Uso de polinômios fracionários nos modelos mistos
Conteúdo do artigo principal
Resumo
Os Polinômios Fracionários (FPs), propostos por Royston e colaboradores na década de 1990, tem sido amplamente estudados e aplicados no contexto de modelos de regressão para resolver o problema de não linearidade entre a resposta e covariáveis contínuas. As funções FP fornecem uma alternativa aos polinômios de alta ordem ou splines para lidar com a falta de ajuste. Modelos mistos também podem se beneficiar dessa classe de curvas na presença de não linearidade. A inclusão de funções FP na estrutura de modelos lineares mistos já foi explorada por alguns autores, mas em situações simples, por exemplo, uma única covariável no modelo de intercepto aleatório. Este artigo propõe uma estratégia geral para construção de modelos e seleção de variáveis que incorpora os PFs dentro da estrutura de modelos lineares mistos. A aplicação do método a três conjuntos de dados da literatura, conhecidos por violar a suposição de linearidade, ilustra que é possível resolver o problema de falta de ajuste usando um menor número de termos no modelo do que a abordagem usual via expansões de polinômios convencionais.
Detalhes do artigo
Este trabalho está licenciado sob uma licença Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
Referências
Ambler, G. & Royston, P. Fractional polynomial model selection procedures: investigation of Type I error rate. Journal of Statistical Computation and Simulation 69, 89–108 (2001).
Ambler, G. & Benner, A.mfp: Multivariable Fractional Polynomials. R package version 1.5.2.2 (2022). https://CRAN.R-project.org/package=mfp.
Aregay, M., Shkedy, Z., Molenberghs, G., David, M.-P. & Tibaldi, F. Model-based estimates oflong-term persistence of induced HPV antibodies: a flexible subject-specific approach. Journal of Biopharmaceutical Statistics 23, 1228–1248 (2013).
Bates, D., Mächler, M., Bolker, B. & Walker, S. Fitting Linear Mixed-Effects Models Using lme4. Journal of Statistical Software 67, 1–48 (2015).
Belsley, D. A., Kuh, E. & Welsch, R. E.Regression Diagnostics: Identifying Influential Data and Sources of Collinearity (John Wiley & Sons, New York, 1980).
Box, G. E. P. & Tidwell, P. W. Transformation of the independent variables. Technometrics 4, 531–550 (1962).
Crainiceanu, C. M. & Ruppert, D. Likelihood ratio tests in linear mixed models with one variance component. Journal of the Royal Statistical Society: Series B (Statistical Methodology) 66, 165–185 (2004).
Demidenko, E.Mixed Models: Theory and Applications with R (John Wiley & Sons, Hoboken,New Jersey, 2013).
Drikvandi, R., Verbeke, G., Khodadadi, A. & Nia, V. P. Testing multiple variance components in linear mixed-effects models. Biostatistics 14, 144–159 (2013).
Garcia, E. P. Uso de polinômios fracionários nos modelos mistos. Doutorado em Biometria (Instituto de Biociências/ UNESP, Botucatu, 2019).
Harrison, D. & Rubinfeld, D. L. Hedonic housing prices and the demand for clean air. Journal of Environmental Economics and Management 5, 81–102 (1978).
Lee, O. E. & Braun, T. M. Permutation tests for random effects in linear mixed models. Biometrics 68, 486–493 (2012).
Littell, R. C., Milliken, G. A., Stroup, W. W. & Wolfinger, R. D. SAS System for Mixed Models (SAS Institute Inc, Cary, North Carolina, 1996).
Long, J. & Ryoo, J. Using fractional polynomials to model non-linear trends in longitudinal data. British Journal of Mathematical and Statistical Psychology 63, 177–203 (2010).
Longford, N. T. Random Coefficient Models (Oxford University Press, Oxford, 1993).
Pinheiro, J. C. & Bates, D. M. Mixed-Effects Models in S and S-PLUS (Springer-Verlag, NewYork, 2000).
Pinheiro, J. C., Bates, D. M., Debroy, S., Sarkar, D. & Team, R. C.nlme: Linear and Nonlinear Mixed Effects Models R package version 3.1-137 (2018). https://CRAN.R-project.org/package=nlme.
R Core Team.R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing (Vienna, Austria, 2022). https://www.R-project.org/.
Regier, M. D. & Parker, R. D. Smoothing using fractional polynomials: an alternative to polynomials and splines in applied research. Wiley Interdisciplinary Reviews: Computational Statistics 7, 275–283 (2015).
Rocha, F. M. M. Seleção de estruturas de covariância para dados com medidas repetidas. Mestrado em Estatística (IME/USP, São Paulo, 2004).
Royston, P. & Altman, D. G. Regression using fractional polynomials of continuous covariates: parsimonious parametric modelling. Applied Statistics 43, 429–467 (1994).
Royston, P. & Sauerbrei, W. Improving the robustness of fractional polynomial models by preliminary covariate transformation: A pragmatic approach. Computational Statistics & Data Analysis 51, 4240–4253 (2007).
Royston, P. & Sauerbrei, W. Multivariable model-building: a pragmatic approach to regression anaylsis based on fractional polynomials for modelling continuous variables (John Wiley & Sons, Chichester, England, 2008).
Saurbrei, W., Meier-Hirmer, C., Benner, A. & Royston, P. Multivariable regression model building by using fractional polynomials: Description of SAS, STATA and R programs. Computational Statistics & Data Analysis 50, 3464–3485 (2006).
Singer, J. M., Rocha, F. M. M. & Nobre, J. S. Graphical tools for detecting departures from linear mixed model assumptions and some remedial measures. International Statistical Review 85, 290–324(2017)
Tilling, K., MacDonald-Wallis, C., Lawlor, D. A., Hughes, R. A. & Howe, L. D. Modelling childhood growth using fractional polynomials and linear splines. Annals of Nutrition and Metabolism 65, 129–138 (2014).
Vonesh, E. F. & Carter, R. L. Mixed-effects nonlinear regression for unbalanced repeated measures. Biometrics 48, 1–17 (1992).