Structural Equation Modeling: A Multidisciplinary Journal最新文献

Abstract

The logistic and confined exponential curves are frequently used in studies of growth and learning. These models, which are nonlinear in their parameters, can be estimated using structural equation modeling software. This paper proposes a single combined model, a weighted combination of both models. Mplus, Proc Calis, and lavaan code for the model are provided. Monte Carlo simulations varying the number of measurement occasions (5, 10, and 15), internal consistency (α = 0.5, 0.7, and 0.8), and sample size (N = 1,000, 500, and 300) were examined to understand whether the model can be successfully fit with SEM software. Convergence failures were appreciable when model parameters were equal to special cases of logistic or confined exponential curves. At least ten measurement occasions and a moderate degree of reliability (α > 0.7) were required to identify the model as superior to its stand-alone alternatives.

Abstract

In Structural Equation Modeling (SEM), the measurement part and the structural part are typically estimated simultaneously via an iterative Maximum Likelihood (ML) procedure. In this study, we compare performance of the standard procedure to the Structural After Measurement (SAM) approach, where the structural part is separated from the measurement part. One appealing feature of the latter multi-step procedure is that it extends the scope of possible estimators, as now also non-iterative methods from factor-analytic literature can be used to estimate the measurement models. In our simulations, the SAM approach outperformed vanilla SEM in small to moderate samples (i.e., no convergence issues, no inadmissible solutions, smaller MSE values). Notably, this held regardless of the estimator used for the measurement part, with negligible differences between iterative and non-iterative estimators. This may call into question the added value of advanced iterative algorithms over closed-form expressions (which generally require less computational time and resources).

Abstract

Intensive longitudinal data has been widely used to examine reciprocal or causal relations between variables. However, these variables may not be temporally aligned. This study examined the consequences and solutions of the problem of temporal misalignment in intensive longitudinal data based on dynamic structural equation models. First the impact of temporal misalignment on parameter estimation were investigated in a simulation study, which showed that temporal misalignment led to incomparable cross-lagged effects between variables. Then, two solutions, model adjustment and data interpolation, were proposed, and their performance was compared with those of the naive estimation which blindly treating temporally misaligned data as aligned. The simulation results supported the effectiveness of the model adjustment method over the other two methods. Finally, all three methods were applied to two empirical data collected by daily diaries and empirical sampling method, and recommendations were made for collecting and analyzing intensive longitudinal data.

Abstract

Several variants of the autoregressive structural equation model were suggested over the past years, including, for example, the random intercept autoregressive panel model, the latent curve model with structured residuals, and the STARTS model. The present work shows how to place these models into a mixed-effects model framework and how to estimate them in mixed-effects model software, namely the R package nlme. We also show how nlme can be used to fit extensions of these models, for example, models that do not assume equally spaced time intervals between measurement occasions (i.e., continuous time models). Overall, our expositions show that autoregressive structural equations models and mixed-effects models are closely related. We think that this insight eases researchers to understand the differences between the variants of the autoregressive structural equation model and also allows them to profitably link the two different modeling perspectives.

Abstract

The latent growth model (LGM) is a popular tool in the social and behavioral sciences to study development processes of continuous and discrete outcome variables. A special case are frequency measurements of behaviors or events, such as doctor visits per month or crimes committed per year. Probability distributions for such outcomes include the Poisson or negative binomial distribution and their zero-inflated extensions to account for excess zero counts. This article demonstrates how to specify, evaluate, and interpret LGMs for count outcomes using the Mplus program in the structural equation modeling framework. The foundations of LGMs for count outcomes are discussed and illustrated using empirical count data on self-reported criminal offenses of adolescents (N = 1,664; age 15–18). Annotated syntax and output are presented for all model variants. A negative binomial LGM is shown to best fit the crime growth process, outperforming Poisson, zero-inflated, and hurdle LGMs.

Abstract

In the past decade, moderated mediation analysis has been extensively and increasingly employed in social and behavioral sciences. With its widespread use, it is particularly important to ensure the moderated mediation analysis will not bring spurious results. Spurious effects have been studied in both mediation and moderation analysis, but this issue remains unexplored in moderated mediation analysis. To fill this gap, we examined the conditions under which a spurious moderated mediation effect in a dual stage moderated mediation model might occur. Specifically, with a hypothetical example and three theorems, we illustrated how the index of moderated moderated mediation may conclude a moderated mediation effect which does not actually exist. As a remedy to rule out the spurious results, we proposed two methods which are simple and easy to implement. Based on the simulation results, we offer researchers some practical guidelines to apply the methods in empirical research.

Abstract

Parallel process latent growth curve mediation models (PP-LGCMMs) are frequently used to longitudinally investigate the mediation effects of treatment on the level and change of outcome through the level and change of mediator. An important but often violated assumption in empirical PP-LGCMM analysis is the absence of omitted confounders of the relationships among treatment, mediator, and outcome. In this study, we analytically examined how omitting pretreatment confounders impacts the inference of mediation from the PP-LGCMM. Using the analytical results, we developed three sensitivity analysis approaches for the PP-LGCMM, including the frequentist, Bayesian, and Monte Carlo approaches. The three approaches help investigate different questions regarding the robustness of mediation results from the PP-LGCMM, and handle the uncertainty in the sensitivity parameters differently. Applications of the three sensitivity analyses are illustrated using a real-data example. A user-friendly Shiny web application is developed to conduct the sensitivity analyses.