Using Descendants as Instrumental Variables for the Identification of Direct Causal Effects in Linear SEMs

Hei Chan, Manabu Kuroki
Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, PMLR 9:73-80, 2010.

Abstract

In this paper, we present an extended set of graphical criteria for the identification of direct causal effects in linear Structural Equation Models (SEMs). Previous methods of graphical identification of direct causal effects in linear SEMs include methods such as the single-door criterion, the instrumental variable and the IV-pair, and the accessory set. However, there remain graphical models where a direct causal effect can be identified and these graphical criteria all fail. As a result, we introduce a new set of graphical criteria which uses descendants of either the cause variable or the effect variable as “path-specific instrumental variables” for the identification of the direct causal effect as long as certain conditions are satisfied. These conditions are based on edge removal and the existing graphical criteria of instrumental variables, and the identifiability of certain other total effects, and thus can be easily checked.

Cite this Paper

BibTeX
@InProceedings{pmlr-v9-chan10a, title = {Using Descendants as Instrumental Variables for the Identification of Direct Causal Effects in Linear SEMs}, author = {Chan, Hei and Kuroki, Manabu}, booktitle = {Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics}, pages = {73--80}, year = {2010}, editor = {Teh, Yee Whye and Titterington, Mike}, volume = {9}, series = {Proceedings of Machine Learning Research}, address = {Chia Laguna Resort, Sardinia, Italy}, month = {13--15 May}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v9/chan10a/chan10a.pdf}, url = {https://proceedings.mlr.press/v9/chan10a.html}, abstract = {In this paper, we present an extended set of graphical criteria for the identification of direct causal effects in linear Structural Equation Models (SEMs). Previous methods of graphical identification of direct causal effects in linear SEMs include methods such as the single-door criterion, the instrumental variable and the IV-pair, and the accessory set. However, there remain graphical models where a direct causal effect can be identified and these graphical criteria all fail. As a result, we introduce a new set of graphical criteria which uses descendants of either the cause variable or the effect variable as “path-specific instrumental variables” for the identification of the direct causal effect as long as certain conditions are satisfied. These conditions are based on edge removal and the existing graphical criteria of instrumental variables, and the identifiability of certain other total effects, and thus can be easily checked.} }
Endnote
%0 Conference Paper %T Using Descendants as Instrumental Variables for the Identification of Direct Causal Effects in Linear SEMs %A Hei Chan %A Manabu Kuroki %B Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2010 %E Yee Whye Teh %E Mike Titterington %F pmlr-v9-chan10a %I PMLR %P 73--80 %U https://proceedings.mlr.press/v9/chan10a.html %V 9 %X In this paper, we present an extended set of graphical criteria for the identification of direct causal effects in linear Structural Equation Models (SEMs). Previous methods of graphical identification of direct causal effects in linear SEMs include methods such as the single-door criterion, the instrumental variable and the IV-pair, and the accessory set. However, there remain graphical models where a direct causal effect can be identified and these graphical criteria all fail. As a result, we introduce a new set of graphical criteria which uses descendants of either the cause variable or the effect variable as “path-specific instrumental variables” for the identification of the direct causal effect as long as certain conditions are satisfied. These conditions are based on edge removal and the existing graphical criteria of instrumental variables, and the identifiability of certain other total effects, and thus can be easily checked.
RIS
TY - CPAPER TI - Using Descendants as Instrumental Variables for the Identification of Direct Causal Effects in Linear SEMs AU - Hei Chan AU - Manabu Kuroki BT - Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics DA - 2010/03/31 ED - Yee Whye Teh ED - Mike Titterington ID - pmlr-v9-chan10a PB - PMLR DP - Proceedings of Machine Learning Research VL - 9 SP - 73 EP - 80 L1 - http://proceedings.mlr.press/v9/chan10a/chan10a.pdf UR - https://proceedings.mlr.press/v9/chan10a.html AB - In this paper, we present an extended set of graphical criteria for the identification of direct causal effects in linear Structural Equation Models (SEMs). Previous methods of graphical identification of direct causal effects in linear SEMs include methods such as the single-door criterion, the instrumental variable and the IV-pair, and the accessory set. However, there remain graphical models where a direct causal effect can be identified and these graphical criteria all fail. As a result, we introduce a new set of graphical criteria which uses descendants of either the cause variable or the effect variable as “path-specific instrumental variables” for the identification of the direct causal effect as long as certain conditions are satisfied. These conditions are based on edge removal and the existing graphical criteria of instrumental variables, and the identifiability of certain other total effects, and thus can be easily checked. ER -
APA
Chan, H. & Kuroki, M.. (2010). Using Descendants as Instrumental Variables for the Identification of Direct Causal Effects in Linear SEMs. Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 9:73-80 Available from https://proceedings.mlr.press/v9/chan10a.html.

Related Material