What counterfactuals can be tested

Ilya Shpitser, Judea Pearl

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Counterfactual statements, e.g., "my headache would be gone had I taken an aspirin" are central to scientific discourse, and are formally interpreted as statements derived from "alternative worlds". However, since they invoke hypothetical states of affairs, often incompatible with what is actually known or observed, testing counterfactuals is fraught with conceptual and practical difficulties. In this paper, we provide a complete characterization of "testable counterfactuals," namely, counterfactual statements whose probabilities can be inferred from physical experiments. We provide complete procedures for discerning whether a given counterfactual is testable and, if so, expressing its probability in terms of experimental data.

Original languageEnglish (US)
Title of host publicationProceedings of the 23rd Conference on Uncertainty in Artificial Intelligence, UAI 2007
Pages352-359
Number of pages8
StatePublished - Dec 1 2007
Externally publishedYes
Event23rd Conference on Uncertainty in Artificial Intelligence, UAI 2007 - Vancouver, BC, Canada
Duration: Jul 19 2007Jul 22 2007

Other

Other23rd Conference on Uncertainty in Artificial Intelligence, UAI 2007
CountryCanada
CityVancouver, BC
Period7/19/077/22/07

Fingerprint

Testing
Experiments

ASJC Scopus subject areas

  • Artificial Intelligence

Cite this

Shpitser, I., & Pearl, J. (2007). What counterfactuals can be tested. In Proceedings of the 23rd Conference on Uncertainty in Artificial Intelligence, UAI 2007 (pp. 352-359)

What counterfactuals can be tested. / Shpitser, Ilya; Pearl, Judea.

Proceedings of the 23rd Conference on Uncertainty in Artificial Intelligence, UAI 2007. 2007. p. 352-359.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Shpitser, I & Pearl, J 2007, What counterfactuals can be tested. in Proceedings of the 23rd Conference on Uncertainty in Artificial Intelligence, UAI 2007. pp. 352-359, 23rd Conference on Uncertainty in Artificial Intelligence, UAI 2007, Vancouver, BC, Canada, 7/19/07.
Shpitser I, Pearl J. What counterfactuals can be tested. In Proceedings of the 23rd Conference on Uncertainty in Artificial Intelligence, UAI 2007. 2007. p. 352-359
Shpitser, Ilya ; Pearl, Judea. / What counterfactuals can be tested. Proceedings of the 23rd Conference on Uncertainty in Artificial Intelligence, UAI 2007. 2007. pp. 352-359
@inproceedings{963fca253bac435ea57b07462b52ac1f,
title = "What counterfactuals can be tested",
abstract = "Counterfactual statements, e.g., {"}my headache would be gone had I taken an aspirin{"} are central to scientific discourse, and are formally interpreted as statements derived from {"}alternative worlds{"}. However, since they invoke hypothetical states of affairs, often incompatible with what is actually known or observed, testing counterfactuals is fraught with conceptual and practical difficulties. In this paper, we provide a complete characterization of {"}testable counterfactuals,{"} namely, counterfactual statements whose probabilities can be inferred from physical experiments. We provide complete procedures for discerning whether a given counterfactual is testable and, if so, expressing its probability in terms of experimental data.",
author = "Ilya Shpitser and Judea Pearl",
year = "2007",
month = "12",
day = "1",
language = "English (US)",
isbn = "0974903930",
pages = "352--359",
booktitle = "Proceedings of the 23rd Conference on Uncertainty in Artificial Intelligence, UAI 2007",

}

TY - GEN

T1 - What counterfactuals can be tested

AU - Shpitser, Ilya

AU - Pearl, Judea

PY - 2007/12/1

Y1 - 2007/12/1

N2 - Counterfactual statements, e.g., "my headache would be gone had I taken an aspirin" are central to scientific discourse, and are formally interpreted as statements derived from "alternative worlds". However, since they invoke hypothetical states of affairs, often incompatible with what is actually known or observed, testing counterfactuals is fraught with conceptual and practical difficulties. In this paper, we provide a complete characterization of "testable counterfactuals," namely, counterfactual statements whose probabilities can be inferred from physical experiments. We provide complete procedures for discerning whether a given counterfactual is testable and, if so, expressing its probability in terms of experimental data.

AB - Counterfactual statements, e.g., "my headache would be gone had I taken an aspirin" are central to scientific discourse, and are formally interpreted as statements derived from "alternative worlds". However, since they invoke hypothetical states of affairs, often incompatible with what is actually known or observed, testing counterfactuals is fraught with conceptual and practical difficulties. In this paper, we provide a complete characterization of "testable counterfactuals," namely, counterfactual statements whose probabilities can be inferred from physical experiments. We provide complete procedures for discerning whether a given counterfactual is testable and, if so, expressing its probability in terms of experimental data.

UR - http://www.scopus.com/inward/record.url?scp=77649296605&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77649296605&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:77649296605

SN - 0974903930

SN - 9780974903934

SP - 352

EP - 359

BT - Proceedings of the 23rd Conference on Uncertainty in Artificial Intelligence, UAI 2007

ER -