Toward a Theory of Assurance Case Confidence
September 2012 • Technical Report
In this report, the authors present a framework for thinking about confidence in assurance case arguments.
Software Engineering Institute
CMU/SEI Report Number
Assurance cases provide an argument and evidence explaining why a claim about some system property holds. This report outlines a framework for justifying confidence in the truth of such an assurance case claim. The framework is based on the notion of eliminative induction—the principle first put forward by Francis Bacon that confidence in the truth of a hypothesis or claim increases as reasons for doubting its truth are identified and eliminated. Possible reasons for doubting the truth of a claim arise from analyzing an assurance case using defeasible reasoning concepts. Finally, the notion of Baconian probability provides a measure of confidence based on how many defeaters have been identified and eliminated.