In the article referred to in Exercise 2, a model was fit to investigate the relationship between the independent variables given in Exercise 2 and the amount of S-isomer collected. The results of a least-squares fit are presented in the following table. (The article did not provide the value of the t statistic for the constant term.)

Of the following, which is the best next step in the analysis? Explain your reasoning.

i. Drop x²₁ , x²₂ , and x²₃ from the model, and then perform an F test.

ii. Nothing needs to be done. This model is fine.

iii. Add cubic terms x³₁ , x³₂ , and x³₃ to the model to try to improve the fit.

iv. Drop x₁x₂, x₁x₃, and x₂x₃ from the model, and then perform an F test.

v. Drop x²₂ , x₁x₂, and x₁x₃ from the model, and then perform an F test.

Predictor Coefficient P Constant 3.367 -0.018 -1.729 0.127 X2 1.396 135.987 0.000 0.104 10.098 0.000 0.017 1.471 0.184 -0.023 -0.909 0.394 -0.030 -0.006 -2.538 -0.466 0.039 0.655 X1X3 0.012 0.943 0.377 0.055 4.194 0.004