摘要
目的了解１９９３年～１９９５年发表的３６０篇临床医学论著统计学方法应用的状况。方法参照研究设计类型及Ｅｍｅｒｓｏｎ统计学方法的分类并用卫生统计学理论加以分析。结果表明观察性和实验性研究设计分别约占２／３和１／３。其中描述性统计方法应用最多，主要采用的假设检验方法是ｔ检验和χ２检验。６８％存在统计学方法上的问题或错误。未说明检验方法或只有Ｐ值无统计量者、未说明α取值者和统计方法误用者，分别占有假设检验论文的７７％，７６％和３２％。临床试验缺乏对照组者占３５％。结论论文投稿的具体统计学要求应参照国际标准。
Objective The purpose of the study was to determine the type of study, the level of statistics and the extent of common statistical problems that were employed by the published articles. Method Three hundreds and sixty original articles published in the Chinese Journal of Internal Medicine from January 1993 through December 1995 were reviewed.. Results Observational and experimental designs were reported in nearly two thirds and one third, respectively. All these articles used descriptive statistics most frequently, followed by ttest and chisquare analysis of contingency table. Sixtyeight per cent of the articles that employed statistical techniques contained at least one statistical problem. The most common inadequacy was the statement of a probability value without a complete summary of the statistical results, failure to list the statistical tests and the alpha level used. The inappropriate use of statistics was reported in 32 per cent of the articles. Thirtyfive per cent were failure to include a control group in clinical trials. Conclusion We recommended that rather more comprehensive statistical guidelines to be reported should be given for contributors.
出处
《安徽医科大学学报》
CAS
1998年第2期108-110,共3页
Acta Universitatis Medicinalis Anhui
