Addressing Heteroscedasticity in Correlated Binary Data: A Bayesian Mixed Effects Location Scale Approach
,
Abstract
Introduction: The mixed effects logistic regression model is a common model for analysing correlated binary data as longitudinal data. The between and within subject variances are typically considered to be homogeneous but longitudinal data often show heterogeneity in these variances. This study proposes a Bayesian mixed effects location scale model to accommodate heteroscedasticity in binary data analysis.
Methods: This study was carried out in two stages; first, the simulation study was used to evaluate the
accuracy of the proposed model with the Bayesian approach and then the proposed model was applied to a real data. In simulation study, the data were generated from the mixed effects location scale model with different correlations between the random location effect and random scale effect and different sample sizes. In order to evaluate the accuracy of the estimations, the Root Mean Square Error, bias and Coverage Probability were calculated and the deviance information criterion was used to select the appropriate model. At the end we utilized this model to analyse uric acid levels of patients with haematological disorders.
Results: The simulation results show the accuracy of model parameter estimates as well as the correlation between random location and scale effects. They also display that if a random scale effect is present in the data, it should be accounted for in model. Otherwise, the model is forced to assign the within subject variation due to these subject random effects to the error term. The results of real data are also in line with this. The odds of having normal UA levels increases by a factor of 26% per week. Due to the positive value of the covariance parameter, patients with higher mean of UA levels show higher variation in UA levels. Furthermore, the significance of the covariates in the between subject and within subject variances model, as well as the significance of the random scale variance determines the heterogeneity across subjects.
Conclusion: Bayesian mixed effects location scale model provides a useful tool for analysing correlated binary data with heteroscedasticity because it considers data correlation and modelling mean and variance simultaneously. Furthermore, it improves the accuracy of statistical inference in longitudinal studies compared to classic mixed effects models.
2. Boateng, EY, Abaye DA. A Review of the Logistic Regression Model with Emphasis on Medical Research. 2019. Journal of Data Analysis and Information Processing.
3. Molenberghs G, Lesaffre E. Marginal modeling of correlated ordinal data using a multivariate plackett distribution. J Am Stat Assoc. 1994;89(426):633–44.
4. Diggle P, Diggle DMSPJ, allgemeine tierzucht F, Press OU, Diggle PJ, Heagerty P, et al. Analysis of Longitudinal Data (2rd ed.). OUP Oxford; 2002. Available from: https://books. google.com/books?id=kKLbyWycRwcC.
5. Verbeke G, Lesaffre E. Linear the Model With Heterogeneity Population in. J Am Stat Assoc. 2012;91(433):217–21.
6. Reuter M, Hennig J, Netter P, Buehner M, Hueppe M. Using Latent Mixed Markov Models for the choice of the best pharmacological treatment. Stat Med. 2004;23(9):1337–49.
7. Chung H, Park YS, Lanza ST. Latent transition analysis with covariates: Pubertal timing and substance use behaviours in adolescent females. Stat Med. 2005;24(18):2895–910.
8. Tang D, Zhou Y, Long C, Tang S. The Association of Midday Napping With Hypertension Among Chinese Adults Older Than 45 Years: Cross-sectional Study. JMIR Public Health Surveill. 2022 Nov 22;8(11):e38782.
9. Iddrisu, AK., Besing Karadaar, I., Gurah Junior, J. et al. Mixed effects logistic regression analysis of blood pressure among Ghanaians and associated risk factors. Scientific Report. 2023; 13(1).
10. Molenberghs G, Verbeke G. Models for Discrete Longitudinal Data [Internet]. Springer New York; 2006. (Springer Series in Statistics). Available from: https://books.google.com/ books?id=aoZ4QljK4YwC
11. Fitzmaurice GM, Laird NM, Ware JH. Applied Longitudinal Analysis [Internet]. Wiley; 2011. (Wiley Series in Probability and Statistics). Available from: https://books. google.com/books?id=qOmxRtdNJpEC
12. Kapur K, Li X, Blood EA, Hedeker D. Bayesian mixed-effects location and scale models for multivariate longitudinal outcomes: An application to ecological momentary assessment data. Stat Med. 2015;34(4):630–51.
13. Carroll RJ, Ruppert D. Transformation and Weighting in Regression [Internet]. Taylor & Francis; 1988. (Chapman & Hall/ CRC Monographs on Statistics & Applied Probability). Available from: https://books. google.com/books?id=I5rGEPJd57AC
14. Balázs K, Hidegkuti I, De Boeck P. Detecting heterogeneity in logistic regression models. Appl Psychol Meas. 2006;30(4):322– 44.
