论文部分内容阅读
一、等级型资料的特点、分析指标和选用的统计分析方法医学上常见的等级型资料,如用某种药物或某种疗法治疗病人的效果按治愈、显效、好转和无效四个等级划分,有人常把其香作计数资料,用卡方X~2分析法。但X~2检验的是二项或多项分布,如血型的A型、B型、AB型和O型,它们之间有质的区别,分析时和项的排列次序无关,这属于纯粹的计数资料,只有抽样误差。而等级型资料不属于纯粹的计数资料,因为各等级之间没有质的区别,分级的多少是根据需要和可能决定的,必须考虑等级的排列次序,等级从低到高或从高到低排列,不能任意颠倒,所以等级资料可说是半计量资料,有抽样误差和判级误差。等级资料要分析的是平均等级,
First, the characteristics of hierarchical data, analysis of indicators and the use of statistical methods commonly used in medical grade data, such as the use of a drug or a therapies for the treatment of patients by cure, markedly effective, improved and invalid four levels, Some people often make their incense for counting data, using Chi square X ~ 2 analysis. However, X ~ 2 test for two or more distribution, such as blood type A, B, AB and O type, there is a qualitative difference between them, the analysis and the items in the order of arrangement, which belongs to the pure Counting data, only the sampling error. While the graded data is not purely count data, because there is no qualitative difference between levels, the number of grading is based on needs and may be decided, we must consider the order of ranking, ranking from low to high or descending order , Can not be arbitrarily reversed, so the level of information can be said to be semi-quantitative information, sampling error and rating error. Level data to be analyzed is the average level,