在生態統計學中，野生動物數量的估計是一件困難的工作，因為資料的搜集需要耗費大量的金額與時間， 而且資料分析往往受限於嚴謹的假設，故統計分析方法是否適當更顯重要。假設吾人對某區域內的動物總數N或動物密度D有興趣時， 線上穿越法為目前常使用的方法之一，因為其資料蒐集快速，研究範圍廣泛；然而在蒐集資料時， 常會出現資料無法精準取得。在統計研究中，不精準的資料使得分析存在有不可忽視的測量誤差問題， 本文採用Chen(1998)所提出的探測函數建構研究模型，並採用Hwang(1986)的乘法測量誤差模式， 發展利用動差法以及使用無母數機率密度函數的估計方法， 另外對於陳(2006)所考慮的加法測量誤差，並提出小測量誤差的近似估計方法。 In ecology research, estimating wildlife abundance is usually a difficult work. It may be expensive and time consuming to collect enough data for analysing. Besides collecting the data, the appropriateness of the analysis depends on whether the model assumption is plausible or not. The line transect approach had been considered as a convenient approach with minor assumptions and it requires relatively less effort in collecting data. In the present paper , we investigate the effects of additive measurement error and multiplicative measurement error on estimating the abundance in a line transect model , respectively . We propose estimators that correct biases using method of moments and nonparametric density estimation. The simulation result shows that these estimators work fine in reducing the biases when measurement error is present. Key words :Line transect model; Additive measurement error; Multiplicative measurement error ; Nonparametric density estimation.
