关键词: 临床研究; 假设检验; 置信区间; 回归分析; 误用

Abstract: he c orrect u se o f s tatistical m ethods a re a ssociated w ith t he r eliability a nd scientificity of research results in clinical research. The key points and common misuses of statistical description, statistical inference, data processing and results interpretation in clinical research were summarized and analyzed through typical case analysis in this article. Misuse of t test and analysis of variance, misuse of χ 2 test and paired χ 2 test, dummy variable s etting of categorical independent variables in regression analysis, precautions for regression analysis of repeated measurement data, multiple imputation of missing data, prudent interpretation of P value and correlation analysis results were also stressed, to provide reference and basis for clinical workers to apply normative statistical methods in scientific research, and avoid the misuse of statistical methods, so as to improve the quality and level of clinical research.

Key words: Clinical research; Hypothesis testing; Confidence interval; Regression analysis;Misuse