UCLA Causality-Blog

http://www.mii.ucla.edu/causality/

This note comments on a paper published in NeuroImage which argues (yes, again) that the potential outcome model is somehow superior, more rigorous or more principled than the structural models used in fMRI research. To further illuminate the logic of such claims I have added a section (4.4.2) in this paper: http://ftp.cs.ucla.edu/pub/stat_ser/r370.pdf which demonstrates how potential outcomes can be generated, on demand, from a simple structural model, and no one can tell where they came from. Enjoy.

1.3 "The Causal Mediation Formula - A practitioner guide to the assessment of causal pathways" http://ftp.cs.ucla.edu/pub/stat_ser/r379.pdf

This paper present mediation analysis to researchers in the tradition of Baron and Kenny (1986), and shows through examples how "the percentage explained by mediation" and "the percentage owed to mediation" are estimated in nonlinear models with both continuous and categorical variables.

1.4 Simpson's Paradox Sander Greenland brought to my attention a recent paper in Synthese (Sept. 28, 2010) claiming that Simpson's paradox is NOT rooted in causal, but in some other kind of illusion. I remain convinced of the former, and have accordingly modified the Simpson Paradox entry in Wikipedia:Simpson Paradox to reinforce the causal-illusion theory. You might wish to add your take on the subject.

1.5 The ETT Paradox (or, the curse of free will) This paradox would be appreciated by those who are fascinated, like me, by our ability to determine, from data alone, if one would have been better off acting differently from the way one actually did. This can lead to a cycle of inevitable regret, and provoke some naughty thoughts. http://ftp.cs.ucla.edu/pub/stat_ser/r375.pdf

2. New Results

A newly posted paper, "On the Control of Selection Bias" http://ftp.cs.ucla.edu/pub/stat_ser/r381.pdf gives graphical and algebraic conditions for the removal of selection bias and the recovery of covariate-specific effect measures.

2.2 A new section (Section 5) in http://ftp.cs.ucla.edu/pub/stat_ser/r372.pdf generalizes the concept of transportability from experimental to observational studies,and shows how one can avoid re-learning things from scratch when moving to a new population, new domain, or a new environment.

2.3 After months of struggling with the literature of "surrogate endpoints" we feel that we now have a fairly satisfactory theory of surrogacy, It is based on the idea that a surrogate should serve not merely as a good predictor of outcomes, but also as ROBUST predictor of effects in the face of changing external conditions. See section 6 in http://ftp.cs.ucla.edu/pub/stat_ser/r372.pdf

3. Courses, Lectures and Conferences, 3.1 Causal Inference Course

Thomas Richardson and Michael Hudgens are once again teaching Causal Inference June 13-15, 2011 in the Summer Institute here at U Washington. The pdf of the brochure is attached.

The website is http://depts.washington.edu/sismid/

They have funds to support tuition waivers and some travel for students and postdocs.

3.2 2011 Atlantic Causal Conference

The 2011 Atlantic Causal Conference will take place at the University of Michigan School of Public Health in Ann Arbor, Michigan, Thursday May 19th and Friday May 20th. See http://www.sph.umich.edu/biostat/2011acic/index.html for

Contact: Mike Elliott at mrelliot@umich.edu or Ben Hansen at ben.hansen@umich.edu

3.3 Errata for Causality (2010) FYI, Cambridge University Press has come up with a new printing of my book Causality, which corrects a few errors in the 2009 edition. Please advise students to insist on a copy saying "reprinted 2010". If you have an older copy, you can find the corrections marked in red here: http://bayes.cs.ucla.edu/BOOK-09/errata_scanned_pages7-28-10.pdf

3.4 Lecture Slides available Slides of my lecture on "What's New in Causal Inference" can be viewed on my home page. http://bayes.cs.ucla.edu/jp_home.html 2nd line from the top. You are welcome to use them in any way you choose. But usage for a good cause is recommended.

Looking forward to postings from you, and may clarity prevail.

Judea Pearl

            UCLA

http://www.mii.ucla.edu/causality/. http://bayes.cs.ucla.edu/csl_papers.html

Spring March 24, 2011 Causality blog

Spring-time Greetings from the Causality blog, From: "Judea Pearl" <judea@CS.UCLA.EDU> Date: Thu, March 24, 2011 5:26 pm

Dear colleague in causality research,

This is an End-of-Winter Greeting from the UCLA Causality blog, welcoming you back to a spring-time discussion in causality-related issues.

This mesg contains 1. Topics under discussion 2. New results 3. Information on courses, lectures, and conferences.

