*** 9661631 Heimberger Software development and maintenance are a consequential part of current business activity in many sectors of the economy activity. Unfortunately, the current state of statistical practice in software engineering is immature because researchers and practitioners have a limited knowledge of statistics. Compounding this problem is that the applicability of methods is not always systematically demonstrated and that methods are not usually presented in a context that is familiar to engineers and managers. This Small Business Innovation Research Phase I project to be carried out by Software Productivity Solutions addresses these problems by developing an innovative quantitative methodology and supporting tool for Decision Support for Managing Performance Risk (DSMPR) in software projects. DSMPR will demonstrate the value of applying statistical and decision analysis techniques to a well-defined software management problem and provide a methodology that describes how they should be applied in the software management process. DSMPR will accomplish its tasks by integrating selected techniques from Forecasting and Multi-Attribute Utility Theory into a methodology and tool for software project management. The result will be a significant improvement in the quality of insight provided to software managers for decisionmaking. DSMPR has a strong commercial potential for satisfying the needs of a broad and largely untapped market. DSMPR will help software project managers to assess the significance of differences between plans and actual performance; quantify the performance risk represented by the deviations; and assist in making tradeoffs in planning corrective actions. By providing such project management and decision-support capabilities, DSMPR will have a broad applicability to both commercial and Government industry sectors where large software projects are common. These sectors include telecommunications, transportation, and distribution. *** .
