摘要: In the 1939 Metro-Goldwyn-Mayer film, The Wizard of Oz, benevolent is asked to solve existential dilemmas Scarecrow, Tin Man and Cowardly Lion. Seemingly unable provide them intangible things that they seek, he instead gives three mundane items happens have in his bag. munificence not cast into doubt by this sleight hand, however, because characters discover traits sought were had possessed all along. article VanderWeele colleagues1 responds a crisis reproductive epidemiology over how handle measures maturity development. Variables such asgestational age birth weight are ubiquitous perinatal research, but problematic for reasons described their paper. Analysts want condition on some measure maturation put neonates an equal developmental footing, exposures interest may affect progress pregnancy, making conditional estimates biased as total effect. At same time, direct effects generally identifiable unmeasured common causes indicator outcome, leading birth-weight paradox authors. Reproductive epidemiologists therefore look eagerly methods wizards magical solution problem: brains, heart courage necessary daily travails. What methodologists got bag?
wkhealth.com参考文章(4)
Marshall M. Joffe, Celia Byrne, Graham A. Colditz, Postmenopausal Hormone Use, Screening, and Breast Cancer: Characterization and Control of a Bias Epidemiology. ,vol. 12, pp. 429- 438 ,(2001) , 10.1097/00001648-200107000-00013
Marshall M. Joffe, Graham A. Colditz, Restriction as a method for reducing bias in the estimation of direct effects Statistics in Medicine. ,vol. 17, pp. 2233- 2249 ,(1998) , 10.1002/(SICI)1097-0258(19981015)17:19<2233::AID-SIM922>3.0.CO;2-0
Constantine E. Frangakis, Donald B. Rubin, Principal stratification in causal inference. Biometrics. ,vol. 58, pp. 21- 29 ,(2002) , 10.1111/J.0006-341X.2002.00021.X
Maya L. Petersen, Sandra E. Sinisi, Mark J. van der Laan, Estimation of direct causal effects. Epidemiology. ,vol. 17, pp. 276- 284 ,(2006) , 10.1097/01.EDE.0000208475.99429.2D