OPTIMAL DESIGN WHEN OUTCOME VALUES ARE NOT MISSING AT RANDOM

The presence of missing values complicates statistical analyses. In design of experiments, missing values are particularly problematic when constructing optimal designs, as it is not known which values are missing at the design stage. When data are missing at random it is possible to incorporate thi...

Full description

Saved in:
Bibliographic Details
Published in:Statistica Sinica 2018-10, Vol.28 (4), p.1821-1838
Main Authors: Lee, Kim May, Mitra, Robin, Biedermann, Stefanie
Format: Article
Language:English
Citations: Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by cdi_FETCH-LOGICAL-c301t-eff2dd898ecc600b398dab2f62319a308241eee50ff5ce3a0f2b035ee94eaf703
cites
container_end_page 1838
container_issue 4
container_start_page 1821
container_title Statistica Sinica
container_volume 28
creator Lee, Kim May
Mitra, Robin
Biedermann, Stefanie
description The presence of missing values complicates statistical analyses. In design of experiments, missing values are particularly problematic when constructing optimal designs, as it is not known which values are missing at the design stage. When data are missing at random it is possible to incorporate this information into the optimality criterion that is used to find designs; Imhof, Song and Wong (2002) develop such a framework. However, when data are not missing at random this framework can lead to inefficient designs. We investigate and address the specific challenges that not missing at random values present when finding optimal designs for linear regression models. We show that the optimality criteria depend on model parameters that traditionally do not affect the design, such as regression coefficients and the residual variance. We also develop a framework that improves efficiency of designs over those found when values are missing at random.
doi_str_mv 10.5705/ss.202016.0526
format article
fullrecord <record><control><sourceid>jstor_cross</sourceid><recordid>TN_cdi_crossref_primary_10_5705_ss_202016_0526</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><jstor_id>26511190</jstor_id><sourcerecordid>26511190</sourcerecordid><originalsourceid>FETCH-LOGICAL-c301t-eff2dd898ecc600b398dab2f62319a308241eee50ff5ce3a0f2b035ee94eaf703</originalsourceid><addsrcrecordid>eNo9j81Kw0AYRQdRsNRu3QnzAonfN5NJZsBNSMc0kB9pEl2G_MxARalkuvHtbUlxde_inguHkEcEX0Qgnp3zGTDA0AfBwhuyQqVCTwqIbs8dMPIgAHFPNs4dBgAFAiXwFXmp3pqsiHO61XWWlvRjp0tatU1SFZq-x3mraxrvNS2rhhZZXWdlSuOG7uNyWxUP5M72X85srrkm7atukp2XV2mWxLk3csCTZ6xl0ySVNOMYAgxcyakfmA0ZR9VzkCxAY4wAa8VoeA-WDcCFMSowvY2Ar4m__I7z0bnZ2O5nPnz382-H0F30O-e6Rb-76J-BpwX4dKfj_L9moUBEBfwP6cxRdg</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>OPTIMAL DESIGN WHEN OUTCOME VALUES ARE NOT MISSING AT RANDOM</title><source>JSTOR Archival Journals and Primary Sources Collection</source><creator>Lee, Kim May ; Mitra, Robin ; Biedermann, Stefanie</creator><creatorcontrib>Lee, Kim May ; Mitra, Robin ; Biedermann, Stefanie</creatorcontrib><description>The presence of missing values complicates statistical analyses. In design of experiments, missing values are particularly problematic when constructing optimal designs, as it is not known which values are missing at the design stage. When data are missing at random it is possible to incorporate this information into the optimality criterion that is used to find designs; Imhof, Song and Wong (2002) develop such a framework. However, when data are not missing at random this framework can lead to inefficient designs. We investigate and address the specific challenges that not missing at random values present when finding optimal designs for linear regression models. We show that the optimality criteria depend on model parameters that traditionally do not affect the design, such as regression coefficients and the residual variance. We also develop a framework that improves efficiency of designs over those found when values are missing at random.</description><identifier>ISSN: 1017-0405</identifier><identifier>EISSN: 1996-8507</identifier><identifier>DOI: 10.5705/ss.202016.0526</identifier><language>eng</language><publisher>Institute of Statistical Science, Academia Sinica and International Chinese Statistical Association</publisher><ispartof>Statistica Sinica, 2018-10, Vol.28 (4), p.1821-1838</ispartof><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c301t-eff2dd898ecc600b398dab2f62319a308241eee50ff5ce3a0f2b035ee94eaf703</citedby></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktopdf>$$Uhttps://www.jstor.org/stable/pdf/26511190$$EPDF$$P50$$Gjstor$$H</linktopdf><linktohtml>$$Uhttps://www.jstor.org/stable/26511190$$EHTML$$P50$$Gjstor$$H</linktohtml><link.rule.ids>314,776,780,27903,27904,58217,58450</link.rule.ids></links><search><creatorcontrib>Lee, Kim May</creatorcontrib><creatorcontrib>Mitra, Robin</creatorcontrib><creatorcontrib>Biedermann, Stefanie</creatorcontrib><title>OPTIMAL DESIGN WHEN OUTCOME VALUES ARE NOT MISSING AT RANDOM</title><title>Statistica Sinica</title><description>The presence of missing values complicates statistical