Home > Research > Publications & Outputs > Using regression mixture models with non-normal...
View graph of relations

Using regression mixture models with non-normal data: examining an ordered polytomous approach

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Published

Standard

Using regression mixture models with non-normal data: examining an ordered polytomous approach. / George, Melissa; Yang, Na; Van Horn, M. Lee et al.
In: Journal of Statistical Computation and Simulation, Vol. 83, No. 4, 2013, p. 757-770.

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Harvard

George, M, Yang, N, Van Horn, ML, Smith, J, Jaki, T, Feaster, D, Maysn, K & Howe, G 2013, 'Using regression mixture models with non-normal data: examining an ordered polytomous approach', Journal of Statistical Computation and Simulation, vol. 83, no. 4, pp. 757-770. https://doi.org/10.1080/00949655.2011.636363

APA

George, M., Yang, N., Van Horn, M. L., Smith, J., Jaki, T., Feaster, D., Maysn, K., & Howe, G. (2013). Using regression mixture models with non-normal data: examining an ordered polytomous approach. Journal of Statistical Computation and Simulation, 83(4), 757-770. https://doi.org/10.1080/00949655.2011.636363

Vancouver

George M, Yang N, Van Horn ML, Smith J, Jaki T, Feaster D et al. Using regression mixture models with non-normal data: examining an ordered polytomous approach. Journal of Statistical Computation and Simulation. 2013;83(4):757-770. Epub 2011 Dec 13. doi: 10.1080/00949655.2011.636363

Author

George, Melissa ; Yang, Na ; Van Horn, M. Lee et al. / Using regression mixture models with non-normal data: examining an ordered polytomous approach. In: Journal of Statistical Computation and Simulation. 2013 ; Vol. 83, No. 4. pp. 757-770.

Bibtex

@article{d3c8ee7d898f4f32838136662e33f080,
title = "Using regression mixture models with non-normal data: examining an ordered polytomous approach",
abstract = "Mild to moderate skew in errors can substantially impact regression mixture model results; one approach for overcoming this includes transforming the outcome into an ordered categorical variable and using a polytomous regression mixture model. This is effective for retaining differential effects in the population; however, bias in parameter estimates and model fit warrant further examination of this approach at higher levels of skew. The current study used Monte Carlo simulations; 3000 observations were drawn from each of two subpopulations differing in the effect of X on Y. Five hundred simulations were performed in each of the 10 scenarios varying in levels of skew in one or both classes. Model comparison criteria supported the accurate two-class model, preserving the differential effects, while parameter estimates were notably biased. The appropriate number of effects can be captured with this approach but we suggest caution when interpreting the magnitude of the effects.",
keywords = "regression mixture models, non-normal errors , differential effects",
author = "Melissa George and Na Yang and {Van Horn}, {M. Lee} and Jessalyn Smith and Thomas Jaki and Daniel Feaster and Katherine Maysn and George Howe",
year = "2013",
doi = "10.1080/00949655.2011.636363",
language = "English",
volume = "83",
pages = "757--770",
journal = "Journal of Statistical Computation and Simulation",
issn = "1563-5163",
publisher = "Taylor and Francis Ltd.",
number = "4",

}

RIS

TY - JOUR

T1 - Using regression mixture models with non-normal data: examining an ordered polytomous approach

AU - George, Melissa

AU - Yang, Na

AU - Van Horn, M. Lee

AU - Smith, Jessalyn

AU - Jaki, Thomas

AU - Feaster, Daniel

AU - Maysn, Katherine

AU - Howe, George

PY - 2013

Y1 - 2013

N2 - Mild to moderate skew in errors can substantially impact regression mixture model results; one approach for overcoming this includes transforming the outcome into an ordered categorical variable and using a polytomous regression mixture model. This is effective for retaining differential effects in the population; however, bias in parameter estimates and model fit warrant further examination of this approach at higher levels of skew. The current study used Monte Carlo simulations; 3000 observations were drawn from each of two subpopulations differing in the effect of X on Y. Five hundred simulations were performed in each of the 10 scenarios varying in levels of skew in one or both classes. Model comparison criteria supported the accurate two-class model, preserving the differential effects, while parameter estimates were notably biased. The appropriate number of effects can be captured with this approach but we suggest caution when interpreting the magnitude of the effects.

AB - Mild to moderate skew in errors can substantially impact regression mixture model results; one approach for overcoming this includes transforming the outcome into an ordered categorical variable and using a polytomous regression mixture model. This is effective for retaining differential effects in the population; however, bias in parameter estimates and model fit warrant further examination of this approach at higher levels of skew. The current study used Monte Carlo simulations; 3000 observations were drawn from each of two subpopulations differing in the effect of X on Y. Five hundred simulations were performed in each of the 10 scenarios varying in levels of skew in one or both classes. Model comparison criteria supported the accurate two-class model, preserving the differential effects, while parameter estimates were notably biased. The appropriate number of effects can be captured with this approach but we suggest caution when interpreting the magnitude of the effects.

KW - regression mixture models

KW - non-normal errors

KW - differential effects

U2 - 10.1080/00949655.2011.636363

DO - 10.1080/00949655.2011.636363

M3 - Journal article

VL - 83

SP - 757

EP - 770

JO - Journal of Statistical Computation and Simulation

JF - Journal of Statistical Computation and Simulation

SN - 1563-5163

IS - 4

ER -