医学科研论文中列联表资料的统计方法及常见问题

The Statistic Methods for Contingency Table and Its Common Problems in Medical Research Papers

文章概要介绍了科研论文中列联表资料的类型及统计方法。结合编辑工作的实践经验,列举了列联表资料统计分析中的常见问题:将χ2检验视为分析列联表资料的万能工具;忽视对列联表中理论频数(T)的观察;对多个样本率或构成比比较的χ2检验不彻底;对双向有序的列联表资料进行直线相关分析等,并逐一进行了解析,以期引起学术期刊编辑对列联表资料统计分析问题的重视,增强对科研论文的审鉴能力,进而促进我国医学学术期刊整体水平的提高。

Contingency table data types and their statistical methods were briefly introduced. The common problems in the statistical analyses of contingency table data encountered in the editorial practice (including using χ2-examination as an omnipotent tool for contingency table analyses, ignoring the observation to the theoretical frequency(T) in contingency tables, incomplete χ2-examination on multiple sample rates or constitutional ratio comparisons, making linear correlation analyses to bi-dimensionally-ordered contingency table data, etc.) were listed and analyzed individually. This work would call academic journal editorial colleagues to pay attention to the problems in statistical analyses of contingency table data and would help them to improve their ability to evaluate research papers, and would also promote the overall level of medical academic journals in our country.