Example 6.17 illustrates how to obtain a statistical design that covers all interaction pairs in 20 configurations instead of 32. (a) Enumerate all test configurations by listing the levels for each...

Example 6.17 illustrates how to obtain a statistical design that covers all interaction pairs in 20 configurations instead of 32. (a) Enumerate all test configurations by listing the levels for each factor. (b) How many interactions pairs are covered more than once? (c) Derive an alternate design that covers all interaction pairs and satisfies the operating system and hardware related constraint. (d) Is your design better than the design in Example 6.17 in any way? (e) Can you think of any error in the AGTCS software that might not be discovered by testing against the 20 configurations derived in Example 6.17?