Content deleted Content added
No edit summary |
Please use some common sense in links. OBVIOUSLY one would not expect an article titled contrast to be on the topic of contrast (statistics). |
||
Line 4:
|year=2009 |month= |pmid= |doi=10.2217/14622416.10.3.345 |url=}}</ref>. Each variable may represent random values in one of multiple groups involved in a comparison. Associated with a contrast variable are two terms: the standardized mean of contrast variable ([[SMCV]]) and [[c+-probability|c<sup>+</sup>-probability]]. The SMCV is the ratio of [[mean]] to [[standard deviation]] of a contrast variable and the c<sup>+</sup>-probability is the probability that a contrast variable obtains a positive value.
Traditional [[contrast (statistics)|contrast]] is a statistical parameter defined only on the group [[mean]]s. A contrast variable allows us to consider not only group [[mean]]s but also group [[variance]]s in a comparison. In addition, the concept of contrast variable can help to derive the [[effect size]]s across any number of groups readily and smoothly.<ref name=ZhangJBiometBiostat2010>{{cite journal |author=Zhang XHD
|title= Contrast variable potentially providing a consistent interpretation to effect sizes
|journal=Journal of Biometrics & Biostatistics |volume=1 |issue= |pages=108
| |||