I want to do Polynomial orthogonal contrasts (quadratic and linear) instead of Duncan's multiple range analysis to analyse all the response datas of my dietary protein requirement experiment.
I am not sure that there is a ready-made orthogonal polynomial fitting algorithm or subprogram in SPSS. However with sufficient programming knowledge and ability one can improvise such a subroutine in SPSS or in any other software. I improvise subroutines as the need arises and if possible.