Key points are not available for this paper at this time.
The steepest ascent, univariate, factorial, and random methods for seeking maxima were compared by trying versions of them in several two-factor situations, with and without experimental error. Of the two sequential methods, the steepest ascent was better than the univariate. Of the nonsequential methods, the factorial was better than the random. The sequential methods were superior to the nonsequential.
Samuel H. Brooks (Sat,) studied this question.