Author Country (or Countries)



Two second order response surface designs (central composite design and extended central composite design) are compared on the basis of their slope estimation capabilities of a response function using quantile plots. A large number of points is randomly generated on a sphere of radius centred at the origin in a given region of interest. The scaled average slope variance is evaluated at each of these points and the quantiles of the resulting values are plotted. Both designs were compared using slope rotatability design criterion and result obtained show that both designs have stable slope variance and the extended central composite design is better than the central composite design for large values of the radius.

Suggested Reviewers