Metrics to Develop High Quality Software

Objectives: The ultimate goal of the proposed metrics is to develop a quality software product, which is possible only when the product is certified at the end of each phase, so that there will be no place for errors and faults. Methods/Statistical Analysis: Project A and B are “School Management Software” in C++ language. The project A is developed using the “Software quality and productivity enhancement model” procedure and verified by metrics at every phase of development and maintenance whereas project B is developed in general procedure by another group. Both the software projects are observed for six months under similar work and conditions. Metrics are applied on observations from time to time and compared. Findings: Results from proposed metrics certify that undefined (hidden) task in product SRS or design phase constitute the major risks associated (product failure, quality, productivity etc.) with the product. Poor requirement identification and management laid greater role in product failure. It is certified in this paper that increase in the value of product failure % decrease the quality and productivity of product. The product whose failure rate is greater than 1% is risky to use, this value varies as per the product use and its working environment. Product effectiveness curve help the user in decision making regarding the working process of the software process, if it moves downward continue the use, but if it moves upward after a time period then the product should be abort or replaced. In this paper the application of proposed metrics on small project in different phases of its development, prerelease stage and in maintenance of the product enhance the quality and productivity of the product. Application/Improvements: The proposed metrics are applied on small projects; they can also be applied on large, complex software products. The application of the proposed metrics reduces the risk associated with the product that enhances the product quality