论文部分内容阅读
目的 :解决世界健康调查 (World Health Survey,WHS)数据中的数据缺失问题 ,进行更加有效的统计推断。 方法 :根据多重填补 (multiple imputation,MI)的原理 ,运用 Am elia软件和 Stata○R中的综合统计推断程序对 WHS数据集进行处理。结果 :由缺失数据造成的信息缺失得到了弥补 ,综合评价结果的质量得到了提高。 结论 :MI具有良好的特性 ,和针对特定问题但操作复杂的方法相比 ,MI是一种解决数据缺失问题的简单和近似的方法。
OBJECTIVE: To address the issue of data loss in the World Health Survey (WHS) data for more effective statistical inference. Methods: According to the principle of multiple imputation (MI), the WHS dataset was processed by the comprehensive statistical inference program of Am elia software and Stata ○ R. Results: The lack of information caused by the missing data has been offset and the quality of the comprehensive evaluation has been improved. Conclusion: MI has good characteristics, and MI is a simple and approximate method to solve the data missing problem compared to the method that is specific to a problem but complicated to operate.