The quality of statistics data is very important.This paper researches the problem mainly from the error of the raw data.It is important that we can judge the error in the raw data if there is it and remove its influence.We can use the technique and the thinking of variance analysis,effect chart and interpenetrating sample.
This paper deals with the quality of statistical data in terms of error.It holds that the quality of statistical data is a matter of error in nature.In view of total error,however,the system error is much more difficult to treat than the random error is.The writer makes a detailed analysis of the characteristics structures and forms of the system error,and points out that what is the important is not measuring the value of the system error,but discovering the data with the system error and removing its effect.
90年代初,在美国亚利桑那州的图森(Tucson)召开了“调查中的计量误差国际研讨会”(The International Conference onMeasurement Errors in Survey),会议的直接成果之一,就是出版了论文集《调查中的计量误差》(Measurement Errors in Survey),目前可看作是这方面的较新成果。很显然,计量误差虽然是抽样调查中系统性误差产生的重要原因,但决不是唯一的原因,系统性误差可能产生在抽样调查的任何一个环节。
