Analysis of Bug Fixing Processes Using Program Slicing Metrics

This paper is a report of a feasibility study into an alternative assessment of software processes at the micro-level. Using the novel approach of applying program slicing metrics to identify bug characteristics, the research studied relationships between bug characteristics and their bug fixing processes. The results suggested that specific characteristics such as cyclomatic complexity may relate to how long it takes to fix a bug. Results serve as a proof of concept and a starting point for this proposed assessment methodology. Future refinement of the metrics and much larger sample data is needed. This research is an initial step in the development of assessment tools to assist with Software Process Improvement.