Breaking down barriers: Helpful breakthrough statistical methods you need to understand better
Breaking down barriers: Helpful breakthrough statistical methods you need to understand better
复制标题
DOI：
10.1067/mtc.2001.117536
复制
发表时间：
2001-09-01
影响因子：
6
通讯作者：
Blackstone, EH
中科院分区：
文献类型：
--
作者：
Blackstone, EH
许多临床研究的编辑立场。然而，两个不同的分析师查看相同的数据集时，往往会发现不同的风险因素。除了不一致的发现，自动化的计算机算法经常被用来以一种逐步的方式辅助(甚至执行)多变量分析。这种做法通常受到统计学家的谴责。幸运的是，对于科学的变量选择来说，现在已经有了一项突破性的技术，它不仅可以从精心构建的、医学上知情的变量集中识别风险因素，还可以量化识别的可靠性。你需要更好地了解这项技术。
EDITORIAL stance of many clinical investigations. Yet, two different analysts looking at the same data set often identify different risk factors. 4-6 In addition to inconsistent findings, automated computer algorithms often are used to assist (or even perform) multivariable analyses in a stepwise fashion. This practice is generally condemned by statisticians. Fortunately for scientific variable selection, a breakthrough technology now exists that not only identifies risk factors from a carefully constructed, medically informed set of variables, but also quantifies the reliability of that identification. You need to understand this technology better.