Question: Consider a 24-1 fractional factorial study.
(a) How many different combinations of levels of the experimental factors are involved?
(b) Suppose three combinations were run twice and the other five were run once each. The 95 % confidence limits for sums of pairs of factorial effects (aliased pairs) are Eˆ ± Δ. Find Δ as a (numeric) multiple of .
(c) How many generators are involved in setting up this study?
(d) Suppose as run, D↔ABC and only the "grand mean plus alias," "A main effect plus alias," "BC interaction plus alias," and "ABC interaction plus alias" sums of effects are clearly detectable. Assuming all two-factor and higher-order interactions are negligible, give a simple interpretation of this result.