Please use this identifier to cite or link to this item:
https://hdl.handle.net/10419/31063
Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Holzmann, Hajo | en |
dc.contributor.author | Min, Aleksey | en |
dc.contributor.author | Czado, Claudia | en |
dc.date.accessioned | 2006-09-07 | - |
dc.date.accessioned | 2010-05-14T10:10:02Z | - |
dc.date.available | 2010-05-14T10:10:02Z | - |
dc.date.issued | 2006 | - |
dc.identifier.pi | doi:10.5282/ubm/epub.1846 | en |
dc.identifier.pi | urn:nbn:de:bvb:19-epub-1846-3 | en |
dc.identifier.uri | http://hdl.handle.net/10419/31063 | - |
dc.description.abstract | A new method for testing linear restrictions in linear regression models is suggested. It allows to validate the linear restriction, up to a specified approximation error and with a specified error probability. The test relies on asymptotic normality of the test statistic, and therefore normality of the errors in the regression model is not required. In a simulation study the performance of the suggested method for model selection purposes, as compared to standard model selection criteria and the t-test, is examined. As an illustration we analyze the US college spending data from 1994. | en |
dc.language.iso | eng | en |
dc.publisher | |aLudwig-Maximilians-Universität München, Sonderforschungsbereich 386 - Statistische Analyse diskreter Strukturen |cMünchen | en |
dc.relation.ispartofseries | |aDiscussion Paper |x478 | en |
dc.subject.ddc | 519 | en |
dc.subject.keyword | asymptotic normality | en |
dc.subject.keyword | linear regression | en |
dc.subject.keyword | model selection | en |
dc.subject.keyword | model validation | en |
dc.title | Validating linear restrictions in linear regression models with general error structure | - |
dc.type | |aWorking Paper | en |
dc.identifier.ppn | 51717104X | 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.