A code inspection model for software quality management and prediction
Dennis A. Christenson, S.T. Huang · 2003
Theoretical considerations pertaining to the development and refinement of a code inspection model are presented. A code inspection model seeks to relate input parameters, such as the inspection effort, to the resulting quality of the code. The authors argue that finding a high or low density of errors in code yields little information about the quality of the code after the inspection has been completed. They suggest that a model which relates the amount of effort put into the inspection with the resulting fraction of errors found would be very useful. On the basis of the known code size, effort, and number of coding errors found, it would be possible to estimate the effectiveness of the inspection and the density of coding errors remaining in the code after the inspection was completed. Thus, some decision could be made about reinspecting the code or continuing on into the test phase.>