Home > Research > Publications & Outputs > Using mixtures of t densities to make inference...

Research

Associated organisational unit

Links

Text available via DOI:

Keywords

View graph of relations

Using mixtures of t densities to make inferences in the presence of missing data with a small number of multiply imputed data sets

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Published

Standard

Using mixtures of t densities to make inferences in the presence of missing data with a small number of multiply imputed data sets. / Rashid, S.; Mitra, Robin; Steele, R.J.
In: Computational Statistics and Data Analysis, Vol. 92, 01.12.2015, p. 84-96.

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Harvard

Rashid, S, Mitra, R & Steele, RJ 2015, 'Using mixtures of t densities to make inferences in the presence of missing data with a small number of multiply imputed data sets', Computational Statistics and Data Analysis, vol. 92, pp. 84-96. https://doi.org/10.1016/j.csda.2015.05.009

APA

Rashid, S., Mitra, R., & Steele, R. J. (2015). Using mixtures of t densities to make inferences in the presence of missing data with a small number of multiply imputed data sets. Computational Statistics and Data Analysis, 92, 84-96. https://doi.org/10.1016/j.csda.2015.05.009

Vancouver

Rashid S, Mitra R, Steele RJ. Using mixtures of t densities to make inferences in the presence of missing data with a small number of multiply imputed data sets. Computational Statistics and Data Analysis. 2015 Dec 1;92:84-96. Epub 2015 Jun 19. doi: 10.1016/j.csda.2015.05.009

Author

Rashid, S. ; Mitra, Robin ; Steele, R.J. / Using mixtures of t densities to make inferences in the presence of missing data with a small number of multiply imputed data sets. In: Computational Statistics and Data Analysis. 2015 ; Vol. 92. pp. 84-96.

Bibtex

@article{2f9ad2729fb34c379e0fcb5ee1b7a9ae,
title = "Using mixtures of t densities to make inferences in the presence of missing data with a small number of multiply imputed data sets",
abstract = "Strategies for making inference in the presence of missing data after conducting a Multiple Imputation (MI) procedure are considered. An approach which approximates the posterior distribution for parameters using a mixture of tt-distributions is proposed. Simulated experiments show this approach improves inferences in some aspects, making them more stable over repeated analysis and creating narrower bounds for certain common statistics of interest. Extensions to the existing literature have been executed that provide further stability to inferences and also a strong potential to identify ways to make the analysis procedure more flexible. The competing methods have been first compared using simulated data sets and then a real data set concerning analysis of the effect of breastfeeding duration on children?s cognitive ability. R code to implement the methods used is available as online supplementary material.",
keywords = "MISSING DATA, multiple imputation, bayesian statistics, disclosure avoidance, mixture distribution, monte carlo",
author = "S. Rashid and Robin Mitra and R.J. Steele",
year = "2015",
month = dec,
day = "1",
doi = "10.1016/j.csda.2015.05.009",
language = "English",
volume = "92",
pages = "84--96",
journal = "Computational Statistics and Data Analysis",
issn = "0167-9473",
publisher = "Elsevier",

}

RIS

TY - JOUR

T1 - Using mixtures of t densities to make inferences in the presence of missing data with a small number of multiply imputed data sets

AU - Rashid, S.

AU - Mitra, Robin

AU - Steele, R.J.

PY - 2015/12/1

Y1 - 2015/12/1

N2 - Strategies for making inference in the presence of missing data after conducting a Multiple Imputation (MI) procedure are considered. An approach which approximates the posterior distribution for parameters using a mixture of tt-distributions is proposed. Simulated experiments show this approach improves inferences in some aspects, making them more stable over repeated analysis and creating narrower bounds for certain common statistics of interest. Extensions to the existing literature have been executed that provide further stability to inferences and also a strong potential to identify ways to make the analysis procedure more flexible. The competing methods have been first compared using simulated data sets and then a real data set concerning analysis of the effect of breastfeeding duration on children?s cognitive ability. R code to implement the methods used is available as online supplementary material.

AB - Strategies for making inference in the presence of missing data after conducting a Multiple Imputation (MI) procedure are considered. An approach which approximates the posterior distribution for parameters using a mixture of tt-distributions is proposed. Simulated experiments show this approach improves inferences in some aspects, making them more stable over repeated analysis and creating narrower bounds for certain common statistics of interest. Extensions to the existing literature have been executed that provide further stability to inferences and also a strong potential to identify ways to make the analysis procedure more flexible. The competing methods have been first compared using simulated data sets and then a real data set concerning analysis of the effect of breastfeeding duration on children?s cognitive ability. R code to implement the methods used is available as online supplementary material.

KW - MISSING DATA

KW - multiple imputation

KW - bayesian statistics

KW - disclosure avoidance

KW - mixture distribution

KW - monte carlo

U2 - 10.1016/j.csda.2015.05.009

DO - 10.1016/j.csda.2015.05.009

M3 - Journal article

VL - 92

SP - 84

EP - 96

JO - Computational Statistics and Data Analysis

JF - Computational Statistics and Data Analysis

SN - 0167-9473

ER -