File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: A cautionary note on the use of generalized Cook-type measures

TitleA cautionary note on the use of generalized Cook-type measures
Authors
Fung, WK
KeywordsAdding-back measures
Influential observations
Masking
Multiple deletion measures
Outliers
Regression diagnostics
Issue Date1995
PublisherElsevier BV. The Journal's web site is located at http://www.elsevier.com/locate/csda
Citation
Computational Statistics And Data Analysis, 1995, v. 19 n. 3, p. 321-326 How to Cite?
DOI: http://dx.doi.org/10.1016/0167-9473(93)E0024-X
AbstractThe Cook statistic is a well-known and useful single deletion diagnostic measure for the detection of influential observations in regression. However, the single deletion method may suffer from the masking effect problem. It is generally believed that the multiple deletion Cook-type measures can solve the masking effect problem. We explain and illustrate that this is quite often not true. As the Cook-type measures are so commonly used in applied work, in this note we emphasize that we have to be cautious in interpreting the results obtained by the generalized measures. A possible remedy is discussed. © 1995.
Persistent Identifierhttp://hdl.handle.net/10722/82903
ISSN
0167-9473
2021 Impact Factor: 2.035
2020 SCImago Journal Rankings: 1.093
ISI Accession Number ID
WOS:A1995QK17500005

 

DC FieldValueLanguage
dc.contributor.authorFung, WKen_HK
dc.date.accessioned2010-09-06T08:34:43Z-
dc.date.available2010-09-06T08:34:43Z-
dc.date.issued1995en_HK
dc.identifier.citationComputational Statistics And Data Analysis, 1995, v. 19 n. 3, p. 321-326en_HK
dc.identifier.issn0167-9473en_HK
dc.identifier.urihttp://hdl.handle.net/10722/82903-
dc.description.abstractThe Cook statistic is a well-known and useful single deletion diagnostic measure for the detection of influential observations in regression. However, the single deletion method may suffer from the masking effect problem. It is generally believed that the multiple deletion Cook-type measures can solve the masking effect problem. We explain and illustrate that this is quite often not true. As the Cook-type measures are so commonly used in applied work, in this note we emphasize that we have to be cautious in interpreting the results obtained by the generalized measures. A possible remedy is discussed. © 1995.en_HK
dc.languageengen_HK
dc.publisherElsevier BV. The Journal's web site is located at http://www.elsevier.com/locate/csdaen_HK
dc.relation.ispartofComputational Statistics and Data Analysisen_HK
dc.rightsComputational Statistics & Data Analysis. Copyright © Elsevier BV.en_HK
dc.subjectAdding-back measuresen_HK
dc.subjectInfluential observationsen_HK
dc.subjectMaskingen_HK
dc.subjectMultiple deletion measuresen_HK
dc.subjectOutliersen_HK
dc.subjectRegression diagnosticsen_HK
dc.titleA cautionary note on the use of generalized Cook-type measuresen_HK
dc.typeArticleen_HK
dc.identifier.openurlhttp://library.hku.hk:4550/resserv?sid=HKU:IR&issn=0167-9473&volume=19&spage=321&epage=326&date=1995&atitle=A+cautionary+note+on+the+use+of+generalized+Cook-type+measuresen_HK
dc.identifier.emailFung, WK: wingfung@hku.hken_HK
dc.identifier.authorityFung, WK=rp00696en_HK
dc.description.naturelink_to_subscribed_fulltext-
dc.identifier.doi10.1016/0167-9473(93)E0024-X-
dc.identifier.scopuseid_2-s2.0-58149211372en_HK
dc.identifier.hkuros3973en_HK
dc.identifier.volume19en_HK
dc.identifier.issue3en_HK
dc.identifier.spage321en_HK
dc.identifier.epage326en_HK
dc.identifier.isiWOS:A1995QK17500005-
dc.publisher.placeNetherlandsen_HK
dc.identifier.scopusauthoridFung, WK=13310399400en_HK
dc.identifier.issnl0167-9473-

Export via OAI-PMH Interface in XML Formats

OR

Export to Other Non-XML Formats