Sensitivity of fi ve information criteria to discriminate covariance structures with missing data in repeated measures designs

  1. Pablo Livacic-Rojas 1
  2. Paula Fernández 2
  3. Guillermo Vallejo 2
  4. Ellián Tuero-Herrero 2
  5. Feliciano Ordóñez 2
  1. 1 Universidad de Santiago de Chile
    info

    Universidad de Santiago de Chile

    Santiago de Chile, Chile

    ROR https://ror.org/02ma57s91

  2. 2 Universidad de Oviedo
    info

    Universidad de Oviedo

    Oviedo, España

    ROR https://ror.org/006gksa02

Revista:
Psicothema

ISSN: 0214-9915 1886-144X

Año de publicación: 2020

Volumen: 32

Número: 3

Páginas: 399-409

Tipo: Artículo

DOI: 10.7334/PSICOTHEMA2020.63 DIALNET GOOGLE SCHOLAR lock_openAcceso abierto editor

Otras publicaciones en: Psicothema

Objetivos de desarrollo sostenible

Resumen

Antecedentes: el presente trabajo analiza la efectividad de distintos criterios de información para seleccionar estructuras de covarianza extendiéndolo a diferentes mecanismos de pérdida de datos, la mantención y ajustes de las estructuras de medias y las matrices. Método: se utilizó el método Monte Carlo con 1.000 simulaciones, el software estadístico SAS 9.4 y un diseño de medidas parcialmente repetidas (p=2; q=5). Las variables manipuladas fueron: a) complejidad del modelo; b) tamaño muestral; c) emparejamiento de las matrices de covarianza y tamaño muestral; d) matrices de dispersión; e) forma de distribución de la variable; y f) mecanismo de no respuesta. Resultados: los resultados muestran que todos los criterios de información funcionan bien en el escenario 1 para distribuciones normales y no normales con homogeneidad y heterogeneidad de varianzas. Sin embargo, en los escenarios 2 y 3, todos son precisos con la matriz ARH, aunque, AIC, AICCR y HQICR lo hacen para TOEP y UN. Por otro lado, cuando la distribución no es normal, solo en el escenario 3 funcionan bien AIC y AICCR, matrices más heterogéneas y No Estructurada, con Casos Completo MAR y MCAR. Conclusiones: en consecuencia, para seleccionar la matriz correctamente se recomienda analizar la heterogeneidad, tamaño muestral y distribución de los datos.

Información de financiación

Acknowledgements The authors would like express acknowledge the suggestions of revisors to improve this work has been funded by the Chilean National Fund for Scientific and Technological Development (FONDECYT. Ref.: 1170642) and the Spanish Ministry of Science, Innovation and Universities (Ref: PGC2018-101574-B-I00).

