Exercises

Internet Exercises

1. Find Internet sites that deal with the quality loss function. Now limit the results to sites that deal with it in a software product context.
2. Find Internet sites that refer to the Taguchi seven-step method.

Review Questions

1. Describe the quality loss function and its application.
2. Review and explain the seven steps in a Taguchi analysis.
3. Which steps of this analysis may be iterated to gain a design optimization?
4. What is the contribution of “orthogonality” in Taguchi’s use of orthogonal arrays?
5. Review the statistical terms used in this article, and try to come to an intuitive understanding of them.

Discussion Questions

1. Why is the choice of a signal-to-noise ratio formula special for each application of Taguchi Methods?
2. If a formal methodology is merely the codification of experience and common sense, why is it helpful? List the pros and cons of using a formal method such as Taguchi’s seven-step process.

Problems

1. 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?