simplicity, stability, and structuredness. Consumer and producer attitudes about software quality only slightly differ. Further analysis suggests some possible meta-dimensions of software quality: fulfilling customer requirements, ease of use, and capability (for producers of software); and fulfilling customer requirements, ease of use, and first impression quality (for consumers of software). Consumers appear to be primarily interested in what the software can do for them now, whereas producers appear to be primarily interested in what the software can do for them in the future. The steps involved in the identification of significant dimensions are detailed, and the implications to practice and further research are discussed.
Key words: attitude measurement theory, content analysis, principal component analysis, product reviews, quality dimensions.