Estimation Under Uncertainty

Abstract

Uncertainty and inaccuracy are inherent properties of estimation, in particular predictive estimation. Measuring and managing uncertainty and inaccuracy lies at the heart of good estimation. Flyvbjerg (2006) distinguished three categories of reasons for inaccuracies in project forecasts:

Further Reading

• S. French (1995), “Uncertainty and imprecision: Modeling and analysis,” Journal of the Operational Research Society, vol. 46, pp. 70–79.

  The author provides a general discussion of uncertainty and imprecision in the context of statistics and decision making. He identifies potential sources of uncertainty and their impact on analysis. Furthermore, the author discusses possible ways of modeling and methods for analyzing uncertainty.

• B. Kitchenham and S. Linkman (1997), “Estimates, uncertainty, and risk,” IEEE Software, vol. 14, no. 3, pp. 69–74.

  In this article, the authors discuss in detail the potential sources of uncertainty in the context of software project estimation.

• M. Jørgensen and S. Grimstad (2008), “Avoiding Irrelevant and Misleading Information When Estimating Development Effort,” IEEE Software, vol. 25, no. 3, pp. 78–83.

  In this article, the authors discuss the sources of uncertainty in the context of expert-based effort estimation. In particular, they discuss the impact on estimates if experts provide irrelevant and misleading information.

• S. Grimstad and M. Jørgensen (2006), “A framework for the analysis of software cost estimation accuracy”, Proceedings of the International Symposium on Empirical Software Engineering, Rio de Janeiro, Brazil, 2006, pp. 58–65.

  Based upon an analysis of effort estimation studies, the authors propose a framework for analyzing effort estimation effort. Within the framework, they propose a list of typical sources of estimation error.

• T. Little (2006), “Schedule Estimation and Uncertainty Surrounding the Cone of Uncertainty.” IEEE Software, vol. 23, no. 3, pp. 48–54.

  The author discusses aspects of estimation uncertainty (here schedule estimation) based on example industrial data.

• Z. Xu and T.M. Khoshgoftaar (2004), “Identification of Fuzzy Models of Software Cost Estimation,” Fuzzy Sets and Systems, vol. 145, no. 1, pp. 141–163.

  Authors describe a fuzzy effort modeling technique that deals with linguistic data and uses historical project data for automatically generating fuzzy membership functions and fuzzy rules.

• J. Li and G. Ruhe (2006), “A Comparative Study of Attribute Weighting Heuristics for Effort Estimation by Analogy,” Proceedings of the International Symposium on Empirical Software Engineering, Rio de Janeiro, Brazil, pp. 66–74.

  Authors compare in an empirical study several techniques for quantifying the impact of effort drivers on development effort. Investigated techniques are used in the context of an analogy-based estimation method called AQUA+. One of the techniques they employ for modeling the uncertain impact of effort drivers is based on rough sets theory.

• S. McConnell (2006), Software Estimation: Demystifying the Black Art, Microsoft Press.

  In Chap. 4 of his book, McConnell considers several ways of addressing estimation uncertainty. He provides several means for narrowing the cone of uncertainty and for considering the remaining uncertainty in project estimates, targets, and commitments.

• A. Stellman and J. Greene (2005), Applied Software Project Management, O’Reilly Media.

  In Chap. 9 of their book, the authors discuss project change. They investigate the most common reasons of failed project changes and provide guidelines on how to successfully run projects through change. Additionally, in Sect. 6.4, the authors discuss a change control process for dealing with changes in requirements. They highlight the need for implementing only those changes that are worth pursuing and for preventing unnecessary or overly costly changes. In practice, many changes that apparently look like good ideas ultimately get thrown out once the true cost-benefit ratio of the change is known.

• E.T. Wang, P. Ju, J.J. Jiang, and G. Klein (2008), “The effects of change control and management review on software flexibility and project performance,” Information & Management, vol. 45, no. 7, pp. 438–443.

  This article discusses the link between project performance, software flexibility, and management interventions. In particular, they investigate software flexibility as a mediator between project performance and two commonly recommended management control mechanisms, management review and change control. In their survey among over 200 project managers from the Project Management Institute, the authors confirmed that the level of control activities during software development is a significant facilitator of software flexibility, which, in turn, improves project performance.