Presents the results of a study aimed at investigating the antecedent software development factors that affect quality of final products. By monitoring those causal factors from the early phases of development, one can have a final product of enhanced quality and reduced costs.
The study considered an unprecedentedly large number of 30 C++ object‐oriented systems of varied size and application domains, a comprehensive suite of large number of predictive software design or code measures in one study, and compared their results on a common platform.
It was found that many of the software design or code measures have a significant positive or negative relationship with quality.
The value of the paper lies in the fact that it addresses some of the major problems from which most of the studies conducted in this research domain suffer. The objective and justification of this paper are to address these deficiencies, in addition to validating some of the results obtained in earlier studies. Another important value of the paper lies in the fact that, based on the results of the study, the paper enlists useful lessons learned that can provide some practical insight for practitioners and quality managers.
