...

ch11-toc-plus-p331-f..

by user

on
Category:

science

41

views

Report

Comments

Transcript

ch11-toc-plus-p331-f..
From Causality, Second edition, 2009.
xii
11
Contents
10.3.4 Path-Switching Causation
10.3.5 Temporal Preemption
10.4 Conclusions
324
325
327
Reflections, Elaborations, and Discussions with Readers
11.1 Causal, Statistical, and Graphical Vocabulary
11.1.1 Is the Causal-Statistical Dichotomy Necessary?
11.1.2 d-Separation without Tears (Chapter 1, pp. 16–18)
11.2 Reversing Statistical Time (Chapter 2, p. 58–59)
11.3 Estimating Causal Effects
11.3.1 The Intuition behind the Back-Door Criterion
(Chapter 3, p. 79)
11.3.2 Demystifying “Strong Ignorability”
11.3.3 Alternative Proof of the Back-Door Criterion
11.3.4 Data vs. Knowledge in Covariate Selection
11.3.5 Understanding Propensity Scores
11.3.6 The Intuition behind do-Calculus
11.3.7 The Validity of G-Estimation
11.4 Policy Evaluation and the do-Operator
11.4.1 Identifying Conditional Plans (Section 4.2, p. 113)
11.4.2 The Meaning of Indirect Effects
11.4.3 Can do(x) Represent Practical Experiments?
11.4.4 Is the do(x) Operator Universal?
11.4.5 Causation without Manipulation!!!
11.4.6 Hunting Causes with Cartwright
11.4.7 The Illusion of Nonmodularity
11.5 Causal Analysis in Linear Structural Models
11.5.1 General Criterion for Parameter Identification
(Chapter 5, pp. 149–54)
11.5.2 The Causal Interpretation of Structural Coefficients
11.5.3 Defending the Causal Interpretation of SEM (or, SEM
Survival Kit)
11.5.4 Where Is Economic Modeling Today? – Courting
Causes with Heckman
11.5.5 External Variation vs. Surgery
11.6 Decisions and Confounding (Chapter 6)
11.6.1 Simpson’s Paradox and Decision Trees
11.6.2 Is Chronological Information Sufficient for
Decision Trees?
11.6.3 Lindley on Causality, Decision Trees, and Bayesianism
11.6.4 Why Isn’t Confounding a Statistical Concept?
11.7 The Calculus of Counterfactuals
11.7.1 Counterfactuals in Linear Systems
11.7.2 The Meaning of Counterfactuals
11.7.3 d-Separation of Counterfactuals
331
331
331
335
337
338
338
341
344
346
348
352
352
354
354
355
358
359
361
362
364
366
366
366
368
374
376
380
380
382
384
387
389
389
391
393
Contents
11.8
11.9
xiii
Instrumental Variables and Noncompliance
11.8.1 Tight Bounds under Noncompliance
More on Probabilities of Causation
11.9.1 Is “Guilty with Probability One” Ever Possible?
11.9.2 Tightening the Bounds on Probabilities of Causation
395
395
396
396
398
Epilogue The Art and Science of Cause and Effect
A public lecture delivered November 1996 as part of
the UCLA Faculty Research Lectureship Program
401
Bibliography
429
Name Index
453
Subject Index
459
CHAPTER ELEVEN
Reflections, Elaborations, and
Discussions with Readers
As X-rays are to the surgeon,
graphs are for causation.
The author
In this chapter, I reflect back on the material covered in Chapters 1 to 10, discuss issues
that require further elaboration, introduce new results obtained in the past eight years,
and answer questions of general interest posed to me by readers of the first edition.
These range from clarification of specific passages in the text, to conceptual and philosophical issues concerning the controversial status of causation, how it is taught in classrooms and how it is treated in textbooks and research articles.
The discussions follow roughly the order in which these issues are presented in the
book, with section numbers indicating the corresponding chapters.
11.1
CAUSAL, STATISTICAL, AND GRAPHICAL VOCABULARY
11.1.1 Is the Causal–Statistical Dichotomy Necessary?
Question to Author (from many readers)
Chapter 1 (Section 1.5) insists on a sharp distinction between statistical and causal concepts; the former are definable in terms of a joint distribution function (of observed variables), the latter are not. Considering that many concepts which the book classifies as
“causal” (e.g., “randomization,” “confounding,” and “instrumental variables”) are commonly discussed in the statistical literature, is this distinction crisp? Is it necessary? Is it
useful?
Author Answer
The distinction is crisp,1 necessary, and useful, and, as I tell audiences in all my lectures:
“If you get nothing out of this lecture except the importance of keeping statistical and
causal concepts apart, I would consider it a success.” Here, I would dare go even further:
1
The basic distinction has been given a variety of other nomenclatures, e.g., descriptive vs. etiological, associational vs. causal, empirical vs. theoretical, observational vs. experimental, and
many others. I am not satisfied with any of these surrogates, partly because they were not as crisply
defined, partly because their boundaries got blurred through the years, and partly because the concatenation “nonstatistical” triggers openness to new perspectives.
331
Fly UP