×

Comparison and robustness of the REML, ML, MIVQUE estimators for multi-level random mediation model. (English) Zbl 1516.62423

Summary: This article concentrates on the multi-level random mediation effects model (1-1-1) and reviews the maximum likelihood (ML), restricted maximum likelihood (REML), and minimum variance quadratic unbiased estimation (MIVQUE) estimation methods provided by the SAS MIXED process. This paper uses Monte Carlo simulation to make a comparison of the performance of these estimators under a wide variety of different conditions. First, REML and ML produced equivalent results and both of them outperformed MIVQUE, no matter whether the normality assumption was satisfied. Second, the results indicated that the distribution of the \( \boldsymbol{e_{Yij}}\) does not influence the mediation effect. The deviation of the normal distribution of \( \boldsymbol{b_j}\) or ‘\( \boldsymbol{a_j}\) and \( \boldsymbol{b_j} \)’ affected the mediation effect, particularly in condition that not only the magnitude of the deviation but also the covariance between these two effects were large. This thesis ends with the implications, suggestions and recommendations for the application.

MSC:

62-XX Statistics

Software:

SASmixed; gformula; Stata
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] H. Aguinis, B.K. Boyd, C.A. Pierce, and J.C. Short, Walking new avenues in management research methods and theories: Bridging micro and macro domains, J. Manag. 37 (2011), pp. 395-403.
[2] E.B. Andersen, Asymptotic properties of conditional maximum likelihood estimators, J. R. Stat. Soc. 32 (1970), pp. 283-301. · Zbl 0204.51902
[3] A.R. Barbara, S. Venkatraman, and C. Schlauch, A multi-level exploration of empowerment mediators, Lead. Organ. Dev. J. 35 (2014), pp. 649-667. doi: 10.1108/LODJ-07-2012-0085 · doi:10.1108/LODJ-07-2012-0085
[4] R.M. Baron and D.A. Kenny, The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations, J. Pers. Soc. Psychol. 51 (1986), pp. 1173-1182. doi: 10.1037/0022-3514.51.6.1173 · doi:10.1037/0022-3514.51.6.1173
[5] D.J. Bauer, K.J. Preacher, and K.M. Gil, Conceptualizing and testing random indirect effects and moderated mediation in multilevel models: New procedures and recommendations, Psychol. Methods 11 (2006), pp. 142-163. doi: 10.1037/1082-989X.11.2.142 · doi:10.1037/1082-989X.11.2.142
[6] D.M. Bonner, A study of the technological, instructional, and motivational factors affecting PHR certification exam outcomes, Unpublished dissertation, University of North Texas, 2012.
[7] A.S. Bryk and S.W. Raudenbush, Hierarchical Linear Models: Applications and Data Analysis Methods, 2nd ed., Sage Publications, Thousand Oaks, 2002. · Zbl 1001.62004
[8] R. Daniel, B. De Stavola, and S. Cousens, Gformula: Estimating causal effects in the presence of time-varying confounding or mediation using the g-computation formula, Stata J. 11 (2011), pp. 479-517.
[9] J. De Leeuw and G.G. Kreft, Random coefficient models for multilevel analysis, J. Educ. Stat. 11 (1986), pp. 57-85. · doi:10.2307/1164848
[10] B. De Stavola, R. Daniel, G. Ploubidis, and N. Micali, Mediation analysis with intermediate confounding: Structural equation modeling viewed through the causal inference lens, Am. J. Epidemiol. 181 (2015), pp. 64-80. doi: 10.1093/aje/kwu239 · doi:10.1093/aje/kwu239
[11] X. Fan, A. Felsovalyi, S.A. Sivo, and S.C. Keenan, SAS for Monte Carlo Studies: A Guide for Quantitative Researchers, SAS Institute, Cary, NC, 2002.
[12] L.A. Goodman, On the exact variance of products, J. Amer. Statist. Assoc. 55 (1960), pp. 708-713. doi: 10.1080/01621459.1960.10483369 · Zbl 0099.13603
[13] D.A. Harville, Maximum likelihood approaches to variance component estimation and to related problems, J. Amer. Statist. Assoc. 72 (1977), pp. 320-338. doi: 10.1080/01621459.1977.10480998 · Zbl 0373.62040
[14] M.A. Hitt, P.W. Beamish, S.E. Jackson, and J.E. Mathieu, Building theoretical and empirical bridges across levels: Multilevel research in management, Acad. Manage. J. 50 (2007), pp. 1385-1399. doi: 10.5465/AMJ.2007.28166219 · doi:10.5465/AMJ.2007.28166219
[15] K. Imai, L. Keele, and D. Tingley, A general approach to causal mediation analysis, Psychol. Methods 15 (2010), pp. 309-334. doi: 10.1037/a0020761 · doi:10.1037/a0020761
