umu.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Covariate selection and propensity score specification in causal inference
Umeå University, Faculty of Social Sciences, Department of Statistics.
2008 (English)Doctoral thesis, comprehensive summary (Other academic)
Abstract [en]

This thesis makes contributions to the statistical research field of causal inference in observational studies. The results obtained are directly applicable in many scientific fields where effects of treatments are investigated and yet controlled experiments are difficult or impossible to implement.

In the first paper we define a partially specified directed acyclic graph (DAG) describing the independence structure of the variables under study. Using the DAG we show that given that unconfoundedness holds we can use the observed data to select minimal sets of covariates to control for. General covariate selection algorithms are proposed to target the defined minimal subsets.

The results of the first paper are generalized in Paper II to include the presence of unobserved covariates. Morevoer, the identification assumptions from the first paper are relaxed.

To implement the covariate selection without parametric assumptions we propose in the third paper the use of a model-free variable selection method from the framework of sufficient dimension reduction. By simulation the performance of the proposed selection methods are investigated. Additionally, we study finite sample properties of treatment effect estimators based on the selected covariate sets.

In paper IV we investigate misspecifications of parametric models of a scalar summary of the covariates, the propensity score. Motivated by common model specification strategies we describe misspecifications of parametric models for which unbiased estimators of the treatment effect are available. Consequences of the misspecification for the efficiency of treatment effect estimators are also studied.

Place, publisher, year, edition, pages
Umeå: Statistik , 2008.
Series
Statistical studies, ISSN 1100-8989 ; 38
Keyword [en]
Covariate selection, graphical models, matching, observational studies, treatment effects, unconfoundedness
National Category
Probability Theory and Statistics
Identifiers
URN: urn:nbn:se:umu:diva-1688ISBN: 978-91-7264-564-6 (print)OAI: oai:DiVA.org:umu-1688DiVA: diva2:141832
Public defence
2008-09-12, Hörsal D, Samhällsvetarhuset, Umeå universitet, 901 87, Umeå, 13:15
Opponent
Supervisors
Available from: 2008-05-30 Created: 2008-05-30 Last updated: 2011-03-17Bibliographically approved
List of papers
1. Covariate selection for non-parametric estimation of treatment effects
Open this publication in new window or tab >>Covariate selection for non-parametric estimation of treatment effects
Manuscript (Other academic)
Identifiers
urn:nbn:se:umu:diva-3287 (URN)
Available from: 2008-05-30 Created: 2008-05-30 Last updated: 2010-01-13Bibliographically approved
2. Identification of minimal sets of covariates for the non-parametric estimation of an average treatment effect
Open this publication in new window or tab >>Identification of minimal sets of covariates for the non-parametric estimation of an average treatment effect
Article in journal (Refereed) Submitted
Identifiers
urn:nbn:se:umu:diva-3288 (URN)
Available from: 2008-05-30 Created: 2008-05-30Bibliographically approved
3. Model-free variable selection in causal inference
Open this publication in new window or tab >>Model-free variable selection in causal inference
Manuscript (Other academic)
Identifiers
urn:nbn:se:umu:diva-3289 (URN)
Available from: 2008-05-30 Created: 2008-05-30 Last updated: 2010-01-13Bibliographically approved
4. Propensity score model specification for estimation of average treatment effects
Open this publication in new window or tab >>Propensity score model specification for estimation of average treatment effects
2010 (English)In: Journal of Statistical Planning and Inference, ISSN 0378-3758, Vol. 140, no 7, 1948-1956 p.Article in journal (Refereed) Published
Abstract [en]

Treatment effect estimators that utilize the propensity score as a balancing score, e.g., matching and blocking estimators are robust to misspecifications of the propensity score model when the misspecification is a balancing score. Such misspecifications arise from using the balancing property of the propensity score in the specification procedure. Here, we study misspecifications of a parametric propensity score model written as a linear predictor in a strictly monotonic function, e.g. a generalized linear model representation. Under mild assumptions we show that for misspecifications, such as not adding enough higher order terms or choosing the wrong link function, the true propensity score is a function of the misspecified model. Hence, the latter does not bring bias to the treatment effect estimator. It is also shown that a misspecification of the propensity score does not necessarily lead to less efficient estimation of the treatment effect. The results of the paper are highlighted in simulations where different misspecifications are studied.

Place, publisher, year, edition, pages
Amsterdam: Elsevier, 2010
Keyword
Causal effect, Observational studies, Matching, Propensity score
National Category
Probability Theory and Statistics
Research subject
Statistics
Identifiers
urn:nbn:se:umu:diva-3290 (URN)10.1016/j.jspi.2010.01.033 (DOI)000276369000028 ()
Available from: 2008-05-30 Created: 2008-05-30 Last updated: 2010-04-22Bibliographically approved

Open Access in DiVA

fulltext(214 kB)2128 downloads
File information
File name FULLTEXT01.pdfFile size 214 kBChecksum SHA-1
aad9c76dccc9826b1a97b9e3b83d98cb5491c45f899de48e4ff0b58c4db54cebafbbb5f2
Type fulltextMimetype application/pdf

Authority records BETA

Waernbaum, Ingeborg

Search in DiVA

By author/editor
Waernbaum, Ingeborg
By organisation
Department of Statistics
Probability Theory and Statistics

Search outside of DiVA

GoogleGoogle Scholar
Total: 2128 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

isbn
urn-nbn

Altmetric score

isbn
urn-nbn
Total: 1712 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf