In this dissertation, we have used Bayesian networks to model uncertainty in software development. Alternative approaches to uncertainty modeling and logical inference exist, however, that must be weighed and compared against Bayesian methods. Those include Certainty-Factor approaches, Dempster-Shafer approaches, fuzzy logic, and default, monotonic, and nonmonotonic logic [Ste95]. Specifically, relative merits and pitfalls of these techniques should be evaluated against the Bayesian approach in the context of software engineering situations.