Abstract

A recent survey [4] shows that 35% to 50% of software projects had an overrun in terms of the development effort and schedule. Such late deliveries of software tend to cause project backlogs on the order of 374% ofcurrent capacity [2]. Many factors have been suggested as root causes for going over the budget, such as inaccurate estimates, inept scheduling, and failure to recognize risks and plan accordingly. Considerable effort has been devoted to the study of software project management techniques to cure these problems [1]. On the other hand, some suggest that we should seek ways to make programmers more productive, just as Henry Ford made auto assembly workers more productive. For instance, Boehm [3] believes that there are opportunities to increase software productivity, and predicts that by 1995 a 20% improvement in software productivity will be worth US$45 billion in the US and US$90 billion world-wide. As a result, new methodologies and tools for software development have been suggested and developed in order to increase productivity and subsequently alleviate project backlogs. Representative examples include structured analysis and design, computer-aided software engineering (CASE), and fourth generation languages (4GLs). However, the impacts of these new methodologies and tools on productivity have not been clearly realized, and their use is often controversial [8]. Previous research [6,7] suggests that improving programming productivity requires much more than the isolated implementation of new technologies and policies. Unfortunately, the integration of the new technologies with strategies for managing the software development process, such as project and configuration management, characterization of project teams, and process control, has been largely overlooked [5] as a critical contributor to software productivity improvement. Thus, an integrated analysis, which covers both technological and managerial aspects of software development, is needed. Since there has been almost no research of this problem, we began with an exploratory study [9] that examined current practices in application systems development, focusing on utilization of productivity tools in information systems (I/S) organizations. The aim ofthe study was to develop an appreciation of the factors that might affect application development productivity. We found that an integration of technical and managerial interventions accounted for increased productivity. In the present study, we follow up on the exploratory study. Using hypotheses developed in [9], we are collecting data from a large sample of companies through a mail survey. Before proceeding to a detailed description of the research in progress, we present some concepts basic to understanding our research approach.

Share

COinS