1. Discussions inviting comments

1.1 Principal Stratification - A goal or a tool?' http://ftp.cs.ucla.edu/pub/stat_ser/r382.pdf Posted for discussion by the International Journal of Biostatistics (IJB), this paper questions whether studies based on Principal Stratification target quantities that researchers truly care about.

If you have comments, ideas or objections, you are invited to communicate them to the IJB's Editor, "Nicholas P. Jewell" <mm-11332-3261687@bepress.com> or/and, if you wish, cross-post them on this blog.

1.2 Comments and Controversies: Graphical models, potential outcomes and causal inference: Comment on Lindquist and Sobel" http://ftp.cs.ucla.edu/pub/stat_ser/r380.pdf

This note comments on a paper published in NeuroImage which argues (yes, again) that the potential outcome model is somehow superior, more rigorous or more principled than the structural models used in fMRI research. To further illuminate the logic of such claims I have added a section (4.4.2) in this paper: http://ftp.cs.ucla.edu/pub/stat_ser/r370.pdf which demonstrates how potential outcomes can be generated, on demand, from a simple structural model, and no one can tell where they came from. Enjoy.

1.3 "The Causal Mediation Formula - A practitioner guide to the assessment of causal pathways" http://ftp.cs.ucla.edu/pub/stat_ser/r379.pdf

This paper present mediation analysis to researchers in the tradition of Baron and Kenny (1986), and shows through examples how "the percentage explained by mediation" and "the percentage owed to mediation" are estimated in nonlinear models with both continuous and categorical variables.

1.4 Simpson's Paradox Sander Greenland brought to my attention a recent paper in Synthese (Sept. 28, 2010) claiming that Simpson's paradox is NOT rooted in causal, but in some other kind of illusion. I remain convinced of the former, and have accordingly modified the Simpson Paradox entry in Wikipedia to reinforce the causal-illusion theory. You might wish to add your take on the subject.

1.5 The ETT Paradox (or, the curse of free will) This paradox would be appreciated by those who are fascinated, like me, by our ability to determine, from data alone, if one would have been better off acting differently from the way one actually did. This can lead to a cycle of inevitable regret, and provoke some naughty thoughts. http://ftp.cs.ucla.edu/pub/stat_ser/r375.pdf

2. New Results

A newly posted paper, "On the Control of Selection Bias" http://ftp.cs.ucla.edu/pub/stat_ser/r381.pdf gives graphical and algebraic conditions for the removal of selection bias and the recovery of covariate-specific effect measures.

2.2 A new section (Section 5) in http://ftp.cs.ucla.edu/pub/stat_ser/r372.pdf generalizes the concept of transportability from experimental to observational studies,and shows how one can avoid re-learning things from scratch when moving to a new population, new domain, or a new environment.

2.3 After months of struggling with the literature of "surrogate endpoints" we feel that we now have a fairly satisfactory theory of surrogacy, It is based on the idea that a surrogate should serve not merely as a good predictor of outcomes, but also as ROBUST predictor of effects in the face of changing external conditions. See section 6 in http://ftp.cs.ucla.edu/pub/stat_ser/r372.pdf

3. Courses, Lectures and Conferences, 3.1 Causal Inference Course

Thomas Richardson and Michael Hudgens are once again teaching Causal Inference June 13-15, 2011 in the Summer Institute here at U Washington. The pdf of the brochure is attached.

The website is http://depts.washington.edu/sismid/

They have funds to support tuition waivers and some travel for students and postdocs.

3.2 2011 Atlantic Causal Conference

The 2011 Atlantic Causal Conference will take place at the University of Michigan School of Public Health in Ann Arbor, Michigan, Thursday May 19th and Friday May 20th. See http://www.sph.umich.edu/biostat/2011acic/index.html for

Contact: Mike Elliott at mrelliot@umich.edu or Ben Hansen at ben.hansen@umich.edu

3.3 Errata for Causality (2010) FYI, Cambridge University Press has come up with a new printing of my book Causality, which corrects a few errors in the 2009 edition. Please advise students to insist on a copy saying "reprinted 2010". If you have an older copy, you can find the corrections marked in red here: http://bayes.cs.ucla.edu/BOOK-09/errata_scanned_pages7-28-10.pdf

3.4 Lecture Slides available Slides of my lecture on "What's New in Causal Inference" can be viewed on my home page. http://bayes.cs.ucla.edu/jp_home.html 2nd line from the top. You are welcome to use them in any way you choose. But usage for a good cause is recommended.

Looking forward to postings from you, and may clarity prevail.