Referencias bibliográficas

  • Aberson, C.L. (2019). Applied Power Analysis for the Behavioral Sciences. Routledge.
  • Blanca, M. J., Arnau, J., López-Montiel, D., Bono, R., & Bendayan, R. (2013). Skewness and kurtosis in real data samples. Methodology:
  • European Journal of Research Methods for the Behavioral and Social Sciences, 9(2), 78-84. https://doi.org/10.1027/1614-2241/a000057.
  • Bono, R., Blanca, M. J., Arnau, J., & Gómez-Benito, J. (2017). Non-normal distributions commonly used in health, education, and social sciences: A systematic review. Frontiers in Psychology-Quantitative Psychology and Measurement, 8, Article 1602, 1-6. https://doi.org/10.3389/ fpsyg.2017.01602
  • Bono, R., Arnau, J., Alarcón, R., & Blanca, M. J. (2020). Bias, precision and accuracy of skewness and kurtosis estimators for frequently used continuous distributions. Symmetry, 12(1), 19. https://doi.org/10.3390/ sym12010019
  • Dong & Peng (2013). Principled missing data methods for researchers. SpringerPlus, 2, 222. https://doi.org/10.1186/2193-1801-2-222
  • Fernández, P., Vallejo G., & Livacic-Rojas P. (2010). Robustez de cinco estadísticos univariados para analizar diseños Split Plot en condiciones adversas [Robustness of fi ve univariate statistics to analyze Split Plot designs in adverse conditions]. Revista Latinoamericana de Psicología, 42(2), 289-309.
  • Fernández, P., Vallejo, G., Livacic-Rojas, P., & Tuero-Herrero, E. (2018). The (Ir)Responsibility of (Under)Estimating Missing Data. Frontiers in Psychology. 9, 556-568. https://doi.org/10.3389/fpsyg.2018.00556
  • Fleishman, A.I. (1978). A method for simulating non-normal distributions. Psychometrika, 43, 521-532. https://doi.org/10.1007/BF02293811
  • Funatogawa, I., & Funatogawa, T. (2019). Longitudinal Data Analysis. Autoregresive Linear Mixed Effects Models. Springer.
  • Garson, G.D. (2020). Multilevel Modeling: Applications in STATA®, IBM® SPSS®, SAS®, R, & HLM™. Sage Publications.
  • Grissom, R., & Kim, J. (2012). Effect size for research: Univariate and multivariate applications. 2nd edition. Routledge.
  • Grund, S., Lüdtke, O., & Robitzsch, A. (2019). Missing data in Multilevel Research. In S.E. Humphrey & J.M. LeBreton (Editors-in-Chief), The Handbook of Multilevel Theory, Measurement and Analysis (pp. 365- 386). American Psychological Association.
  • Heck, R., & Thomas, S.L. (2020). An Introduction to Multilevel Modeling Techniques. MLM and SEM Approaches Using Mplus. Routledge.
  • Leke, C. A., & Marwala, T. (2019). Deep Learning and Missing Data in Engineering Systems (Studies in Big Data Book 48). Springer.
  • Little, R.J.A., & Rubin D.B. (2020). Statistical Analysis with Missing Data (3rd edition). Wiley-Blackwell.
  • Livacic-Rojas P., Vallejo G., & Fernández, P. (2013). Covariance structures selection and Type I Error rates in Split Plot designs. Methodology. European Journal of Research Methods for the Behavioural and Social Sciences, 9(4), 129-136. https://doi.org/10.1027/1614-2241/a000058.
  • Livacic-Rojas P., Vallejo G., & Fernández, P. (2017). Power of modifi ed Brown-Forsythe and Mixed-Model Approaches in Split-Plot Designs. Methodology. European Journal of Research Methods for the Behavioural and Social Sciences, 13(1), 9-22. https://doi.org/10.1027/1614-2241/ a000124
  • Liu, X. (2016). Methods and Applications of Longitudinal Data Analysis. Academic Press.
  • Micceri, T. (1989). The unicorn, the normal curve, and other improbable creatures. Psychological Bulletin, 105, 156-166. https://doi.org/10.1037/ 0033-2909.105.1.156
  • Molenberghs, G., Fitzmaurice, G., Kenward, M., Tsiatis, A., & Verbeke, G. (2015). Handbook of Missing Data Methodology. CRC Press.
  • Raghunathan, T. (2016). Missing Data Analysis in Practice. CRC Press.
  • Raghunathan, T., Berglund, P., & Solenberger, P. (2018). Multiple Imputation in Practice. CRC Press. Ripley, B. D. (1987). Statistical Inference for Spatial Processes. Cambrigde University Press.
  • Sullivan, T.R., Yelland L.N., Lee K.L., Ryan, P., & Salter, A.B. (2017). Treatment of missing data follow-up studies of randomized controlled trials: A systematic review of the literature. Clinical Trials, 14(4), 387- 395. https://doi.org/10.1177/1740774517703319
  • Vale, C. D., & Maurelli, V. A. (1983). Simulating multivariate non normal distributions. Psychometrika, 48, 465-471. https://doi.org/10.1007/ BF02293687.
  • Vallejo, G., Fernández, M. P., Livacic-Rojas, P. E., & Tuero-Herrero, E. (2011a). Selecting the best unbalanced repeated measures model. Behavior Research Methods, 43, 18-36. https://doi.org/10.3758/ s13428-010-0040-1
  • Vallejo, G., Fernández, P., Livacic-Rojas, P., & Tuero-Herrero, E.(2011b). Comparison of modern methods for analyzing repeated measures data with missing values. Multivariate Behavioral Research, 46, 900-937. https://doi.org/10.1080/00273171.2011.625320
  • Vallejo G., Tuero-Herrero E., Núñez J.C., & Rosário P. (2014). Performance evaluation of recent information criteria for selecting multilevel models in Behavioral and Social Sciences. International Journal of Clinical and Health Psychology, 14, 48-57. https://doi.org/10.1016/S1697- 2600(14)70036-5
  • Vallejo, G., Ato, M., Fernández, M.P., Livacic-Rojas, P.E., & Tuero-Herrero, E. (2016). Power analysis to detect treatment effect in longitudinal studies with heterogeneous errors and incomplete data. Psicothema, 28(3), 330-339. https://doi.org/10.7334/psicothema2016.92
  • Vallejo, G., Fernández, M. P., & Livacic-Rojas, P. E. (2018). Data analysis of incomplete repeated measures using a multivariate extension of the Brown-Forsythe procedure. Psicothema, 30(4), 434-441. https://doi. org/10.7334/psicothema2018
  • Vallejo, G., Fernández, M. P., & Livacic-Rojas, P. E. (2019). Sample size estimation for heterogeneous growth curve models with attrition. Behavior Research Methods, 51(3), 1216-1243. https://doi.org/10.3758/ s13428-018-1059-y
  • Wilcox, R. (2017). Introduction to Robust Estimation & Hypothesis Testing (4rd ed.). Elsevier.
  • Yoo, J.E. (2013). Multiple Imputation with Structural Equation Modeling. Using auxiliary variables when data are missing. Academic Publishing.
  • Zhang, Z., & Yuan, K. (2018). Practical Statistical Power Analysis. Using Web Power and R. Isdsa Press.
  • Zhou, X.H., Zhou, Ch., Liu, D., & Ding, X. (2014). Applied Missing Data Analysis in the Health Sciences. Wiley & Sons.