Causal Analysis in Theory and Practice

Dear friends in causality,

This greeting from UCLA Causality blog contains:
A. News items concerning causality research,
B. New postings, publications, slides and videos,
C. New questions and some answers.

1. Journal of Causal Inference – Vol. 1, Issue 2
The second issue of the Journal of Causal Inference is on its way, and an on-line posting date has been set for December 23 2013. The first issue, can be viewed here:
http://www.degruyter.com/view/j/jci.2013.1.issue-1/issue-files/jci.2013.1.issue-1.xml
or here:
http://www.degruyter.com/view/j/jci

As always, submissions are welcome on all aspects of causal analysis, especially those deemed heretical.

2. Causality book – new printing
The 3rd printing of Causality (2009, 2nd ed.) is finally out (as of Sept. 1), corrected and improved.

If you have an older printing and do not wish to buy another copy, all changes are marked in red here:
http://bayes.cs.ucla.edu/BOOK-09/causality2-errata-updated7_3_13.pdf

3. Special Issue on Counterfactuals
The latest issue of Cognitive Science is dedicated to counterfactual reasoning. Edited by Stephen Sloman, the Table of Content and on-line version are available here:
http://onlinelibrary.wiley.com/doi/10.1111/cogs.2013.37.issue-6/issuetoc

4. A Strange article in Science Magazine.
Two articles in Science Magazine and Nature were brought to our attention:
http://www.sciencemag.org/content/338/6106/496.abstract
http://www.nature.com/nphys/journal/v8/n12/full/nphys2497.html
The author claims that causal effects can be inferred from correlation, using an extended version of Granger Causality.
To me this sounds like squaring the circle; perhaps one of our readers can illuminate us.

5. Calls for papers on Causality
5.1 Isabell Guyon sent us a call for papers for Special Topic of JMLR on Causality and Experimental Design. See:
http://clopinet.com/isabelle/Projects/NIPS2013/Causality_Special_Topic_2014.html

5.2 The ACM TIST journal is planning a Special Issue on Causal Discovery and Inference, and has issued a call for papers.
See: http://tist.acm.org/CFPs/TIST-SI-CDI.html

6. Dennis Lindley, dead at 90
On a sad note, Dennis Lindley, a pioneer in Bayesian inference died last week, at age 90.
Lindley brought the “SEEING vs. DOING” distinction to the attention of the statistics community:
http://bayes.cs.ucla.edu/BOOK-2K/lindley-rev.pdf
He also adapted the causal interpretation of Simpson’s paradox ahead of his peers.
We will miss his intellect, curiosity and integrity.
A true gentleman.

7. New postings on this blog.
Since our last greetings, the following items were posted on this blog (you can view them below).
Aug. 9 , 2013, Larry Wasserman on JSM 2013
Oct. 26, 2013, Comments on Kenny’s Summary of Causal Mediation
Nov. 10, 2013, On Heckman and Pinto
Nov. 19, 2013, The Key to Understanding Mediation
Dec. 14 2013, “But where does the graph come from?”

8. New slides and videos available
* Slides of the tutorial on “Causes and Counterfactuals” preseted at NIPS-2013 (by Pearl and Bareinboim) are available here:
http://ftp.cs.ucla.edu/pub/stat_ser/nips-dec2013-pearl-bareinboim-tutorial.pdf
http://www.cs.ucla.edu/~eb/nips-dec2013-pearl-bareinboim-tutorial-full.pdf

* Video of an introductory lecture presented to economists (at Stanford) is available here:
http://gsb-mediasite.stanford.edu/mediasite/Play/4ee5b390c7ef456c9ba0b11d1a519d4b1d

9. New scientific questions and some answers
There are new postings on my home page
http://bayes.cs.ucla.edu/csl_papers.html
which might earn your attention. Among them:

420 – J. Pearl, “Reflections on Heckman and Pinto’s ‘Causal Analysis after Haavelmo”,
where I defend Haavelmo’s original theory of intervention against a Fisherian surrogate proposed in Heckman and Pinto (2013).
http://ftp.cs.ucla.edu/pub/stat_ser/r420.pdf

419 – Bareinboin, Lee, Honavar, Pearl “Transportability from Multiple Enironments with Limited Experiments”
where we ask (and answer) whether it is possible to combine experimental findings from many heterogeneous studies to get what we need.
http://ftp.cs.ucla.edu/pub/stat_ser/r419.pdf

417 – Pearl and Mohan “Recoverability and Testability of Missing Data.”
where we explain missing-data problems to the uninitiated using graphical models, and illustrate the concepts of recoverability and testability.
http://ftp.cs.ucla.edu/pub/stat_ser/r417.pdf

416 – J. Pearl “The Mathematics of Causal Inference”
A summary of a Lecture given at JSM-2013, which compiles the main mathematical results in causal inference.
http://ftp.cs.ucla.edu/pub/stat_ser/r416.pdf

414 – J. Pearl, “Understanding Simpson’s Paradox”
where I introduce a guessing game that exhibits perpetual reversals and argue that the paradox can safely be titled: “resolved”
http://ftp.cs.ucla.edu/pub/stat_ser/r414.pdf

410 – Mohan, Pearl and Tian “Graphical Models for Inference with Missing Data” (Newly Revised)
where we take a fresh look at missing data problems from a causal inference perspective and propose a new taxonomy for misssing data mechanisms.
http://ftp.cs.ucla.edu/pub/stat_ser/r410.pdf

363 – Pearl and Paz, “Confounding Equivalence in Causal Analysis” (Revised Oct 2013)
We ask: When would an adjustment for T introduce the same bias as an adjustment for Z, and we answer it by extending the results of the 2010 version of this paper.
http://ftp.cs.ucla.edu/pub/stat_ser/r343w.pdf

Wishing you a happy and productive new year,
Judea Pearl

Judea Pearl Writes:

Researchers using causal diagrams have surely noticed that, despite a tremendous progress in causal modeling in the past three decades, editors and reviewers persist in raising questions about the usefulness of causal diagrams, noting that their structure is based largely on untested or untestable assumptions and, hence, that they could not serve as a basis for policy evaluation or personal decision-making.

Questions such as:
“What if I do not have the graph?”
“What if I am not sure about the absence of this or that confounder?”
“What if I do not have the scientific knowledge required to construct the graph,”
and more, come up again and again, especially from editors and reviewers who have not had first-hand experience in causal inference research.

As a service to readers of this blog, I would like to share the way I usually answer such questions.

To continue to the full post, click here.