Document Detail


The sensitivity of variance component estimates to underreporting: method and application to substance abuse data.
MedLine Citation:
PMID:  17903113     Owner:  NLM     Status:  MEDLINE    
Abstract/OtherAbstract:
Many studies of human behavior and psychological constructs rely on subjects' willingness to disclose information about themselves. This is problematic for phenotypes that require the disclosure of sensitive information, such as sexual behavior or illicit drug use, which are likely to be underreported. We describe a method for evaluating how sensitive variance component estimates are to underreporting. The method involves estimating, by maximum likelihood, the original population proportions of the response classes, and adjusting them for a set of hypothesized underreporting parameters. If the true values of the underreporting parameters were known, the researcher could estimate the variance components based on these values. Usually, underreporting levels are not known with certainty. However, it is possible to assume a specific value for the underreporting rate, obtain response pattern proportions adjusted for this rate, and then to conduct the analyses on these revised estimates. By repeating the procedure across the range of plausible underreporting values, the researcher can assess how sensitive the variance component estimates are to variation in underreporting. We apply this method to a sample of male-male twin pairs who reported on themselves and their co-twins for illicit drug abuse and dependence (DAD). We show how underreporting influences estimates of additive genetic, common environment, and specific environment variance components (A, C, and E) obtained for DAD in a classical twin design.
Authors:
Paul W Andrews; Kenneth S Kendler; Nathan Gillespie; Michael C Neale
Related Documents :
15912873 - The variance of d' estimates obtained in yes-no and two-interval forced choice procedures.
16795483 - Forcing square pegs into round holes: some comments on "an analysis-of-variance model f...
1868593 - Predictive, error-compensating kinetic method for enzymatic quantification of creatinin...
Publication Detail:
Type:  Journal Article; Research Support, N.I.H., Extramural; Twin Study    
Journal Detail:
Title:  Twin research and human genetics : the official journal of the International Society for Twin Studies     Volume:  10     ISSN:  1832-4274     ISO Abbreviation:  -     Publication Date:  2007 Oct 
Date Detail:
Created Date:  2007-10-01     Completed Date:  2008-03-14     Revised Date:  -    
Medline Journal Info:
Nlm Unique ID:  101244624     Medline TA:  Twin Res Hum Genet     Country:  Australia    
Other Details:
Languages:  eng     Pagination:  721-8     Citation Subset:  IM    
Affiliation:
Virginia Institute for Psychiatric and Behavioral Genetics,Virginia Commonwealth University, Richmond,Virginia 23298-0126, USA. pandrews@vcu.edu
Export Citation:
APA/MLA Format     Download EndNote     Download BibTex
MeSH Terms
Descriptor/Qualifier:
Analysis of Variance
Diseases in Twins / genetics*,  psychology*
Female
Humans
Interviews as Topic
Likelihood Functions
Male
Models, Psychological
Prevalence
Registries
Substance-Related Disorders / genetics*,  psychology*
Grant Support
ID/Acronym/Agency:
P32 MH-20030/MH/NIMH NIH HHS

From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine


Previous Document:  Differences in smoking habits of MZ and DZ twins: a commentary on Tishler and Carey.
Next Document:  Twin analyses of fatigue.