An Alternative Diagnostic Procedure for Meta-Regression

  • Ali Hassan Abuzaid Department of Mathematics, Faculty of Science, Al Azhar University C Gaza, Palestine
  • Enass Abed Department of Applied Statistics, Faculty of Economics and Administrative Sciences Al Azhar University- Gaza, Palestine.
  • Abdu Atta Department of Mathematics, Statistics and Physics, School of Arts and Sciences, Qatar University, Doha, Qatar
  • Esam Mahdi Department of Mathematics, Statistics and Physics, School of Arts and Sciences, Qatar University, Doha, Qatar.
Keywords: Coordinate descent algorithm;, penalized maximum likelihood, meta-analysis, Effect-size, Outliers

Abstract

This paper proposes an alternative procedure for detecting outliers in meta-regression using the penalized maximum likelihood with smoothly clipped absolute deviation penalty function. The coordinate descent algorithm is implemented to estimate the parameters where the cross-validation criterion is used to determine the tuning parameter. Extensive simulation experiments demonstrate the usefulness of our proposed procedure as well as its improved power performance compared to previous procedures. Simulation results demonstrate that the performance has a direct relationship with the number of studies and an inverse relationship with the heterogeneity between studies. An illustrative application with real data, implementing the proposed procedure and others, is given.

Author Biographies

Ali Hassan Abuzaid, Department of Mathematics, Faculty of Science, Al Azhar University C Gaza, Palestine
Ali H. Abuzaid is an Associate Professor in the Department of Mathematics with a concentration in statistics at Al-Azhar University- Gaza, Palestine. He holds a PhD and MSc in Statistics from University of Malaya, Malaysia. He is interested in the development of outlier detection procedures in different types of data and the application on real data. Currently, he is the dean of planning and quality assurance at Al Azhar University.
Enass Abed, Department of Applied Statistics, Faculty of Economics and Administrative Sciences Al Azhar University- Gaza, Palestine.
Department of Applied Statistics, Faculty of Economics and Administrative Sciences, Al Azhar University- Gaza, Palestine
Abdu Atta, Department of Mathematics, Statistics and Physics, School of Arts and Sciences, Qatar University, Doha, Qatar
Department of Mathematics, Statistics and Physics, School of Arts and Sciences, Qatar University, Doha, Qatar
Esam Mahdi, Department of Mathematics, Statistics and Physics, School of Arts and Sciences, Qatar University, Doha, Qatar.
Department of Mathematics, Statistics and Physics, School of Arts and Sciences, Qatar University, Doha, Qatar.

References

E. Abed, On the Detection of Outliers in Meta-regression, Master Thesis, Al Azhar University - Gaza, Palestine, 2018.

H. Alavi Majd, Kh. Najafi Ghobadi, A. Akbarzadeh Baghban, N. Ahmadi, and E. Sajjadi, Detecting and accommodating outliers in meta-analysis for evaluating effect of Albendazole on Ascaris lumbricoides infection, Iran Red. Crescent Medical Journal, vol. 16,no. 5, e17648, 2014.

Z. Y. Algamal, Variable Selection in Count Data Regression Model based on Firefly Algorithm, Statistics, Optimization,& Information Computing, vol. 7, pp. 520-529, 2019.

RD. Baker, and D. Jackson, A new approach to outliers in meta-analysis. Healthcare Management Science, vol. 11, pp. 121–131,2008.

R. Baker, and D. Jackson, New models for describing outliers in meta-analysis, Journal of Research Synthesis Method, vol. 7, no.3, pp. 314–328, 2016.

R. Bangert-Drowns, M. Hurley, and B. Wilkinson, The Effects of School-Based Writing-to-Learn Interventions on Academic Achievement. A Meta-Analysis, vol. 74, pp. 29-58, 2004.

K. Beath, A finite mixture method for outlier detection and robustness in meta-analysis. Research Synthesis Methods, vol. 5, no. 4,pp. 285–293, 2014.

D. A. Belsley, E. Kuh, and R. E. Welsch, Regression Diagnostics: Identifying Influential Data and Sources of Collinearity. : New York: John Wiley & Sons, 1980 .

M. Borenstein, L. V. Hedges, J.P. T. Higgins, and H. R. Rothstein, Introduction to Meta-Analysis. Chichester: John Wiley Sons,2009.

P. Breheny, ncvreg: Regularization Paths for SCAD and MCP Penalized Regression Models. R package version 3.11-1,2019, URL https://cran.r-project.org/web/packages/ncvreg/.

S. E. Brockwell, and I. R. Gordon, A comparison of statistical methods for meta-analysis. Statistics in Medicine, vol. 20, pp. 825–840, 2001.

Gumedze F. N., and D. Jackson A random effects variance shift model for detecting and accommodating outliers in meta-analysis.BMC Medical Research Methodology, vol. 11, no. 19, pp. 1-9, 2011.

L. V. Hedges, and I. Olkin, Statistical Method for Meta-Analysis. Academic Press, INC. Harcouri Brace Jovanovich, Publishers,1985.

J. P. T. Higgins, and S. Green, Cochrane Handbook for Systematic Reviews of Interventions, Cochrane Book Series. John Wiley Sons, 2008.

J. Huang, and H. Xie, Asymptotic oracle properties of SCAD-penalized least squares estimators, Particles, Processes and Inverse Problems, vol. 55, pp. 149-166, 2007.

K. L. Lee, and S. G. Thompson, . Flexible parametric models for random-effect distributions. Statistics in Medicine, vol. 27, pp.418–434, 2008.

S. Saleh, and A. H. Abuzaid, Alternative Robust Variable Selection Procedures in Multiple Regression, Statistics, Optimization & Information Computing, vol. 7, no. 4, pp. 816-825, 2019.

Y. She, and A. Owen, Outlier Detection Using Nonconvex Penalized Regression. Journal of the American Statistical Association,vol. 106, no. 494, pp. 626-639, 2011.

M. Ueki, and K. Fueda, Optimal tunning parameter estimation in maximum penalized method. Annals of the Institute of statistical Mathematics, vol. 63, pp. 413-438, 2010.

W. Viechtbauer, metafor-Meta-Analysis, Package for R, version 2.1-0, 2019. URL http://cran.r-project.org/package=metafor.

W. Viechtbauer, and M. C. Cheung, Outlier and influence diagnostics for meta-analysis. Journal of Research Synthesis Method,vol. 1, pp. 112-125, 2010.

J. Zhang, Y. Liu, and Y. Wu, Exponential regression for censored data with outliers. Journal of Statistical Computation and Simulation, vol. 86, pp. 431-442, 2015.

C. Zou, S. Tseng, and Z. Wang, Outlier detection in general profiles using penalized regression method, IE Transactions,vol. 45,no. 2, pp. 106-117, 2014.

Published
2020-02-17
How to Cite
Abuzaid, A. H., Abed, E., Atta, A., & Mahdi, E. (2020). An Alternative Diagnostic Procedure for Meta-Regression. Statistics, Optimization & Information Computing, 8(1), 54-65. https://doi.org/10.19139/soic-2310-5070-864
Section
Research Articles