Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Data Science, Design, and Causal Inference

Data Science, Design, and Causal Inference Curr Epidemiol Rep (2017) 4:269–270 https://doi.org/10.1007/s40471-017-0129-5 INVITED COMMENTARY Robert W. Platt Published online: 24 October 2017 Springer International Publishing AG 2017 Jamie Robins’ 2001 paper “Data, Design, and Background different causal structures. To simplify, he assumes that all Knowledge in Etiologic Inference” [1] is one of his simplest other confounders have been dealt with by stratification, so and most straightforward, but in my view, most important, that different results between the different designs cannot be contributions to the epidemiologic literature. Every student explained by confounding. He then demonstrates that depend- ing on the design of the study and on external information of epidemiology, or other sciences that involve interpretation of quantitative results, should read this paper, preferably more about E, E*, and D, the interpretation of the study would be than once. The concepts discussed in the paper are fairly quite different. straightforward, but the core point of the paper is extremely The point of this paper is simple and obvious to a reader important: study design matters, and etiologic (i.e., causal) with good training in epidemiology, regardless of whether one inference requires assumptions and external sources of infor- is a devotee of causal diagrams or not. External information http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Current Epidemiology Reports Springer Journals

Data Science, Design, and Causal Inference

Platt, Robert
Current Epidemiology Reports , Volume 4 (4) – Oct 24, 2017
Download PDF
2 pages

Loading next page...
1
 
/lp/springer_journal/data-science-design-and-causal-inference-bXOXA3J1z7

References (13)

Publisher
Springer Journals
Copyright
Copyright © 2017 by Springer International Publishing AG
Subject
Medicine & Public Health; Epidemiology
eISSN
2196-2995
DOI
10.1007/s40471-017-0129-5
Publisher site
See Article on Publisher Site

Abstract

Curr Epidemiol Rep (2017) 4:269–270 https://doi.org/10.1007/s40471-017-0129-5 INVITED COMMENTARY Robert W. Platt Published online: 24 October 2017 Springer International Publishing AG 2017 Jamie Robins’ 2001 paper “Data, Design, and Background different causal structures. To simplify, he assumes that all Knowledge in Etiologic Inference” [1] is one of his simplest other confounders have been dealt with by stratification, so and most straightforward, but in my view, most important, that different results between the different designs cannot be contributions to the epidemiologic literature. Every student explained by confounding. He then demonstrates that depend- ing on the design of the study and on external information of epidemiology, or other sciences that involve interpretation of quantitative results, should read this paper, preferably more about E, E*, and D, the interpretation of the study would be than once. The concepts discussed in the paper are fairly quite different. straightforward, but the core point of the paper is extremely The point of this paper is simple and obvious to a reader important: study design matters, and etiologic (i.e., causal) with good training in epidemiology, regardless of whether one inference requires assumptions and external sources of infor- is a devotee of causal diagrams or not. External information

Journal

Current Epidemiology Reports – Springer Journals

Published: Oct 24, 2017

There are no references for this article.