analyses. In design of experiments, missing values are particularly problematic when constructing optimal designs, as it is not known which values are missing at the design stage. When data are missing at random it is possible to incorporate this information into the optimality criterion that is used to find designs; Imhof, Song and Wong (2002) develop such a framework. However, when data are not missing at random this framework can lead to inefficient designs. We investigate and address the specific challenges that not missing at random values present when finding optimal designs for linear regression models. We show that the optimality criteria depend on model parameters that traditionally do not affect the design, such as regression coefficients and the residual variance. We also develop a framework that improves efficiency of designs over those found when values are missing at random.</description><issn>1017-0405</issn><issn>1996-8507</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2018</creationdate><recordtype>article</recordtype><recordid>eNo9j81Kw0AYRQdRsNRu3QnzAonfN5NJZsBNSMc0kB9pEl2G_MxARalkuvHtbUlxde_inguHkEcEX0Qgnp3zGTDA0AfBwhuyQqVCTwqIbs8dMPIgAHFPNs4dBgAFAiXwFXmp3pqsiHO61XWWlvRjp0tatU1SFZq-x3mraxrvNS2rhhZZXWdlSuOG7uNyWxUP5M72X85srrkm7atukp2XV2mWxLk3csCTZ6xl0ySVNOMYAgxcyakfmA0ZR9VzkCxAY4wAa8VoeA-WDcCFMSowvY2Ar4m__I7z0bnZ2O5nPnz382-H0F30O-e6Rb-76J-BpwX4dKfj_L9moUBEBfwP6cxRdg</recordid><startdate>20181001</startdate><enddate>20181001</enddate><creator>Lee, Kim May</creator><creator>Mitra, Robin</creator><creator>Biedermann, Stefanie</creator><general>Institute of Statistical Science, Academia Sinica and International Chinese Statistical Association</general><scope>AAYXX</scope><scope>CITATION</scope></search><sort><creationdate>20181001</creationdate><title>OPTIMAL DESIGN WHEN OUTCOME VALUES ARE NOT MISSING AT RANDOM</title><author>Lee, Kim May ; Mitra, Robin ; Biedermann, Stefanie</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c301t-eff2dd898ecc600b398dab2f62319a308241eee50ff5ce3a0f2b035ee94eaf703</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2018</creationdate><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Lee, Kim May</creatorcontrib><creatorcontrib>Mitra, Robin</creatorcontrib><creatorcontrib>Biedermann, Stefanie</creatorcontrib><collection>CrossRef</collection><jtitle>Statistica Sinica</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Lee, Kim May</au><au>Mitra, Robin</au><au>Biedermann, Stefanie</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>OPTIMAL DESIGN WHEN OUTCOME VALUES ARE NOT MISSING AT RANDOM</atitle><jtitle>Statistica Sinica</jtitle><date>2018-10-01</date><risdate>2018</risdate><volume>28</volume><issue>4</issue><spage>1821</spage><epage>1838</epage><pages>1821-1838</pages><issn>1017-0405</issn><eissn>1996-8507</eissn><abstract>The presence of missing values complicates statistical analyses. In design of experiments, missing values are particularly problematic when constructing optimal designs, as it is not known which values are missing at the design stage. When data are missing at random it is possible to incorporate this information into the optimality criterion that is used to find designs; Imhof, Song and Wong (2002) develop such a framework. However, when data are not missing at random this framework can lead to inefficient designs. We investigate and address the specific challenges that not missing at random values present when finding optimal designs for linear regression models. We show that the optimality criteria depend on model parameters that traditionally do not affect the design, such as regression coefficients and the residual variance. We also develop a framework that improves efficiency of designs over those found when values are missing at random.</abstract><pub>Institute of Statistical Science, Academia Sinica and International Chinese Statistical Association</pub><doi>10.5705/ss.202016.0526</doi><tpages>18</tpages><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier ISSN: 1017-0405
ispartof Statistica Sinica, 2018-10, Vol.28 (4), p.1821-1838
issn 1017-0405
1996-8507
language eng
recordid cdi_crossref_primary_10_5705_ss_202016_0526
source JSTOR Archival Journals and Primary Sources Collection
title OPTIMAL DESIGN WHEN OUTCOME VALUES ARE NOT MISSING AT RANDOM
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-22T18%3A19%3A24IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-jstor_cross&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=OPTIMAL%20DESIGN%20WHEN%20OUTCOME%20VALUES%20ARE%20NOT%20MISSING%20AT%20RANDOM&rft.jtitle=Statistica%20Sinica&rft.au=Lee,%20Kim%20May&rft.date=2018-10-01&rft.volume=28&rft.issue=4&rft.spage=1821&rft.epage=1838&rft.pages=1821-1838&rft.issn=1017-0405&rft.eissn=1996-8507&rft_id=info:doi/10.5705/ss.202016.0526&rft_dat=%3Cjstor_cross%3E26511190%3C/jstor_cross%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-c301t-eff2dd898ecc600b398dab2f62319a308241eee50ff5ce3a0f2b035ee94eaf703%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_id=info:pmid/&rft_jstor_id=26511190&rfr_iscdi=true