Person-mean centering has been recommended for disaggregating between-person and within-person effects when modeling time-varying predictors. Multilevel modeling textbooks recommended global standardization for standardizing fixed effects. An aim of this study is to evaluate whether and when person-mean centering followed by global standardization can accurately estimate fixed-effects within-person relations (the estimand of interest in this study) in multilevel modeling. We analytically derived that global standardization generally yields inconsistent (asymptotically biased) estimates for the estimand when between-person differences in within-person standard deviations exist and the average within-person relation is nonzero. Alternatively, a person-mean-SD standardization (P-S) approach yields consistent estimates. Our simulation results further revealed (1) how misleading the results from global standardization were under various circumstances and (2) the P-S approach had accurate estimates and satisfactory coverage rates of fixed-effects within-person relations when the number of occasions is 30 or more (in many conditions, performance was satisfactory with 10 or 20 occasions). A daily diary data example, focused on emotional complexity, was used to empirically illustrate the approaches. Researchers should choose standardization approaches based on theoretical considerations and should clearly describe the purpose and procedure of standardization in research articles.
The most widely used statistical model for conducting moderation analysis is the moderated multiple regression (MMR) model. In MMR modeling, missing data could pose a challenge, mainly because the interaction term is a product of two or more variables and thus is a nonlinear function of the involved variables. In this study, we consider a simple MMR model, where the effect of the focal predictor X on the outcome Y is moderated by a moderator U. The primary interest is to find ways of estimating and testing the moderation effect with the existence of missing data in X. We mainly focus on cases when X is missing completely at random (MCAR) and missing at random (MAR). Three methods are compared: (a) Normal-distribution-based maximum likelihood estimation (NML); (b) Normal-distribution-based multiple imputation (NMI); and (c) Bayesian estimation (BE). Via simulations, we found that NML and NMI could lead to biased estimates of moderation effects under MAR missingness mechanism. The BE method outperformed NMI and NML for MMR modeling with missing data in the focal predictor, missingness depending on the moderator and/or auxiliary variables, and correctly specified distributions for the focal predictor. In addition, more robust BE methods are needed in terms of the distribution mis-specification problem of the focal predictor. An empirical example was used to illustrate the applications of the methods with a simple sensitivity analysis. (PsycINFO Database Record
1 September 2018 in Psychology of Sport and Exercise
From a longitudinal study, we have repeatedly measured data from multiple individuals at multiple occasions. For each individual, the relation between 2 variables can be measured by the Pearson’s correlation. The question is how to aggregate the multiple correlations and conduct statistical inference on the aggregated intra-individual correlation. Several methods are proposed to aggregate and test intra-individual correlations: (a) a meta-analysis method based on Fisher’s Z transformed correlations, (b) a meta-analysis method based on the Pearson’s correlations, and (c) a multilevel modeling method using data standardized within each individual. The performance of the methods after bias corrections was compared using simulations with considering factors including numbers of individuals, numbers of time points, population effect sizes, and their distribution forms (homogeneous vs heterogeneous). The results from the simulation studies show that estimation biases were found using the meta-analytic methods and suggestions on when and how to correct biases were provided based on the simulation results. Furthermore, the performance of the 3 methods after necessary bias corrections was found to be comparable and reasonably good, indicating that all 3 methods worked for aggregating and testing intra-individual correlations. An empirical daily diary data set was then used to illustrate the applications of the 3 methods. The assumptions, advantages and disadvantages, and possible extensions of the 3 methods were discussed.
In the current study, extending from the cross-lagged panel models (CLPMs) in Cole and Maxwell (2003), we proposed the multilevel autoregressive mediation models (MAMMs) by allowing the coefficients to differ across individuals. In addition, Level-2 covariates can be included to explain the interindividual differences of mediation effects. Given the complexity of the proposed models, Bayesian estimation was used. Both a CLPM and an unconditional MAMM were fitted to daily diary data. The 2 models yielded different statistical conclusions regarding the average mediation effect. A simulation study was conducted to examine the estimation accuracy of Bayesian estimation for MAMMs and consequences of model mis-specifications. Factors considered included the sample size (N), number of time points (T), fixed indirect and direct effect sizes, and Level-2 variances and covariances. Results indicated that the fixed effect estimates for the indirect effect components (a and b) and the fixed effects of Level-2 covariates were accurate when N >= 50 and T >= 5. For estimating Level-2 variances and covariances, they were accurate provided a sufficiently large N and T (e.g., N >= 500 and T >= 50). Estimates of the average mediation effect were generally accurate when N >= 100 and T >= 10, or N >= 50 and T >= 20. Furthermore, we found that when Level-2 variances were zero, MAMMs yielded valid inferences about the fixed effects, whereas when random effects existed, CLPMs had low coverage rates for fixed effects. DIC can be used for model selection. Limitations and future directions were discussed.