A new approach to handle missing covariate data in twin research: With an application to educational achievement data

I. Schwabe, Dorret I. Boomsma, Eveline L. De Zeeuw, Stéphanie M. Van Den Berg

Research output: Contribution to journalArticleScientificpeer-review

11 Downloads (Pure)

Abstract

The often-used ACE model which decomposes phenotypic variance into additive genetic (A), common environmental (C) and unique-environmental (E) parts can be extended to include covariates. Collection of these variables however often leads to a large amount of missing data, for example when self-reports (e.g. questionnaires) are not fully completed. The usual approach to handle missing covariate data in twin research results in reduced power to detect statistical effects, as only phenotypic and covariate data of individual twins with complete data can be used. Here we present a full information approach to handle missing covariate data that makes it possible to use all available data. A simulation study shows that, independent of missingness scenario, number of covariates or amount of missingness, the full information approach is more powerful than the usual approach. To illustrate the new method, we applied it to test scores on a Dutch national school achievement test (Eindtoets Basisonderwijs) in the final grade of primary school of 990 twin pairs. The effects of school-aggregated measures (e.g. school denomination, pedagogical philosophy, school size) and the effect of the sex of a twin on these test scores were tested. None of the covariates had a significant effect on individual differences in test scores.
Original languageEnglish
Pages (from-to)583-595
JournalBehavior Genetics
Volume46
Issue number4
DOIs
Publication statusPublished - 1 Jul 2016

Fingerprint

Educational Status
Individuality
Self Report
school
test
effect
simulation

Cite this

Schwabe, I. ; Boomsma, Dorret I. ; Zeeuw, Eveline L. De ; Berg, Stéphanie M. Van Den. / A new approach to handle missing covariate data in twin research : With an application to educational achievement data. In: Behavior Genetics. 2016 ; Vol. 46, No. 4. pp. 583-595.
@article{2f21576f9b43441e8dc5773b960c4ca3,
title = "A new approach to handle missing covariate data in twin research: With an application to educational achievement data",
abstract = "The often-used ACE model which decomposes phenotypic variance into additive genetic (A), common environmental (C) and unique-environmental (E) parts can be extended to include covariates. Collection of these variables however often leads to a large amount of missing data, for example when self-reports (e.g. questionnaires) are not fully completed. The usual approach to handle missing covariate data in twin research results in reduced power to detect statistical effects, as only phenotypic and covariate data of individual twins with complete data can be used. Here we present a full information approach to handle missing covariate data that makes it possible to use all available data. A simulation study shows that, independent of missingness scenario, number of covariates or amount of missingness, the full information approach is more powerful than the usual approach. To illustrate the new method, we applied it to test scores on a Dutch national school achievement test (Eindtoets Basisonderwijs) in the final grade of primary school of 990 twin pairs. The effects of school-aggregated measures (e.g. school denomination, pedagogical philosophy, school size) and the effect of the sex of a twin on these test scores were tested. None of the covariates had a significant effect on individual differences in test scores.",
author = "I. Schwabe and Boomsma, {Dorret I.} and Zeeuw, {Eveline L. De} and Berg, {St{\'e}phanie M. Van Den}",
year = "2016",
month = "7",
day = "1",
doi = "10.1007/s10519-015-9771-1",
language = "English",
volume = "46",
pages = "583--595",
journal = "Behavior Genetics",
issn = "0001-8244",
publisher = "Springer New York",
number = "4",

}

A new approach to handle missing covariate data in twin research : With an application to educational achievement data. / Schwabe, I.; Boomsma, Dorret I.; Zeeuw, Eveline L. De; Berg, Stéphanie M. Van Den.

In: Behavior Genetics, Vol. 46, No. 4, 01.07.2016, p. 583-595.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - A new approach to handle missing covariate data in twin research

T2 - With an application to educational achievement data

AU - Schwabe, I.

AU - Boomsma, Dorret I.

AU - Zeeuw, Eveline L. De

AU - Berg, Stéphanie M. Van Den

PY - 2016/7/1

Y1 - 2016/7/1

N2 - The often-used ACE model which decomposes phenotypic variance into additive genetic (A), common environmental (C) and unique-environmental (E) parts can be extended to include covariates. Collection of these variables however often leads to a large amount of missing data, for example when self-reports (e.g. questionnaires) are not fully completed. The usual approach to handle missing covariate data in twin research results in reduced power to detect statistical effects, as only phenotypic and covariate data of individual twins with complete data can be used. Here we present a full information approach to handle missing covariate data that makes it possible to use all available data. A simulation study shows that, independent of missingness scenario, number of covariates or amount of missingness, the full information approach is more powerful than the usual approach. To illustrate the new method, we applied it to test scores on a Dutch national school achievement test (Eindtoets Basisonderwijs) in the final grade of primary school of 990 twin pairs. The effects of school-aggregated measures (e.g. school denomination, pedagogical philosophy, school size) and the effect of the sex of a twin on these test scores were tested. None of the covariates had a significant effect on individual differences in test scores.

AB - The often-used ACE model which decomposes phenotypic variance into additive genetic (A), common environmental (C) and unique-environmental (E) parts can be extended to include covariates. Collection of these variables however often leads to a large amount of missing data, for example when self-reports (e.g. questionnaires) are not fully completed. The usual approach to handle missing covariate data in twin research results in reduced power to detect statistical effects, as only phenotypic and covariate data of individual twins with complete data can be used. Here we present a full information approach to handle missing covariate data that makes it possible to use all available data. A simulation study shows that, independent of missingness scenario, number of covariates or amount of missingness, the full information approach is more powerful than the usual approach. To illustrate the new method, we applied it to test scores on a Dutch national school achievement test (Eindtoets Basisonderwijs) in the final grade of primary school of 990 twin pairs. The effects of school-aggregated measures (e.g. school denomination, pedagogical philosophy, school size) and the effect of the sex of a twin on these test scores were tested. None of the covariates had a significant effect on individual differences in test scores.

U2 - 10.1007/s10519-015-9771-1

DO - 10.1007/s10519-015-9771-1

M3 - Article

VL - 46

SP - 583

EP - 595

JO - Behavior Genetics

JF - Behavior Genetics

SN - 0001-8244

IS - 4

ER -