15. Hedeker D, Mermelstein RJ, Demirtas H. An application of a mixed- effects location scale model for analysis of ecological momentary assessment (EMA) data. Biometrics. 2008;64(2):627–34.
16. Hoff PD, Niu X. A covariance regression model. Stat Sin. 2012;22(2):729–53.
17. Rast P, Hofer SM, Sparks C. Modeling Individual Differences in Within-Person Variation of Negative and Positive Affect in a Mixed Effects Location Scale Model Using BUGS/JAGS. Multivariate Behav Res. 2012;47(2):177–200.
18. Vahabi N, Kazemnejad A, Datta S. A Marginalized Overdispersed Location Scale Model for Clustered Ordinal Data. Sankhya B. 2018;80:103–34.
19. Aitkin M. Modelling Variance Heterogeneity in Normal Regression Using GLIM. Appl Stat. 1987;36(3):332.
20. White H. Estimating Regression Models with Multiplicative Heteroscedasticity. Econometrica. 1976;48(4):817–38.
21. Vasseur H. Prediction of tropospheric scintillation on satellite links from radiosonde data. IEEE Trans Antennas Propag. 1999;47(2):293–301.
22. Gill N, Hedeker D. Fast estimation of mixed-effects location-scale regression models. Stat Med. 2023 Apr 30;42(9):1430-1444.
23. Renò R, Rizza R. Is volatility lognormal? Evidence from Italian futures. Phys A Stat Mech its Appl. 2003;322:620–8.
24. Spiegelhalter D, Thomas A, Best N, Lunn D. WinBUGS User Manual.2003. Available from: internet: http://www.mrc-bsu. cam.ac.uk/bugs.
25. Gelman A, Carlin JB, Stern HS, Dunson DB, Vehtari A, & Rubin DB. Bayesian Data Analysis (3rd ed.). Chapman and Hall/CRC. 2013.
26. Spiegelhalter DJ, Best NG, Carlin BP, Van Der Linde A. Bayesian measures of model complexity and fit. J R Stat Soc Ser B Stat Methodol. 2002;64(4):583–616.
27. Meyer, R. Deviance Information Criterion (DIC). Wiley StatsRef: Statistics Reference Online.2016; 1–6.
28. Aljurf M, Weisdorf D, Alfraih F, Szer J, Müller C, Confer D, et al. “Worldwide Network for Blood & Marrow Transplantation (WBMT) special article, challenges facing emerging alternate donor registries.” Bone Marrow Transplant [Internet]. 2019;54(8):1179–88.
29. Chao N. Clinical manifestations, diagnosis and grading of acute graft-versus- host disease. UpToDate, Negrin R, Accessed sep 2022. Available from: https://www.uptodate.com/contents/clinical-manifestations- diagnosis-and-grading-of-acute-graft-versus- host-disease.
30. Jagasia M, Arora M, Flowers MED, Chao NJ, McCarthy PL, Cutler CS, et al. Risk factors for acute GVHD and survival after hematopoietic cell transplantation. Blood. 2012;119(1):296–307.
31. Cannell P, Herrmann R. Urate metabolismduringbonemarrowtransplantation. Bone Marrow Transplant. 1992;10:337–9.
32. Haen SP, Eyb V, Mirza N, Naumann A, Peter A, Löffler MW, et al. Uric acid as a novel biomarker for bone-marrow function and incipient hematopoietic reconstitution after aplasia in patients with hematologic malignancies. J Cancer Res Clin Oncol. 2017;143(5):759–71.
33. Ostendorf BN, Blau O, Uharek L, Blau IW, Penack O. Association between low uric acid levels and acute graft-versus-host disease. Ann Hematol. 2015;94(1):139–44.
34. Liu L, Yu Z. A likelihood reformulation method in non-normal random effects models. Stat Med. 2018;27:3105–3124.
| Files | ||
| Issue | Vol 9 No 2 (2023) | |
| Section | Original Article(s) | |
| DOI | https://doi.org/10.18502/jbe.v9i2.14628 | |
| Keywords | ||
| Correlated binary data Heteroscedasticity Bayesian mixed-effects model Variance modelling Random scale effects Longitudinal data analysis | ||
| Rights and permissions | |
|
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
