Repeated measures regression mixture models

M. Kim*, M.L. Van Horn*, T. Jaki, J. Vermunt, D. Feaster, K.L. Lichstein, D.J. Taylor, B.W. Riedel, A.J. Bush

*Corresponding author for this work

Research output: Contribution to journalArticleScientificpeer-review

Abstract

Regression mixture models are one increasingly utilized approach for developing theories about and exploring the heterogeneity of effects. In this study we aimed to extend the current use of regression mixtures to a repeated regression mixture method when repeated measures, such as diary-type and experience-sampling method, data are available. We hypothesized that additional information borrowed from the repeated measures would improve the model performance, in terms of class enumeration and accuracy of the parameter estimates. We specifically compared three types of model specifications in regression mixtures: (a) traditional single-outcome model; (b) repeated measures models with three, five, and seven measures; and (c) a single-outcome model with the average of seven repeated measures. The results showed that the repeated measures regression mixture models substantially outperformed the traditional and average single-outcome models in class enumeration, with less bias in the parameter estimates. For sample size, whereas prior recommendations have suggested that regression mixtures require samples of well over 1,000 participants, even for classes at a large distance from each other (classes with regression weights of .20 vs. .70), the present repeated measures regression mixture models allow for samples as low as 200 participants with an increased number (i.e., seven) of repeated measures. We also demonstrate an application of the proposed repeated measures approach using data from the Sleep Research Project. Implications and limitations of the study are discussed.
Original languageEnglish
JournalBehavior Research Methods
DOIs
Publication statusE-pub ahead of print - 2020

Fingerprint

Weights and Measures
Mixture Model
Ecological Momentary Assessment
Enumeration
Sample Size
Research Projects
Diary
Sleep
Sampling

Cite this

Kim, M., Van Horn, M. L., Jaki, T., Vermunt, J., Feaster, D., Lichstein, K. L., ... Bush, A. J. (2020). Repeated measures regression mixture models. Behavior Research Methods. https://doi.org/10.3758/s13428-019-01257-7
Kim, M. ; Van Horn, M.L. ; Jaki, T. ; Vermunt, J. ; Feaster, D. ; Lichstein, K.L. ; Taylor, D.J. ; Riedel, B.W. ; Bush, A.J. / Repeated measures regression mixture models. In: Behavior Research Methods. 2020.
@article{d007b1e16d49444988c208dad0354fbd,
title = "Repeated measures regression mixture models",
abstract = "Regression mixture models are one increasingly utilized approach for developing theories about and exploring the heterogeneity of effects. In this study we aimed to extend the current use of regression mixtures to a repeated regression mixture method when repeated measures, such as diary-type and experience-sampling method, data are available. We hypothesized that additional information borrowed from the repeated measures would improve the model performance, in terms of class enumeration and accuracy of the parameter estimates. We specifically compared three types of model specifications in regression mixtures: (a) traditional single-outcome model; (b) repeated measures models with three, five, and seven measures; and (c) a single-outcome model with the average of seven repeated measures. The results showed that the repeated measures regression mixture models substantially outperformed the traditional and average single-outcome models in class enumeration, with less bias in the parameter estimates. For sample size, whereas prior recommendations have suggested that regression mixtures require samples of well over 1,000 participants, even for classes at a large distance from each other (classes with regression weights of .20 vs. .70), the present repeated measures regression mixture models allow for samples as low as 200 participants with an increased number (i.e., seven) of repeated measures. We also demonstrate an application of the proposed repeated measures approach using data from the Sleep Research Project. Implications and limitations of the study are discussed.",
author = "M. Kim and {Van Horn}, M.L. and T. Jaki and J. Vermunt and D. Feaster and K.L. Lichstein and D.J. Taylor and B.W. Riedel and A.J. Bush",
year = "2020",
doi = "10.3758/s13428-019-01257-7",
language = "English",
journal = "Behavior Research Methods",
issn = "1554-351X",
publisher = "Springer",

}

Kim, M, Van Horn, ML, Jaki, T, Vermunt, J, Feaster, D, Lichstein, KL, Taylor, DJ, Riedel, BW & Bush, AJ 2020, 'Repeated measures regression mixture models', Behavior Research Methods. https://doi.org/10.3758/s13428-019-01257-7

Repeated measures regression mixture models. / Kim, M.; Van Horn, M.L.; Jaki, T.; Vermunt, J.; Feaster, D.; Lichstein, K.L.; Taylor, D.J.; Riedel, B.W.; Bush, A.J.

In: Behavior Research Methods, 2020.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Repeated measures regression mixture models

AU - Kim, M.

AU - Van Horn, M.L.

AU - Jaki, T.

AU - Vermunt, J.

AU - Feaster, D.

AU - Lichstein, K.L.

AU - Taylor, D.J.

AU - Riedel, B.W.

AU - Bush, A.J.

PY - 2020

Y1 - 2020

N2 - Regression mixture models are one increasingly utilized approach for developing theories about and exploring the heterogeneity of effects. In this study we aimed to extend the current use of regression mixtures to a repeated regression mixture method when repeated measures, such as diary-type and experience-sampling method, data are available. We hypothesized that additional information borrowed from the repeated measures would improve the model performance, in terms of class enumeration and accuracy of the parameter estimates. We specifically compared three types of model specifications in regression mixtures: (a) traditional single-outcome model; (b) repeated measures models with three, five, and seven measures; and (c) a single-outcome model with the average of seven repeated measures. The results showed that the repeated measures regression mixture models substantially outperformed the traditional and average single-outcome models in class enumeration, with less bias in the parameter estimates. For sample size, whereas prior recommendations have suggested that regression mixtures require samples of well over 1,000 participants, even for classes at a large distance from each other (classes with regression weights of .20 vs. .70), the present repeated measures regression mixture models allow for samples as low as 200 participants with an increased number (i.e., seven) of repeated measures. We also demonstrate an application of the proposed repeated measures approach using data from the Sleep Research Project. Implications and limitations of the study are discussed.

AB - Regression mixture models are one increasingly utilized approach for developing theories about and exploring the heterogeneity of effects. In this study we aimed to extend the current use of regression mixtures to a repeated regression mixture method when repeated measures, such as diary-type and experience-sampling method, data are available. We hypothesized that additional information borrowed from the repeated measures would improve the model performance, in terms of class enumeration and accuracy of the parameter estimates. We specifically compared three types of model specifications in regression mixtures: (a) traditional single-outcome model; (b) repeated measures models with three, five, and seven measures; and (c) a single-outcome model with the average of seven repeated measures. The results showed that the repeated measures regression mixture models substantially outperformed the traditional and average single-outcome models in class enumeration, with less bias in the parameter estimates. For sample size, whereas prior recommendations have suggested that regression mixtures require samples of well over 1,000 participants, even for classes at a large distance from each other (classes with regression weights of .20 vs. .70), the present repeated measures regression mixture models allow for samples as low as 200 participants with an increased number (i.e., seven) of repeated measures. We also demonstrate an application of the proposed repeated measures approach using data from the Sleep Research Project. Implications and limitations of the study are discussed.

U2 - 10.3758/s13428-019-01257-7

DO - 10.3758/s13428-019-01257-7

M3 - Article

JO - Behavior Research Methods

JF - Behavior Research Methods

SN - 1554-351X

ER -