Consider the last software application you debugged, enhanced, modified, or upgraded. Analyze it following the paradigm that is used for Kanchana's example, and construct the appropriate orthogonal array for the statistical experiment.
2.
Define a suitable signal-to-noise ratio for your problem, and perform the statistical experiment to rank the error-prone modules. Now compare the results of this formal statistical paradigm with your intuitive experience. Do you arrive at the same result? If not, why not?
