Abstract
在這一篇論文中，我們修正 Hwang 及 Hu (1994) 的指數樣本變異係數機率密度函數的遞迴關係式，並導出指數樣本變異係數在樣本數為三、四、五之下的機率密度函數。隨著樣本數的增加，我們探討其尾部機率的發展情況；並呈述為何研究指數變異係數的原因及其與其他分配的樣本變異係數之轉換關係。 最後， 將結果應用到聚合體分子的異質性(Heterogeneity) 判斷。In this paper, we modify the recursive relationship of theprobability densityfunction of the sample coefficient ofvariation under exponential population obtained by Hwang and Hu(1994), and derive the p.d.f. of the coefficient of variationfor sample size n=3,4, and 5. The tail probability is discussedwhenn gets bigger. Furthermore, we give some reasons of usingthe exponential coefficient of variation. At last, we present anapplication to polymer molecules.