No abstract

This paper describes analysis techniques used to quantitatively assess the software maintenance process of a large military contractor, and the results obtained. The analysis techniques make use of basic data collected throughout the maintenance process. The data collected are extensive and allow a set of functional enhancements to be traced to r-x?; activities and product impact. * nonparametric statistical techniques are then applied to test relationships between data items. The results provide valuable information for predicting process and product charaCteristics, and assessing the maintenance process.

SUMMARYThis paper describes the measurement-based software maintenance process defined and implemented at Lockheed-Martin, Moorestown, NJ. The documented process includes extensive data collection, a tightly controlled but highly accessible database, data analysis techniques supported by software tools, and process assessment and improvement activities. The methods and techniques used are presented in a 'how to' fashion so that other organizations can leverage our efforts to define and implement a measurement-based process of their own. Our approach is an evolutionary one, rather than a revolutionary organizational upheaval. We describe the benefits gained from our process, including statistically validated metric results, and the subsequent process improvements implemented. This paper describes solutions to the 'real-world' issues faced by an organization which successfully implemented a measurement-based software maintenance process.

This paper presents a quantitative comparison of perfective and corrective software maintenance performed by a large military contractor using a formal program release process. The analysis techniques used in the comparison make use of basic data collected throughout the maintenance process. The data collected allow the impact of performing perfective and corrective maintenance to be quantitatively compared. Both parametric and non‐parametric statistical techniques are applied to test relationships between and among process and product data. The results provide valuable information for predicting future process and product characteristics, assessing perfective and corrective maintenance impact, and quantitatively comparing the impact of both types of requirements volatility. The results also support one common rule of thumb, cast some doubt on another, and lead to the formulation of a new one. © 1997 John Wiley & Sons, Ltd.

No abstract