Error Abstraction Accuracy and Fixation during Error-Based Requirements Inspections

Software inspections are widely used as a requirements verification technique. Our research uses the tried-and-tested perspective of cognitive failures (i.e., human errors) to improve the effectiveness of fault detection during requirements inspections. We have previously shown that inspection effectiveness can be significantly improved by augmenting the current fault-based inspection technique with the proposed Error Abstraction and Inspection (supported by a Human Error Taxonomy). This paper investigates the impact of an inspector's ability to accurately abstract human errors on their fault-detection effectiveness.