论统计调查中的误差控制

Error Control in Statistical Investigation

摘要: 统计调查被应用在生产生活的方方面面，随着当代信息技术的发展，日益成为一项重要的数据分析方法，其全面性、准确性和普遍性使其成为实验的普适选择。而其中，数据质量的控制对于后续的统计分析起着关键性作用。本文将社会调查研究工作与数据分析具体方法相融合，以实际调查“浙江省大学生状态转变的探究”为例，从问卷数据回收分析的流程出发，分别探寻数据筛选、信效度检验、内在逻辑检验、随机性检验四个角度中误差控制的实时体现，通过数据、控制流程以及结论的共同体现，论证统计调查中的误差控制的重要性及其实际应用。

Abstract: Statistical survey is used in all aspects of production and life. With the development of modern information technology, it has increasingly become an important data analysis method. Its comprehensiveness, accuracy and universality make it a popular choice for experiments. Among them, the control of data quality plays a key role in the subsequent statistical analysis. This paper combines social investigation and research work with specific methods of data analysis, takes the actual investigation “Investigation on the state transformation of college students in Zhejiang Province” as an example, and starts from the process of questionnaire data recovery and analysis, respectively to explore the real-time manifestation of error control in four aspects: data screening, reliability and validity test, internal logic test and randomness test. The importance of error control in statistical investigation and its practical application are demonstrated through data, control process and conclusion.