This question already has an answer here:
I have used a Box-Behnken experimental design. I have a full quadratic model.
However, I had to transform the response, $Y$ for the model to fit; I did this using a Box-Cox transformation with $\lambda=0.5$.
For example, one of the regressions is like this:
$$Y = 1.28 - 0.008X_1-0.025X_2-0.05X_3-0.13X_1^2-0.01X_2^2-0.006X_3^2\\ +0.02X_1X_2-0.05X_1X_3+0.09X_2X_3$$
How do I interpret the terms?