The simplified s ̂ and y ̂ models, shown below, were obtained from the DOE KISS regression results shown in Tables 8.20 and 8.21, respectively. Note that in using the ROT to detect significant factors for the s ̂ model, namely “|coef|” ≥ (constant/2), we see that|coef|” for A is 0.198 which is close to but not at least half of 0.4348. However, ROTs also require judgment and by looking at the size of the coefficients, we see that the coefficient for A is significantly larger than all of the other coefficients. Hence, A was deemed to be significant and was included in the s ̂ model, confirmation runs would indeed verify this decision. Using the ROT for y ̂ models, namely “P(2 TALL) < .05,” we see that the only significant effects are from A, C, and AC, which are all included in the y ̂ model.