Design of Experiment QUESTION 1 An experimenter wants to conduct a factorial experiment involving six factors each at two levels. She is planning to use an experiment with exactly eight runs. Answer the following questions about this experiment. The recommended generators for this designs are D = AB, E = AC, and F = BC. True False QUESTION 2 An experimenter wants to conduct a factorial experiment involving six factors each at two levels. She is planning to use an experiment with exactly eight runs. Answer the following questions about this experiment. The resolution of this fractional factorial is IV II III V None of the above QUESTION 3 An experimenter wants to conduct a factorial experiment involving six factors each at two levels. She is planning to use an experiment with exactly eight runs. Answer the following questions about this experiment. The number of words in the complete defining relation is Four Eight Seven Three None of the above QUESTION 4 An experimenter wants to conduct a factorial experiment involving six factors each at two levels. She is planning to use an experiment with exactly eight runs. Answer the following questions about this experiment. The longest word in the defining relation has four letters. True False QUESTION 4 An experimenter wants to conduct a factorial experiment involving six factors each at two levels. She is planning to use an experiment with exactly eight runs. Answer the following questions about this experiment. The longest word in the defining relation has four letters. True False QUESTION 5 An experimenter wants to conduct a factorial experiment involving six factors each at two levels. She is planning to use an experiment with exactly eight runs. Answer the following questions about this experiment. Suppose that the generators that the experimenter actually used were D = AB, E = AC, and F =A BC. The resolution of this design would be: IV II III V None of the above QUESTION 6 An experimenter wants to conduct a factorial experiment involving six factors each at two levels. She is planning to use an experiment with exactly eight runs. Answer the following questions about this experiment. Suppose that the generators that were actually used in the design were D=AB, E=AC and F=ABC. After analyzing the data, the three largest effects were A, B, and E. Dropping the negligible factors results in a full factorial design in factors A, B, and E. True False QUESTION 7 An experimenter wants to conduct a factorial experiment involving six factors each at two levels. She is planning to use an experiment with exactly eight runs. Answer the following questions about this experiment. Suppose that the generators D=AB, E=AC and F=ABC were actually used in the design. After analyzing the data, the three largest effects were A, C, and E. Dropping the negligible factors results in a full factorial design in factors A, C, and E. True False QUESTION 8 An experimenter wants to conduct a factorial experiment involving six factors each at two levels. She is planning to use an experiment with exactly eight runs. Answer the following questions about this experiment. Suppose that after analyzing the data, the experimenter decides to conduct a full fold over to augment the original eight runs. This would involve creating a new fraction by changing the signs on all of the original generators. True False QUESTION 9 An experimenter wants to conduct a factorial experiment involving six factors each at two levels. She is planning to use an experiment with exactly eight runs. Answer the following questions about this experiment. There are three alternate fractions for the design in this problem. True False QUESTION 10 In constructing a fractional factorial, the basic design is always a full factorial that has the correct number of runs required for the fractional True False QUESTION 11 A 2k-p fractional factorial design of resolution R will always project into a full factorial in a subset of R - 1 of the original k factors. True False QUESTION 12 Center points can be added to any two-level design, just so long as there is at least one continuous (numeric) factor. True False QUESTION 13 The recommended number of center points to add to an unreplicated two-level factorial or fractional design is: Between 1 and 3 Between 3 and 5 At least 6 At lease 4 QUESTION 14 To design an unreplicated two-level factorial design in two blocks, a good rule of thumb is to always confound the highest-order interaction with blocks. True False QUESTION 15 Plackett-Burman designs are resolution IV designs. True False QUESTION 16 The resolution of a 2k-p fractional factorial is equal to the number of letters in the shortest word in the complete defining relation for the design. True False QUESTION 17 Every 2k-p fractional factorial design contains a replicated full factorial design in some subset of the original factors. True False QUESTION 18 If there is no replication in a factorial experiment, the recommended approach to the analysis is to fit the full model, including all and use the analysis of variance to identify significant effects. True False QUESTION 19 The half-normal probability plot of effects plots cumulative normal probability versus the absolute effect magnitude. True False QUESTION 20 A split-plot design allows the experimenter to incorporate easy-to-change and hard-to-change factors by running the hard-to-change factor(s) in the subplots. True False QUESTION 21 The levels of a random factor are sampled from a large population of possible levels. True False QUESTION 22 The adjusted R-squared statistic always increases as variables are added to a model. True False QUESTION 23 A common application or random and fixed ANOVA models is in measurement system capability experiments. True False QUESTION 24 A resolution III fractional factorial design can always be folded-over to form a resolution IV design always contains a full factorial design in any two of the original factors can always be used to fit a model in the main effects if interactions are ignored none of the above (a-c) are true all of the above (a-c) are true QUESTION 25 In an experiment that has at least one nested factor : one may identify the nested factor by arbitrarily renumbering its levels the primary factor will always be a fixed effect the nested factor will always be a random effect all of the above statements (a - c) are true all of the above statements (a - c) are false