数据的数学耦合：一个常见的误差来源。
Mathematic coupling of data: a common source of error.
作者信息
Archie J P
出版信息
Ann Surg. 1981 Mar;193(3):296-303. doi: 10.1097/00000658-198103000-00008.
The relationship between two variables may be mathematically coupled if either one or both variables are derived and/or calculated, and this can lead to erroneous results and invalid conclusions. The purpose of this report is to identify four types of mathematic coupling of data. Type 1 coupling involves directional changes in two variables which are mathematically coupled. Type 2 coupling is the functional relationship between two calculated variables which have one or more common component variables. Type 3, the most common type of mathematic coupling, is direct algebraic coupling between two variables, when one or more of the variables is derived and/or calculated. Type 4 is indirect coupling or physiologic coupling. The common problem in each type of mathematic coupling is that one variable either directly or indirectly contains the whole or components of the second variable. Statistical techniques, when properly applied to the relationship between the two variables, further obscure the underlying mathematic coupling, and tend to support the erroneous results. Recognition of mathematic coupling is imperative for correct data analysis and accurate interpretation.
如果一个变量或两个变量都是派生的和/或计算得出的，那么这两个变量之间的关系在数学上可能是耦合的，这可能会导致错误的结果和无效的结论。本报告的目的是识别数据的四种数学耦合类型。类型1耦合涉及两个在数学上耦合的变量的方向变化。类型2耦合是两个计算变量之间的函数关系，这两个变量有一个或多个共同的组成变量。类型3是最常见的数学耦合类型，当一个或多个变量是派生的和/或计算得出时，两个变量之间存在直接代数耦合。类型4是间接耦合或生理耦合。每种数学耦合类型的常见问题是，一个变量直接或间接地包含第二个变量的全部或部分。统计技术在正确应用于两个变量之间的关系时，会进一步掩盖潜在的数学耦合，并倾向于支持错误的结果。认识到数学耦合对于正确的数据分析和准确的解释至关重要。
Zotero 插件