Judea Pearl

            UCLA

http://www.mii.ucla.edu/causality/. http://bayes.cs.ucla.edu/csl_papers.html

Thu, October 7, 2010 9:16 pm

To:  	cw@CS.UCLA.EDU
Cc:  	judea@CS.UCLA.EDU
Subject:  	Greeting from UCLA Causality-Blog
From:  	"Judea Pearl" <judea@CS.UCLA.EDU>

Dear colleagues in causality research,

This is a belated End-of-Summer greeting from the UCLA Causality-Blog, http://www.mii.ucla.edu/causality/. welcoming you back to an open discussion of causality-related issues

We open the new season with four new postings and three "hot" topics for discussion.

1. New postings:

The following papers and videos have been posted on our website.

1.1

Pearl and Bareinboim, "Transportability across studies: A formal approach", October 2010. http://www.cs.ucla.edu/~eb/r372.pdf The paper introduces a formal representation for encoding differences between populations and derives procedures for deciding whether (and how) causal effects in the target environment can be inferred from experimental findings in another.

1.2

J. Pearl, "The Causal Foundations of Structural Equation Modeling" August 2010. http://ftp.cs.ucla.edu/pub/stat_ser/r370.pdf The paper summarizes how traditional SEM methods can be enriched by modern advances in causal and counterfactual inference.

1.3

Greenland and Pearl, "Graphical Analysis of Full and Partial Covariate Adjustment" June 2010. http://ftp.cs.ucla.edu/pub/stat_ser/r369.pdf The paper answers a commonly asked question: Would adjustment for one variable reduce, increase, or leave unchanged the effect of a second variable on a third. A complete answer is given in terms of causal diagrams.

1.4

Videos of Symposium Lectures: All lectures given at the UCLA Symposium on Heuristics, Probability and Causality (March 12, 2010) are now available on you-tube.

Heuristics Session: http://bayes.cs.ucla.edu/TRIBUTE/videos-heuristics.htm

Probability Session: http://bayes.cs.ucla.edu/TRIBUTE/videos-prob-reasoning.htm

Causality Session: http://bayes.cs.ucla.edu/TRIBUTE/tribute-videos.htm

2. Discussions.

2.1

My open letter to Nancy Cartwright (posted June, 2010) has received Cartwright's response, http://www.mii.ucla.edu/causality/wp-content/uploads/2010/10/Cartwright2.doc accompanied by two of her recent addresses to the American Philosophical Society. http://www.mii.ucla.edu/causality/wp-content/uploads/2010/10/Cartwright1.doc http://www.mii.ucla.edu/causality/wp-content/uploads/2010/10/Cartwright3.doc

As you can see, Cartwright maintains that the

structure-based theory of counterfactuals does not answer the questions that policy makers wish answered, yet she does not provide (an example of) an input-output description of such a policy question. Can we conclude perhaps that, one we cast a problem in an input-output description it becomes solvable by the structure-based theory?? I think so.

2.2

This summer witnessed an interesting discussion on causal inference between two camps of economists: the "structuralists" and the "experimentalists," the former acknowledge their reliance on modelling assumptions, the latter pretend they dont. The discussion was published in the Spring 2010 issue of the Journal of Econometric Perspectives (vol 24 No 2), with Angrist and Pischke representing the "experimentalist" position and Leamer, Nevo and Keane defending the structural approach.

Worth reading.

My view: To the extent that the "experimental" approach is valid, it is merely a routine exercise in structural economics.

However, the philosophical basis of the "experimentalist" approach, as it is currently marketed, is misguided and potentially dangerous, for it takes semblance to the RCT ideal to be its main guiding principle. The fallibility of this paradigm has surfaced in a number of examples (e.g., http://ftp.cs.ucla.edu/pub/stat_ser/r363.pdf) and has given birth to a school of research that avoids making modelling assumptions transparent.

DRW: RCT= Randomized Controlled Trials. see Angus Deaton 2010 Instruments, Randomization, and Learning about Development. Journal of Economic Literature 48 (June 2010): 424–455. http:www.aeaweb.org/articles.php?doi=10.1257/jel.48.2.424

2.3

Another take on the "experimental - structural" debate is provided by Heckman, http://www.mii.ucla.edu/causality/wp-content/uploads/2010/10/heckman.pdf who reiterates the superiority of the structural over the Neyman-Rubin model, but stops short of identifying the key reason for that superiority.

This is strange because, after all, the

structural and potential-outcome approaches are logically equivalent, differing only in conceptual transparency (see Causality pages 230-34). If I had Heckman's platform, I would cite the inability of the "experimentalist" approach to encode counterfactual modeling assumptions in a transparent way, the bad practical advice that emerges from this deficiency (see http://ftp.cs.ucla.edu/pub/stat_ser/r363.pdf and the insecure, dismissive attitude that this deficiency engenders among its carriers (e.g., http://ftp.cs.ucla.edu/pub/stat_ser/r348.pdf).

