The Bias Correction of the Parameters Given a Censored Normal Sample

• Main Authors: Hsieh AI-RU, 謝璦如
• Other Authors: Tsai TIEN-LUNG
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/46972997736571442646

碩士 === 國立臺北大學 === 統計學系 === 92 === Many instruments are used in biological, industrial and medical experiments. Every instrument has its Detection Limit (DL) and the observation values below Detection Limit (DL) cannot be detected. These non-detected values are also known as Left Censoring data. Maximum Likelihood Estimator (MLE) is in common use to estimate parameters. When the estimator has a bias, a process called Bias Correction will be conducted. The bias will become more severe in small samples. The purpose of this thesis is to conduct Bias Correction based on the normal distribution of Type Ⅱ Censor Ordered Sample. The bias correction formula can be applied on the areas which experimental instruments are used, such as Quality Control, Data Analysis, Bio-statistics, Medical Test, etc.