Multiple Imputation for Missing Data in Life Cycle Inventory
The major environmental loads of mineral separation process in China iron production (with missing data) are analyzed. And the inner relationship between these loads data is qualified and the missing data are imputed using a statistic method called multiple imputation (MI), aimed to improve the quality of LCA datasets and allow industry to easily conduct a highly reliable LCA. By using computer simulation, MI replaces each missing value with a set of plausible values which represent the uncertainty of the missing data. The multiply imputed datasets are then analyzed by the standard procedures for completing data and combining the results from these analyses. The result proves that MI Method is an effective and reasonable method to solve the problem of missing data and therefore can ensure the validity and reliability of LCA.
Zhong Wei Gu, Yafang Han, Fu Sheng Pan, Xitao Wang, Duan Weng and Shaoxiong Zhou
Y. Liu et al., "Multiple Imputation for Missing Data in Life Cycle Inventory", Materials Science Forum, Vols. 610-613, pp. 21-27, 2009