正态模型缺失数据的贝叶斯和Jackknife多重插补法的比较

数据缺失是统计调查中经常存在的问题,若是少量缺失则可以利用删除法;若缺失值较多,利用删除法则会丢失大量有用信息,这时候就需利用插补法来补全数据,从而减少对统计分析的影响。根据统计年鉴上近几年的粮食产量、种植规模、有效灌溉面积等系列数据,分别采用贝叶斯多重插值法和刀切多重插值法展开了模拟研究,通过对两种方法所得数据的比对分析,来进一步掌握实际的插值效果。研究发现,利用这两种方法构建的模型都有较好的估计结果,但是贝叶斯多重插补法更为精确,而Jackknife法在操作方面则更为简单。

Comparison of Bayesian and Jackknife Multiple Interpolation Methods with Missing Data in Normal Model

Missing data is a common problem in statistical surveys.If there are a few missing,you can use the deletion method.If there are many missing values,the deletion method will lose a lot of useful information.In this case,you need to use the interpolation method to complete the data.Thereby reducing the impact on statistical analysis.This paper simulates the data of grain yield,planting area,effective irrigated area and chemical fertilizer application by using Bayesian multiple imputation method and Jackknife multiple imputation method to compare these two methods in agricultural survey.The study found that the models constructed by these two methods have good estimation results,but the Bayesian multiple interpolation method is more accurate,and the Jackknife method is simpler in operation.