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

Learn More →

Essential concepts of causal inference: a remarkable history and an intriguing future

Essential concepts of causal inference: a remarkable history and an intriguing future Causal inference refers to the process of inferring what would happen in the future if we change what we are doing, or inferring what would have happened in the past, if we had done something different in the distant past. Humans adjust our behaviors by anticipating what will happen if we act in different ways, using past experiences to inform these choices. ‘Essential’ here means in the mathematical sense of excluding the unnecessary and including only the necessary, e.g. stating that the Pythagorean theorem works for an isosceles right triangle is bad mathematics because it includes the unnecessary adjective isosceles; of course this is not as bad as omitting the adjective ‘right.’ I find much of what is written about causal inference to be mathematically inapposite in one of these senses because the descriptions either include irrelevant clutter or omit conditions required for the correctness of the assertions. The history of formal causal inference is remarkable because its correct formulation is so recent, a twentieth century phenomenon, and its future is intriguing because it is currently undeveloped when applied to investigate interventions applied to conscious humans, and moreover will utilize tools impossible without modern computing. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Biostatistics & Epidemiology Taylor & Francis

Essential concepts of causal inference: a remarkable history and an intriguing future

Rubin, Donald B.
Biostatistics & Epidemiology , Volume 3 (1): 16 – Jan 1, 2019
Download PDF

Essential concepts of causal inference: a remarkable history and an intriguing future

Abstract

Causal inference refers to the process of inferring what would happen in the future if we change what we are doing, or inferring what would have happened in the past, if we had done something different in the distant past. Humans adjust our behaviors by anticipating what will happen if we act in different ways, using past experiences to inform these choices. ‘Essential’ here means in the mathematical sense of excluding the unnecessary and including only the necessary, e.g. stating...
Loading next page...
 
/lp/taylor-francis/essential-concepts-of-causal-inference-a-remarkable-history-and-an-DxEwsO511Z
Publisher
Taylor & Francis
Copyright
© 2019 International Biometric Society – Chinese Region
ISSN
2470-9379
eISSN
2470-9360
DOI
10.1080/24709360.2019.1670513
Publisher site
See Article on Publisher Site

Abstract

Causal inference refers to the process of inferring what would happen in the future if we change what we are doing, or inferring what would have happened in the past, if we had done something different in the distant past. Humans adjust our behaviors by anticipating what will happen if we act in different ways, using past experiences to inform these choices. ‘Essential’ here means in the mathematical sense of excluding the unnecessary and including only the necessary, e.g. stating that the Pythagorean theorem works for an isosceles right triangle is bad mathematics because it includes the unnecessary adjective isosceles; of course this is not as bad as omitting the adjective ‘right.’ I find much of what is written about causal inference to be mathematically inapposite in one of these senses because the descriptions either include irrelevant clutter or omit conditions required for the correctness of the assertions. The history of formal causal inference is remarkable because its correct formulation is so recent, a twentieth century phenomenon, and its future is intriguing because it is currently undeveloped when applied to investigate interventions applied to conscious humans, and moreover will utilize tools impossible without modern computing.

Journal

Biostatistics & Epidemiology – Taylor & Francis

Published: Jan 1, 2019

Keywords: Experiments; observational studies; randomization; RCM (Rubin causal model); RCT (Randomized controlled trial)

References

  • Science and statistics
  • Controlling bias in observational studies: a review
  • Comment: Neyman (1923) and causal inference in experiments and observational studies
  • The causes of human variability
  • Comment on “Randomization analysis of experimental data in the Fisher randomization test” by Basu
  • Causal inference using potential outcomes: design, modeling, decisions
  • On the Application of probability theory to agricultural experiments: essay on principles, Section 9
  • Introductory remarks on Neyman (1923)
  • Statistical problems in agricultural experimentation (with discussion)
  • Experimental designs for identifying causal mechanisms
  • Discussion on the paper “Experimental designs for identifying causal mechanisms” by Imai, Tingley and Yamamoto
  • Discussion of “experimental designs for identifying causal mechanisms” by Imai. Tingley and Yamamoto
  • Reflections stimulated by the comments of Shadish (2010) and West and Thoemmes (2010). Pshchol
  • On the two different aspects of the representative method: the method of stratified sampling and the method of purposive selection
  • Studies in crop variation. II. The manorial response of different potato varieties
  • On small differences in sensation
  • An investigation into the causes of changes in Pauperism in England, chiefly during the last two intercensal decades (part I.)
  • Principles of research
  • Statistics and causal inference
  • Estimating causal effects of treatments in randomized and nonrandomized studies
  • Bayesian inference for causality: the importance of randomization
  • Inference and missing data
  • Assignment to treatment group on the basis of a covariate
  • Bayesian inference for causal effects: the role of randomization
  • The practical importance of understanding placebo effects and their role when approving drugs and recommending dosages
  • A case-study of the robustness of Bayesian methods of inference: estimating the total in a finite population using transformations to normality
  • Estimation of parameters in the multivariate normal distribution with missing observations
  • Maximum likelihood estimation with incomplete multivariate data
  • Subjective Bayesian models in sampling finite populations
  • Identification of causal effects using instrumental variables
  • The central role of the propensity score in observational studies for causal effects
  • Assessing sensitivity to an unobserved binary covariate in an observational study with binary outcome
  • Estimating the effects caused by treatments.” discussion of “On the nature and discovery of structure
  • Sensitivity of Bayes inference with data-dependent stopping rules
  • Difficulties with the regression analysis of age–adjusted rates
  • Reducing bias in observational studies using subclassification on the propensity score
  • Constructing a control group using multivariate matched sampling incorporating the propensity score
  • The bias due to incomplete matching
  • A difficulty with regression analysis of regional test–score averages. discussion of on state education statistics
  • The role of a second control group in an observational study
  • Heterogeneity and causality: unit heterogeneity and design sensitivity in observational studies. The Am
  • Sensitivity analysis for m-estimates, tests, and confidence intervals in matched observational studies
  • Combining propensity score matching and group-based trajectory analysis in an observational study
  • Split samples and design sensitivity in observational studies
  • Sensitivity of two-sample permutation inferences in observational studies
  • Assessing the effect of an influenza vaccine in an encouragement design
  • Identifiability and estimation of causal effects in randomized trials with noncompliance and completely nonignorable missing data (with discussion)
  • Methods for clustered encouragement design studies with noncompliance and missing data
  • Causal analysis in multi-arm trials with truncation by death
  • Identification and estimation of causal effects with outcomes Truncated by Death
  • ITT analysis of randomized encouragement design studies with missing data
  • Rerandomization to improve covariate balance in experiments