医学论文计数资料假设检验常见的统计错误

COMMON STATISTICAL MISTAKES IN HYPOTHESIS TESTING OF ENUMERATION DATA IN MEDICAL LITERATURE

为了确保假设检验在计数资料统计推断中的正确应用,本文以实例分析了计数资料假设检验中常见的统计错误;未经假设检验下结论;校正x^2检验误用x^2检验;列联表x^2检验使用不当;显著性水准的高低误解为指标差异的大小.对如何正确进行假设检验进行了讨论.

In order to ensure the correct use of the hypothesis testing in the statistical inference of enumeration data, this paper analyses by example the common statistica mistakes in the hypothesis testing of enumeration data. The mistakes includel conclusions drawn without making a test of hypothesis, corrected chi-square wrongly used in stead of chi-square test, chi-square test of contingency table data used: incorrectly, the significance level misunderstood as the extent of difference among indices. Discussion is made as to how to make tests of hypothesis correctly.