Please use this identifier to cite or link to this item:
https://hdl.handle.net/10419/21122
Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Augurzky, Boris | en |
dc.contributor.author | Schmidt, Christoph M. | en |
dc.date.accessioned | 2009-01-28T16:19:38Z | - |
dc.date.available | 2009-01-28T16:19:38Z | - |
dc.date.issued | 2001 | - |
dc.identifier.uri | http://hdl.handle.net/10419/21122 | - |
dc.description.abstract | Propensity score matching is a prominent strategy to reduce imbalance in observational studies.However, if imbalance is considerable and the control reservoir is small, either one has to matchone control to several treated units or, alternatively, discard many treated persons. The firststrategy tends to increase standard errors of the estimated treatment effects while the secondmight produce a matched sample that is not anymore representative of the original one. As analternative approach, this paper argues to carefully reconsider the selection equation upon whichthe propensity score estimates are based. Often, all available variables that rule the selectionprocess are included into the selection equation. Yet, it would suffice to concentrate on onlythose exhibiting a large impact on the outcome under scrutiny, as well. This would introducemore stochastic noise making treatment and comparison group more similar. We assess theadvantages and disadvantages of the latter approach in a simulation study. | en |
dc.language.iso | eng | en |
dc.publisher | |aInstitute for the Study of Labor (IZA) |cBonn | en |
dc.relation.ispartofseries | |aIZA Discussion Papers |x271 | en |
dc.subject.jel | C14 | en |
dc.subject.jel | C15 | en |
dc.subject.ddc | 330 | en |
dc.subject.keyword | Estimation of the propensity score | en |
dc.subject.keyword | balance of relevant covariates | en |
dc.subject.keyword | simulation study | en |
dc.subject.stw | Stichprobenverfahren | en |
dc.subject.stw | Mikroökonometrie | en |
dc.subject.stw | Theorie | en |
dc.title | The Propensity Score: A Means to An End | - |
dc.type | Working Paper | en |
dc.identifier.ppn | 843638524 | en |
dc.rights | http://www.econstor.eu/dspace/Nutzungsbedingungen | en |
Files in This Item:
Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.