EconStor >
Humboldt-Universität Berlin >
Sonderforschungsbereich 373: Quantification and Simulation of Economic Processes, Humboldt-Universität Berlin >
Discussion Papers, SFB 373, HU Berlin >

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10419/66235
  

Full metadata record

DC FieldValueLanguage
dc.contributor.authorCarroll, Raymond J.en_US
dc.contributor.authorFreedman, Laurenceen_US
dc.contributor.authorPee, Daviden_US
dc.date.accessioned2012-10-23en_US
dc.date.accessioned2012-11-19T15:22:02Z-
dc.date.available2012-11-19T15:22:02Z-
dc.date.issued1997en_US
dc.identifier.piurn:nbn:de:kobv:11-10063721en_US
dc.identifier.urihttp://hdl.handle.net/10419/66235-
dc.description.abstractMotivated by an example in nutritional epidemiology, we investigate some design and analysis aspects of linear measurement error models with missing surrogate data. The specific problem investigated consists of an initial large sample in which the response (a food frequency questionnaire, FFQ) is observed, and then a smaller calibration study in which replicates of the error prone predictor are observed (food records or recalls, FR). The difference between our analysis and most of the measurement error model literature is that in our study, the selection into the calibration study can depend upon the value of the response. Rationale for this type of design is given. Two major problems are investigated. In the design of a calibration study, one has the option of larger sample sizes and fewer replicates, or smaller sample sizes and more replicates. Somewhat surprisingly, neither strategy is uniformly preferable in cases of practical interest. The answers depend on the instrument used (recalls or records) and the parameters of interest. The second problem investigated is one of analysis. In the usual linear model with no missing data, method of moments estimates and normal-theory maximum likelihood estimates are approximately equivalent, with the former method in most use because it can be calculated easily and explicitly. Both estimates are valid without any distributional assumptions. In contrast, in the missing data problem under consideration, only the moments estimate is distribution-free, but the maximum likelihood estimate has at least 50% greater precision in practical situations when normality obtains. Implications for the design of nutritional calibration studies are discussed.en_US
dc.language.isoengen_US
dc.publisherHumboldt-Universität Berlinen_US
dc.relation.ispartofseriesDiscussion Papers, Interdisciplinary Research Project 373: Quantification and Simulation of Economic Processes 1997,12en_US
dc.subject.ddc330en_US
dc.subject.keywordMeasurement Erroren_US
dc.subject.keywordErrors-in-Variablesen_US
dc.subject.keywordEstimating Equationsen_US
dc.subject.keywordNutritionen_US
dc.subject.keywordSampling Designsen_US
dc.subject.keywordLinear regressionen_US
dc.subject.keywordMaximum Likelihooden_US
dc.subject.keywordMethod of Momentsen_US
dc.subject.keywordMissing Dataen_US
dc.subject.keywordModel Robustnessen_US
dc.subject.keywordSemiparametricsen_US
dc.subject.keywordStratified Samplingen_US
dc.subject.keywordWeightingen_US
dc.titleDesign aspects of calibration studies in nutrition, with analysis of missing data in linear measurement error modelsen_US
dc.typeWorking Paperen_US
dc.identifier.ppn728322153en_US
dc.rightshttp://www.econstor.eu/dspace/Nutzungsbedingungenen_US
dc.identifier.repecRePEc:zbw:sfb373:199712-
Appears in Collections:Discussion Papers, SFB 373, HU Berlin

Files in This Item:
File Description SizeFormat
728322153.pdf253.24 kBAdobe PDF
No. of Downloads: Counter Stats
Show simple item record
Download bibliographical data as: BibTeX

Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.