An Earned-Value Approach to Assess and Monitor Software Project Uncertainty: A Case Study in Software Test Execution

摘要

In this study, we proposed an earned-value approach to assess and monitor software uncertainty. It combines the value-at-risk method in financial field and the earned-value-feedback process of earned value project management and proposes a framework which contains value-at-uncertainty system, consumed-value-feedback system and experienced base to support in-process uncertainty measurement and contingency buffer management. Value-at-uncertainty system is used to assess the uncertainty in progress of a project and consumed-value-feedback system is applied to provide the decision support of how to deal with the buffer at that specific time. A case study in software test execution of 24 projects is conducted to evaluate the approach. The study shows how our approach works to measure the uncertainty and manage the buffer size in different project shapes. With the approach, both the accuracy and the effectiveness of uncertainty assessment can be improved along with the test execution progress.