Latent Class Multiple Imputation for multiply observed variables in a combined dataset

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientific

Abstract

Both registers and sample surveys can contain measurement error. While
some errors are invisibly present, others become visible when logical
relations in the data are investigated. When a variable is measured in multiple
datasets within a combined dataset, we can get an indication of the errors
which are invisibly present within the separate datasets. We propose a new
method (MILC) based on latent class modelling that estimates the number of
measurement errors in the multiple sources, and simultaneously takes
impossible combinations with other variables into account. We then use the
latent class model to multiply impute the latent “true” variable. Whether
MILC can be applied depends on the entropy R2 of the LC model and the
type of analysis you are interested in.
Original languageEnglish
Title of host publicationQ 2016
Subtitle of host publicationEuropean Conference on Quality in Official Statistics
Place of PublicationMadrid
Pages1-10
Number of pages10
Publication statusPublished - 2 Jun 2016

Fingerprint

Measurement errors
Entropy

Keywords

  • latent class models
  • multiple imputation
  • combined dataset

Cite this

Boeschoten, L., Oberski, D. L., & de Waal, A. G. (2016). Latent Class Multiple Imputation for multiply observed variables in a combined dataset. In Q 2016: European Conference on Quality in Official Statistics (pp. 1-10). Madrid.
Boeschoten, L. ; Oberski, D.L. ; de Waal, A.G. / Latent Class Multiple Imputation for multiply observed variables in a combined dataset. Q 2016: European Conference on Quality in Official Statistics . Madrid, 2016. pp. 1-10
@inproceedings{d3736fe878414756993c78ca99768529,
title = "Latent Class Multiple Imputation for multiply observed variables in a combined dataset",
abstract = "Both registers and sample surveys can contain measurement error. Whilesome errors are invisibly present, others become visible when logicalrelations in the data are investigated. When a variable is measured in multipledatasets within a combined dataset, we can get an indication of the errorswhich are invisibly present within the separate datasets. We propose a newmethod (MILC) based on latent class modelling that estimates the number ofmeasurement errors in the multiple sources, and simultaneously takesimpossible combinations with other variables into account. We then use thelatent class model to multiply impute the latent “true” variable. WhetherMILC can be applied depends on the entropy R2 of the LC model and thetype of analysis you are interested in.",
keywords = "latent class models, multiple imputation, combined dataset",
author = "L. Boeschoten and D.L. Oberski and {de Waal}, A.G.",
year = "2016",
month = "6",
day = "2",
language = "English",
pages = "1--10",
booktitle = "Q 2016",

}

Boeschoten, L, Oberski, DL & de Waal, AG 2016, Latent Class Multiple Imputation for multiply observed variables in a combined dataset. in Q 2016: European Conference on Quality in Official Statistics . Madrid, pp. 1-10.

Latent Class Multiple Imputation for multiply observed variables in a combined dataset. / Boeschoten, L.; Oberski, D.L.; de Waal, A.G.

Q 2016: European Conference on Quality in Official Statistics . Madrid, 2016. p. 1-10.

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientific

TY - GEN

T1 - Latent Class Multiple Imputation for multiply observed variables in a combined dataset

AU - Boeschoten, L.

AU - Oberski, D.L.

AU - de Waal, A.G.

PY - 2016/6/2

Y1 - 2016/6/2

N2 - Both registers and sample surveys can contain measurement error. Whilesome errors are invisibly present, others become visible when logicalrelations in the data are investigated. When a variable is measured in multipledatasets within a combined dataset, we can get an indication of the errorswhich are invisibly present within the separate datasets. We propose a newmethod (MILC) based on latent class modelling that estimates the number ofmeasurement errors in the multiple sources, and simultaneously takesimpossible combinations with other variables into account. We then use thelatent class model to multiply impute the latent “true” variable. WhetherMILC can be applied depends on the entropy R2 of the LC model and thetype of analysis you are interested in.

AB - Both registers and sample surveys can contain measurement error. Whilesome errors are invisibly present, others become visible when logicalrelations in the data are investigated. When a variable is measured in multipledatasets within a combined dataset, we can get an indication of the errorswhich are invisibly present within the separate datasets. We propose a newmethod (MILC) based on latent class modelling that estimates the number ofmeasurement errors in the multiple sources, and simultaneously takesimpossible combinations with other variables into account. We then use thelatent class model to multiply impute the latent “true” variable. WhetherMILC can be applied depends on the entropy R2 of the LC model and thetype of analysis you are interested in.

KW - latent class models

KW - multiple imputation

KW - combined dataset

M3 - Conference contribution

SP - 1

EP - 10

BT - Q 2016

CY - Madrid

ER -

Boeschoten L, Oberski DL, de Waal AG. Latent Class Multiple Imputation for multiply observed variables in a combined dataset. In Q 2016: European Conference on Quality in Official Statistics . Madrid. 2016. p. 1-10