ROBUST DESIGN FOR SOFTWARE DEVELOPMENT - Software Design for Six-Sigma: A Roadmap for Excellence - page 485

point that, of course, all experimental designs (e.g., 2 k ,2 k − p ,3 k − p , etc.) can be used to
analyze SN ratios that you computed. In fact, the many additional diagnostic plots and
other options available for those designs (e.g., estimation of quadratic components,
etc.) may prove very useful when analyzing the variability (SN ratios) in the design.
As a visual summary, an SN ratio plot usually is displayed using the experiment
average SN ratio by factor levels. In this plot, the optimum settings (largest SN ratio)
for each factor easily can be identified.
For prediction purposes, the DFSS team can compute the expected SN ratio given
optimum settings of factors (ignoring factors that were pooled into the error term).
These predicted SN ratios then can be used in a verification experiment in which
the design team actually sets the process accordingly and compares the resulting
observed SN ratio with the predicted SN ratio from the experiment. If major de-
viations occur, then one must conclude that the simple main effect model is not
appropriate. In those cases, Taguchi (1987) recommends transforming the dependent
variable to accomplish additivity of factors, that is, to make the main effects model fit.
Phadke (1989: Chapter 6) also discusses, in detail, methods for achieving additivity of
factors.
A robustness case study is provided in the following section.
18.9 ROBUST DESIGN CASE STUDY NO. 1: STREAMLINING OF
DEBUGGING SOFTWARE USING AN ORTHOGONAL ARRAY 13
Debugging is a methodical process of finding and reducing the number of bugs,
or defects, in a computer program or a piece of electronic hardware, thus making
it behave as expected. Debugging tends to be harder when various subsystems are
tightly coupled, as changes in one may cause bugs to emerge in another. Debuggers
are software tools that enable the programmer to monitor the execution of a program,
stop it, restart it, set breakpoints, change values in memory, and even in some cases,
go back in time. The term debugger also can refer to the person who is doing the
debugging.
Generally, high-level programming languages, such as Java, make debugging
easier because they have features such as exception handling that make real sources
of erratic behavior easier to spot. In lower level programming languages such as C or
assembly, bugs may cause silent problems such as memory corruption, and it is often
difficult to see where the initial problem happened. In those cases, memory debugger
tools may be needed. 14
Software debugging is the process by which DFSS teams attempt to remove
coding defects from a computer program. It is not untypical for the debugging of
software development to take 40%-60% of the overall development time. Ultimately,
a great amount of difficulty and uncertainty surround the crucial process of software
13 Reprinted with permission of John Wiley & Sons, Inc. from Taguchi et al. (2005).
14 http://en.wikipedia.org/wiki/Debugging.