论文部分内容阅读
为了解我国临床医学论著统计学应用情况，采用典型整群抽样法，对５种权威中华医学系列期刊，１９８５～１９９５年出版的１３１６篇临床医学论文原著统计推断应用缺陷进行调查。结果表明：８１６篇论著中至少存在一种统计推断应用缺陷，占６２．０％，其中，１９８５～１９８６年、１９８９～１９９０、１９９４～１９９５年缺陷率分别为７０．２％、６５．７％和４９．２％；最常见的缺陷问题是“统计推断方法未明确交待”，其中，１９８５～１９８６、１９８９～１９９０、１９９４～１９９５年仅存在此类缺陷的论著分别占４３．３％、３５．７％和２０．０％。结论认为，近１０年间，我国临床论著统计推断误用情况未发生明显改善。
In order to understand the application of statistics in clinical medicine in China, a typical cluster sampling method was used to investigate the defects of the original authoritative statistical application of 5 authoritative Chinese medical journal series and 1316 clinical medical articles published from 1985 to 1995. The results show that there are at least one kind of statistical inference application defects in 816 articles, accounting for 62.0%, among which the defect rates in 1985-1986, 1989-1990 and 1994-1995 were 70.2% and 65.7% respectively. And 49.2%; the most common defect problem is “the statistical inference method is not explicitly confessed.” Among them, 1985-1986, 1989-1990, and 1994-1995 only accounted for 43.3% of these defects. 35 .7% and 20.0%. It is concluded that in the past 10 years, there has been no significant improvement in the misuse of statistical inference from clinical treatises in China.