ABSTRACT
One of the primary objectives in a response surface investigation is the determination of the optimum of a response of interest. Such an undertak ing may also be carried out when several responses are under consideration. For example, in a particular chemical experiment, a resin is required to have a certain minimum viscosity, high softpoint temperature, and high percen tage yield (see Chitra, 1990, p. 107). The actual realization of the optimum depends on the nature of the response(s) and the form of the hypothesized (empirical) model(s) being fitted to the data at hand.