[16] D.A. Kenny, J.D. Korchmaros, and N. Bolger, Lower level mediation in multilevel models, Psychol. Methods 8 (2003), pp. 115-128. doi: 10.1037/1082-989X.8.2.115 · doi:10.1037/1082-989X.8.2.115
[17] J.L. Krull and D.P. Mackinnon, Multilevel mediation modeling in group-based intervention studies, Eval. Rev. 23 (1999), pp. 418-444. doi: 10.1177/0193841X9902300404 · doi:10.1177/0193841X9902300404
[18] J.L. Krull and D.P. MacKinnon, Multilevel modeling of individual and group level mediated effects, Multivariate Behav. Res. 36 (2001), pp. 249-277. doi: 10.1207/S15327906MBR3602_06
[19] R. Littell, G. Milliken, W. Stroup, and R. Wolfinger, SAS System for Mixed Models, SAS Institute Inc, Cary, NC, 1996.
[20] N.T. Longford, A fast scoring algorithm for maximum likelihood estimation in unbalanced mixed models with nested random effects, Biometrika 74 (1987), pp. 817-827. doi: 10.1093/biomet/74.4.817 · Zbl 0628.62075 · doi:10.1093/biomet/74.4.817
[21] D.P. MacKinnon, Introduction to Statistical Mediation Analysis, Lawrence Erlbaum Associates, New York, 2008.
[22] D.P. Mackinnon, C.M. Lockwood, and J. Williams, Confidence limits for the indirect effect: Distribution of the product and resampling methods, Multivariate Behav. Res. 39 (2004), pp. 99-128. doi: 10.1207/s15327906mbr3901_4
[23] R.F. Manghi, G.A. Paula, and F.J.A. Cysneiros, On elliptical multilevel models, J. Appl. Stat. (2016). doi:10.1080/02664763.2015.1134445. · Zbl 1514.62737
[24] J.E. Mathieu, R.P. Deshon, and D.D. Bergh, Mediational inferences in organizational research: Then, now, and beyond, Organ. Res. Methods 11 (2008), pp. 203-223. doi: 10.1177/1094428107310089 · doi:10.1177/1094428107310089
[25] A.M.S. Mostafa and J.S. Gould-Williams, Testing the mediation effect of person-organization fit on the relationship between high performance HR practices and employee outcomes in the Egyptian public sector, Int. J. Hum. Resour. Manag. 25 (2014), pp. 276-292. doi: 10.1080/09585192.2013.826917
[26] B. Muthén, Applications of causally defined direct and indirect effects in mediation analysis using SEM in Mplus, Manuscript submitted for publication (2011).
[27] H.D. Patterson and R. Thompson, Recovery of inter-block information when block sizes are unequal, Biometrika 58 (1971), pp. 545-554. doi: 10.1093/biomet/58.3.545 · Zbl 0228.62046 · doi:10.1093/biomet/58.3.545
[28] K.A. Pituch, T.A. Whittaker, and L.M. Stapleton, A comparison of methods to test for mediation in multisite experiments, Multivariate Behav. Res. 40 (2005), pp. 1-23. doi: 10.1207/s15327906mbr4001_1
[29] K.J. Preacher and J.P. Selig, Advantages of Monte Carlo confidence intervals for indirect effects, Commun. Methods Meas. 6 (2012), pp. 77-98. doi: 10.1080/19312458.2012.679848
[30] K.J. Preacher, M.J. Zyphur, and Z. Zhang, A general multilevel SEM framework for assessing multilevel mediation, Psychol. Methods 15 (2010), pp. 209-233. doi: 10.1037/a0020141 · doi:10.1037/a0020141
[31] P.E. Shrout and N. Bolger, Mediation in experimental and nonexperimental studies: New procedures and recommendations, Psychol. Methods 7 (2002), pp. 422-445. doi: 10.1037/1082-989X.7.4.422 · doi:10.1037/1082-989X.7.4.422
[32] W.H. Swallow and J.F. Monahan, Monte Carlo comparison of ANOVA, MIVQUE, REML, and ML estimators of variance components, Technometrics 26 (1984), pp. 47-57. doi: 10.1080/00401706.1984.10487921 · Zbl 0548.62051
[33] G. Verbeke, E. Lesaffre, and L.J. Brant, The detection of residual serial correlation in linear mixed models, Stat. Med. 17 (1998), pp. 1391-1402. doi: 10.1002/(SICI)1097-0258(19980630)17:12<1391::AID-SIM851>3.0.CO;2-4 · doi:10.1002/(SICI)1097-0258(19980630)17:12<1391::AID-SIM851>3.0.CO;2-4
[34] C.-R. Yu, K.H. Zou, M.O. Carlsson, and S. Weerahandi, Generalized estimation of the BLUP in mixed-effects models: A comparison with ML and REML, Comm. Statist. Simulation Comput. 44 (2015), pp. 694-704. doi: 10.1080/03610918.2013.790445 · Zbl 1314.62157
[35] Y. Yuan and D.P. MacKinnon, Robust mediation analysis based on median regression, Psychol. Methods 19 (2014), pp. 1-20. doi: 10.1037/a0033820 · doi:10.1037/a0033820
[36] Z. Zhang, M.J. Zyphur, and K.J. Preacher, Testing multilevel mediation using hierarchical linear models: Problems and solutions, Organ. Res. Methods 12 (2009), pp. 695-719. doi: 10.1177/1094428108327450 · doi:10.1177/1094428108327450
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.