Using the Fault Index Selection Process for Large Software Systems to Guide and Improve Code Inspection Effectiveness
Oakes, Steven C.
Software inspections are one of the most powerful error detection techniques in software development. Traditionally, experienced software inspectors read the source code line-by-line, report the errors as they find them, and document comments and suggestions. Selecting code candidates for an inspection is a challenging and difficult task for the inspection teams due to the enormous size of a software system. It is practically impossible to inspect all of the code. Therefore, a surrogate measure to predict the amount of software errors that are in the system is needed and this measure is called the Fault Index. The Fault Index is a single numeric value calculated from statistical analysis of the variability of source code metrics. Utilizing a Fault Index to select code candidates prior to and during the inspection process increases the effectiveness of the inspection. In addition, incorporating the Fault Index to select code candidates is more efficient than inspectors' opinion. Finally, it is up to the inspection team members to inspect and remove errors from those potentially fault-prone components.
Software maintenance, Computer industry, Software failures, Software engineering, Quality control
Oakes, S. C. (2005). <i>Using the fault index selection process for large software systems to guide and improve code inspection effectiveness</i> (Unpublished thesis). Texas State University-San Marcos, San Marcos, Texas.