As always, your thoughts are welcome and will surely be put into some good cause if conveyed to other blog readers.

Best

Judea

Tue, June 1, 2010 11:59 am

Greetings from UCLA Causality Blog From: "Judea Pearl" <judea@CS.UCLA.EDU> Date: Tue, June 1, 2010 11:59 am To: cw@CS.UCLA.EDU Options: View Full Header | View Printable Version | Download this as a file | Add to Address Book | View Message Details

Dear friends in causality,

Below are a few items you might find to be of some interest and possibly some challenge.

1. A new book containing a collection of recent articles on causation, some tutorial in nature, is now available from College Publications (2010.) Title: Heuristics, Probability and Causality, Editors: R. Dechter, H. Geffner and J. Halpern

For table of contents, preface and more information please click on: http://bayes.cs.ucla.edu/TRIBUTE/pearl-tribute2010.htm As you can see, I have had a natural indirect effect on the cover design, but zero controlled direct effect.

2. A symposium on causality and related topics by some of the contributors to "Heuristics, Probabilities and Causality" was held at UCLA on March 12. Videos of lectures, by: C. Hitchcock, S. Greenland, T. Richardson, J. Robins, R. Scheines, J. Tian, Y. Shoham and J. Pearl, can be viewed here: http://bayes.cs.ucla.edu/TRIBUTE/tribute-videos.htm Videos of additional lectures will be posted in the near future.

3. Recent entries on our Causality-Blog include: 3.1 An open letter from Judea Pearl to Nancy Cartwright concerning "Causal Pluralism", a topic central to a discussion of her book "Hunting Causes" which appeared recently in Economics and Philosophy 26:69-77. (Posted May 31, 2010), and 3.2 A lively discussion by T. Richardson, J. Robins and J. Pearl on the structure of the causal hierarchy and the scientific roll of untestable counterfactual assumptions. (Posted May 3 and May 15, 2010)

Both are posted on http://www.mii.ucla.edu/causality/.

4. A recent posting on my web-page is a paper titled: "The Mediation Formula: A guide to the assessment of causal pathways in non-linear models" which explains why traditional methods of mediation analysis yield distorted results when applied to discrete data, even when correct parametric models are assumed and all parameters are known precisely. The Mediation Formula circumvents these difficulties. http://ftp.cs.ucla.edu/pub/stat_ser/r363.pdf

5. Another posting of potential interest is Technical Report R-364, by T. Kyono (Master Thesis), titled: "Commentator: A Front-End User-Interface Module for Graphical and Structural Equation Modeling". It take a DAG as input and prints (1): all identifiable direct effects, (2) all identifiable causal effects, (3) all (minimal) sets of admissible covariates, (4) all instrumental variables, and (5) (almost) all testable implications of a model. The source code is available upon request. http://ftp.cs.ucla.edu/pub/stat_ser/r364.pdf

6. Finally, I have received inquiries regarding a slide that I used at NYU, in which an instrumental variable poses as an innocent confounder and, upon adjustment, amplifies, rather than reduces confounding bias. The moral of the story was (and is) that "outcome assignment" is safer to model than "treatment assignment". The pertinent paper is R-356, or http://ftp.cs.ucla.edu/pub/stat_ser/r356.pdf

7. As always, your thoughts are welcome and will surely be put into some good cause when conveyed to other blog readers.

Best =======Judea Pearl

       UCLA
      http://bayes.cs.ucla.edu/csl_papers.html

Spring March 24, 2011 Causality blog

Spring-time Greetings from the Causality blog, From: "Judea Pearl" <judea@CS.UCLA.EDU> Date: Thu, March 24, 2011 5:26 pm

Dear colleague in causality research,

This is an End-of-Winter Greeting from the UCLA Causality blog, welcoming you back to a spring-time discussion in causality-related issues.

This mesg contains 1. Topics under discussion 2. New results 3. Information on courses, lectures, and conferences.

1. Discussions inviting comments

1.1 Principal Stratification - A goal or a tool?' http://ftp.cs.ucla.edu/pub/stat_ser/r382.pdf Posted for discussion by the International Journal of Biostatistics (IJB), this paper questions whether studies based on Principal Stratification target quantities that researchers truly care about.

If you have comments, ideas or objections, you are invited to communicate them to the IJB's Editor, "Nicholas P. Jewell" <mm-11332-3261687@bepress.com> or/and, if you wish, cross-post them on this blog.

