當人類眼睛觀看同時存在的幾個色彩時，不會僅對各單一色彩發生感覺，而是同時對所有相關色彩，共同產生「色相對比」、「彩度對比」、「明度對比」三種「色彩對比」的色彩感覺。亦即根據這些色彩的「色相相互關係對比」、「彩度相互關係對比」、「明度相互關係對比」，來感知這些色彩的互相搭配結果。所以作者就在前期研究，根據人類眼睛同時觀看3 個色彩的「色相相互關係」、「彩度相互關係」、「明度相互關係」，建立色彩搭配關係的「實驗樣本群集」，再由所建立的「實驗樣本群集」，抽取合適的許多個「實驗樣本」，組成「配色調查樣本」；最後提出「配色調查樣本」組成模式。
可是若僅根據配色調查目的，就要在所選取的「實驗樣本群集」中，抽取合適的「實驗樣本」組成「配色調查樣本」，仍然太困難。因為相關變數仍太多。所以本文研究目的，就試圖藉由組成變數的合理控制，發展出具系統性的組成修正模式，希望能有效降低「配色調查樣本」組成的困難度。經重複測試，發現“配色調查樣本均勻共用色彩樣本”是一項重要的變數控制，且能有效達到組成簡化的目的。於是就延續前期研究所提出的「配色調查樣本」組成模式架構下，試圖控制該項重要變數。經由變數控制後，就可以將一些不合理的排列結果，在「配色調查樣本」組成之前，先將其刪除，最後達到大幅度簡化「配色調查樣本」組成模式的目的。
When human see several colors they don’t perceive single color only, they also have contrast colors perception about hue, chroma, and value at the same time. It means people can perceive the results of color coordinates by hue, chroma, and value mutually contrast relations. According to above relations supposed people see three colors, the author built color coordinate “experimental samples group” in a former study(2004), and then sampled many proper experimental samples from the group to comprise “color coordinate census samples”. Finally the author proposed the model of “color coordinate census samples” in the former study.
However according to census purpose to derive proper “experimental samples”, which from “experimental samples group”, to comprise “color coordinate samples” is still difficult because of too many relevant variables. In this respect, the purpose of this study is to make the compositive variables under controlling properly, and develops a systematic modified composition model. Hopefully that can reduce difficulties of “color coordinate census samples” composition. Through repeated test, we can find that “color coordinate samples equally using color samples” is a vital variable control. It also can simplify the composition of samples. So this study controls the vital variable under the composition model of “color coordinate census samples” which provided from former study. Thus we can eliminate inapposite results before the formation of “color coordinate census samples”, and largely simplify its composition model.
