A statistical modelling method is presented for yield of maize with fertilizer response using response surface methodology. It was shown that careful consideration of the class of response surface model (RSM), using domain knowledge, give models which are of order two magnitude as good as the more conventional polynomial models. This study shows that the modified inverse polynomial performs better in studying the response of the yield of maize as against the ordinary polynomial.

Keywords: Modified inverse polynomial, ordinary polynomial, response surface model, Maize.