Final answer:
In building a complete 2nd order model, interactions are created between quantitative predictors, both linear and quadratic, and between linear quantitative and qualitative predictors. However, interactions between qualitative predictors and other qualitative predictors are not included, as qualitative variables do not have quantifiable differences that would justify the integration of such terms in the model.
Step-by-step explanation:
When building the complete 2nd order model, you need to create interactions between all of the following except d. qualitative predictors and other qualitative predictors. This is because qualitative (or categorical) variables do not have a natural ordering or quantifiable difference between levels, making the calculation of interactions such as quadratic terms or products of two qualitative predictors not meaningful in the context of a traditional 2nd order response surface model. Second order models typically include terms for the main effects, interaction effects between quantitative predictors, and squared terms of quantitative predictors to capture curvature in the model.
For a 2nd order model, interactions would be between linear quantitative predictors and quadratic quantitative predictors (a), as well as between linear quantitative predictors and qualitative predictors (c). These interactions and quadratic terms help to capture the relationship between predictors in a more complete way, which can be crucial for modeling data accurately and making reliable predictions.