1.2 Comments and Controversies: Graphical models, potential outcomes and causal inference: Comment on Lindquist and Sobel" http://ftp.cs.ucla.edu/pub/stat_ser/r380.pdf

This note comments on a paper published in NeuroImage which argues (yes, again) that the potential outcome model is somehow superior, more rigorous or more principled than the structural models used in fMRI research. To further illuminate the logic of such claims I have added a section (4.4.2) in this paper: http://ftp.cs.ucla.edu/pub/stat_ser/r370.pdf which demonstrates how potential outcomes can be generated, on demand, from a simple structural model, and no one can tell where they came from. Enjoy.

1.3 "The Causal Mediation Formula - A practitioner guide to the assessment of causal pathways" http://ftp.cs.ucla.edu/pub/stat_ser/r379.pdf

This paper present mediation analysis to researchers in the tradition of Baron and Kenny (1986), and shows through examples how "the percentage explained by mediation" and "the percentage owed to mediation" are estimated in nonlinear models with both continuous and categorical variables.

1.4 Simpson's Paradox Sander Greenland brought to my attention a recent paper in Synthese (Sept. 28, 2010) claiming that Simpson's paradox is NOT rooted in causal, but in some other kind of illusion. I remain convinced of the former, and have accordingly modified the Simpson Paradox entry in Wikipedia to reinforce the causal-illusion theory. You might wish to add your take on the subject.

1.5 The ETT Paradox (or, the curse of free will) This paradox would be appreciated by those who are fascinated, like me, by our ability to determine, from data alone, if one would have been better off acting differently from the way one actually did. This can lead to a cycle of inevitable regret, and provoke some naughty thoughts. http://ftp.cs.ucla.edu/pub/stat_ser/r375.pdf

2. New Results

A newly posted paper, "On the Control of Selection Bias" http://ftp.cs.ucla.edu/pub/stat_ser/r381.pdf gives graphical and algebraic conditions for the removal of selection bias and the recovery of covariate-specific effect measures.

2.2 A new section (Section 5) in http://ftp.cs.ucla.edu/pub/stat_ser/r372.pdf generalizes the concept of transportability from experimental to observational studies,and shows how one can avoid re-learning things from scratch when moving to a new population, new domain, or a new environment.

2.3 After months of struggling with the literature of "surrogate endpoints" we feel that we now have a fairly satisfactory theory of surrogacy, It is based on the idea that a surrogate should serve not merely as a good predictor of outcomes, but also as ROBUST predictor of effects in the face of changing external conditions. See section 6 in http://ftp.cs.ucla.edu/pub/stat_ser/r372.pdf

3. Courses, Lectures and Conferences, 3.1 Causal Inference Course

Thomas Richardson and Michael Hudgens are once again teaching Causal Inference June 13-15, 2011 in the Summer Institute here at U Washington. The pdf of the brochure is attached.

The website is http://depts.washington.edu/sismid/

They have funds to support tuition waivers and some travel for students and postdocs.

3.2 2011 Atlantic Causal Conference

The 2011 Atlantic Causal Conference will take place at the University of Michigan School of Public Health in Ann Arbor, Michigan, Thursday May 19th and Friday May 20th. See http://www.sph.umich.edu/biostat/2011acic/index.html for

Contact: Mike Elliott at mrelliot@umich.edu or Ben Hansen at ben.hansen@umich.edu

3.3 Errata for Causality (2010) FYI, Cambridge University Press has come up with a new printing of my book Causality, which corrects a few errors in the 2009 edition. Please advise students to insist on a copy saying "reprinted 2010". If you have an older copy, you can find the corrections marked in red here: http://bayes.cs.ucla.edu/BOOK-09/errata_scanned_pages7-28-10.pdf

3.4 Lecture Slides available Slides of my lecture on "What's New in Causal Inference" can be viewed on my home page. http://bayes.cs.ucla.edu/jp_home.html 2nd line from the top. You are welcome to use them in any way you choose. But usage for a good cause is recommended.

Looking forward to postings from you, and may clarity prevail.

Judea Pearl

            UCLA

http://www.mii.ucla.edu/causality/. http://bayes.cs.ucla.edu/csl_papers.html

Other

Links

Judea Pearl

UCLA Causality-Blog

Contents

Myth, Confusion, and Science in Causal Analysis

March 24, 2011

Spring March 24, 2011 Causality blog

Thu, October 7, 2010 9:16 pm

1. New postings:

2. Discussions.

Tue, June 1, 2010 11:59 am

Spring March 24, 2011 Causality blog

Other

Links

Judea Pearl

Views

Personal tools

Navigation

Search

Toolbox