Lorton L, Rethman M P
Henry M. Jackson Foundation for Military Medicine, Rockville, MD.
J Endod. 1990 Jan;16(1):13-8. doi: 10.1016/S0099-2399(07)80024-0.
Research projects and reports may fail because of investigators' misunderstandings about the principles underlying basic data analysis. Such errors may taint otherwise valid conclusions; related mistakes may serve to falsely support incorrect results. This article explains some of the principles underlying the use of statistical analysis for normally distributed interval data and explores the potential ramifications of misusing summary statistical tests on non-normal distributions. A discussion of the meaning of the statement of probability concerning hypothesis testing and a suggested attitude for researchers toward significance testing is also included.
研究项目和报告可能会失败，原因是研究人员对基础数据分析所依据的原则存在误解。此类错误可能会损害原本有效的结论；相关的错误可能会被用来错误地支持不正确的结果。本文解释了对正态分布区间数据进行统计分析所依据的一些原则，并探讨了对非正态分布滥用汇总统计检验的潜在后果。本文还讨论了关于假设检验的概率陈述的含义，以及对研究人员进行显著性检验的建议态度。
