国民体质监测数据预处理方法研究  

Studies of pretreatment methods of monitoring data about civil constitution

作　　者：刘春艳[1] 沈丽琴[1] 宫川[2] 韩海军[2] 陈希宁[3] 

机构地区：[1]四川大学公共卫生学院儿少卫生与妇幼保健教研室,四川成都610041 [2]四川大学体育学院 [3]四川大学公众健康与社会发展研究所

出　　处：《现代预防医学》2013年第7期1211-1213,共3页Modern Preventive Medicine

摘　　要：目的从对成都市国民体质监测数据的整理过程中探讨出一条处理该类数据的优效、准确的方法,为今后国民体质监测和类似的数据预处理提供可参考的方法。方法对成都市国民体质监测的数据进行了包括纸质版数据的初步筛查、编号、建立Epidata数据库、数据录入、自查数据、交叉检查、数据转换、合并数据等8个步骤的处理。结果通过层层筛查,成都市体质监测数据3～6岁、20～39岁、40～59岁、60～69岁各年龄组剔除率分别为3.00%、2.41%、1.17%、0.64%,总剔除率达到1.95%。结论数据的剔除率与有效样本量是成反比的,数据的收集过程直接影响到数据的有效性和可靠性,也就是数据的质量,而现场调查的质量控制是高质量数据的关键,所以到位的现场质控和恰当的数据预处理,有助于筛选出相对可靠的数据,从而有利于得出真实、准确的统计分析结论。OBJECTIVE To explore a superior, accurate method in process of handling such monitoring data about Chengdu national physique, to provide reference for future national physique and similar data preprocessing. METHODS Chengdu national physique monitoring data were dealt with including the preliminary data screening of paper-based version, number to establish Epidata database, data entry, self -examination data, cross -checking, data conversion, merging data. RESULTS Through screening, removed rate of Chengdu physical monitoring data from 3 to 6 years old, 20 to 39 years old, 40 to 59 years, 60 to 69-year-old age group was 3.00%, 2.41%, 1.17%, 0.64%, total rate was 1.95%. CONCLUSION Removed rate and effective sample size are inversely proportional. The collection process of data directly affects the validity and reliability of the data, that is, the quality of the data, while the quality control of on-site survey is the key for quality data, so the placed on-site quality control and appropriate data preprocessing help filter relatively reliable data, thus contributing to get true, accurate statistical analysis conclusions.

关 键 词：国民体质监测 数据质量 预处理 

分 类 号：R195.1[医药卫生—卫生统计学]