Show simple item record

dc.contributor.authorSteele, M
dc.contributor.authorChaseling, J
dc.contributor.authorHurst, C
dc.contributor.editorZerger, A
dc.contributor.editorArgent, RM
dc.date.accessioned2018-03-07T04:38:12Z
dc.date.available2018-03-07T04:38:12Z
dc.date.issued2005
dc.date.modified2014-06-30T05:03:40Z
dc.identifier.refurihttp://www/mssanz.org.au/modsim05/papers/steele.pdf
dc.identifier.urihttp://hdl.handle.net/10072/2725
dc.description.abstractThe use of goodness-of-fit test statistics for discrete or categorical data is widespread throughout the research community with the ChiSquare the most popular when a researcher aims to determine if observed categorical data differs from a hypothesized multinomial distribution. Even for ordinal categorical data, the use of empirical distribution function (EDF) test statistics such as the Kolmogorov-Smirnov, the three Cramér-von Mises (A2 , W2 and U2 as defined below) and various modifications of these are limited in the literature. Power studies of the EDF type test statistics are even more limited. This paper compares the simulated power of the three Cramér-von Mises test statistics with that of the Chi-Square test statistic for a uniform null hypothesis against a variety of alternative distributions which are summarized in Figure 1. Recommendations are made on which is the most powerful test statistic for the predefined alternative distributions. The results of the simulated power studies in this paper lead to the following general recommendations: • For trend type alternatives A2 and W2 appear much more powerful than U2 and χ2. (See Figure 2 for a uniform null against a decreasing trend alternative distribution). • For all the other investigated alternative distributions U2 and χ2 appear much more powerful than A2 and W2. (See Figure 3 for a uniform null against a leptokurtic type alternative distribution).
dc.description.peerreviewedYes
dc.description.publicationstatusYes
dc.format.extent324524 bytes
dc.format.mimetypeapplication/pdf
dc.languageEnglish
dc.language.isoeng
dc.publisherModelling and Simulation Society of Australia and New Zealand
dc.publisher.placeAustralia
dc.publisher.urihttp://www.mssanz.org.au/modsim05/index.htm
dc.relation.ispartofstudentpublicationN
dc.relation.ispartofconferencenameInternational Congress on Modelling and Simulation (MODSIM05)
dc.relation.ispartofconferencetitleMODSIM 2005: INTERNATIONAL CONGRESS ON MODELLING AND SIMULATION: ADVANCES AND APPLICATIONS FOR MANAGEMENT AND DECISION MAKING
dc.relation.ispartofdatefrom2005-12-12
dc.relation.ispartofdateto2005-12-15
dc.relation.ispartoflocationMelbourne, AUSTRALIA
dc.relation.ispartofpagefrom1300
dc.relation.ispartofpagefrom5 pages
dc.relation.ispartofpageto1304
dc.relation.ispartofpageto5 pages
dc.rights.retentionY
dc.subject.fieldofresearchcode230204
dc.titleSimulated Power of the Discrete Cramer-von Mises Goodness-of-Fit Tests
dc.typeConference output
dc.type.descriptionE1 - Conferences
dc.type.codeE - Conference Publications
gro.rights.copyright© 2005 Modellling & Simulation Society of Australia & New Zealand. The attached file is reproduced here in accordance with the copyright policy of the publisher. For information about this conference please refer to the conference’s website or contact the authors.
gro.date.issued2005
gro.hasfulltextFull Text
gro.griffith.authorChaseling, Janet
gro.griffith.authorSteele, Mike
gro.griffith.authorHurst, Cameron P.


Files in this item

This item appears in the following Collection(s)

  • Conference outputs
    Contains papers delivered by Griffith authors at national and international conferences.

Show simple item record