Zhonglin Wen’s research while affiliated with South China Normal University and other places

Intensive longitudinal data (ILD) is increasing in fields such as psychology and management, yet research on analytical methods for ILD remains relatively scant. Traditionally, the ILD is statistically modeled as a two-level structure, with Level 1 being the time and Level 2 being individuals. Especially, existing analytical methods treat longitudinal moderated mediation as multilevel moderated mediation, without considering the lagged relationship between variables. A possible solution is to use dynamic structural equation modeling (DSEM) for ILD moderated mediation analysis. DSEM has recently been used for analyzing intensive longitudinal mediation (ILMed; McNeish & MacKinnon, 2022; Fang et al., 2024) and intensive longitudinal moderation (ILMod; Speyer et al., 2024). However, it remains unclear how DSEM can be employed in analyzing intensive longitudinal moderated mediation (ILMM). The purpose of this paper is to combine ILMed and ILMod based on DSEM and propose a method of moderated mediation analysis that takes into account the temporal order between variables. For the 1-1-1 ILMed model where all variables are measured at Level 1 (i.e., all variables are ILD), it might be moderated by variables of Level 1 or Level 2. However, for the 2-1-1 ILMed model (i.e., only the independent variable is measured at Level 2) and the 2-2-1 ILMed model (i.e., only the dependent variable is measured at Level 1), they could only be moderated by variables of Level 2. Therefore, there are four basic types of ILMM models: 2-1-1 ILMed moderated by a level 2 moderator, 2-2-1 ILMed moderated by a level 2 moderator, 1-1-1 ILMed moderated by a level 2 moderator, and 1-1-1 ILMed moderated by a level 1 moderator. This paper describes in detail how to construct the above four ILMM models with DSEM, so that empirical researchers can understand which kind of ILMM model meets their needs and how to analyze it. Mplus codes for analyzing all these ILMM models are provided. A simulation study is conducted to examine the estimation accuracy of the 1-1-1 ILMed moderated by a level 2 moderator, with the following factors taken into account: sample size (N), number of time points (T), indirect effect sizes, and Level-2 variances and covariances. Results show that the estimates for the average mediation effect components (a and b) and the average mediation effect are generally accurate when N≥100 and T≥10. However, a sufficiently large N and T (e.g., T≥20) are required in order to obtain accurate estimation of Level-2 variances. Lastly, we discuss assumptions and the extensions of ILMM based on DSEM. As usual, the models used in this paper are based on the assumption that the time series is stationary. Otherwise, residual DSEM can be employed to detrend in ILMM analysis.

With the widespread use of intensive longitudinal data in the social sciences, how to analysis intensive longitudinal mediation (ILM) effect has attracted the attention of many researchers. A conventional approach is using multilevel models or multilevel structural equation models. In that case, the temporal sequence of variables is ignored, with the dynamic relationship between variables remaining unexplored. In this paper, taking 1-1-1 [i.e., 1 (independent at Level 1)–1 (mediator at Level 1)–1 (dependent at Level 1)] ILM model as an example, we summarize five types of ILM analysis approaches: 1) multilevel autoregressive model (MAM); 2) residual multilevel autoregressive model (RMAM); 3) dynamic structural equation model (DSEM); 4) residual dynamic structural equation model (RDSEM). 5) cross-classified dynamic structural equation model (cross-classified DSEM). In both RMAM and RDSEM, we first statistically remove the trend from the variables with a regression of each variable on time, and then construct the ILM model using the residuals from the previous step. There are two methods developed for analyzing the ILM effect. One of them is based on manifest variables. The multilevel model and the temporal sequence of variables are combined to develop MAM, which is extended to RMAM when detrending is required. The other method is based on latent variables. The multilevel structural equation model and the temporal sequence of variables are combined to develop DSEM, which is extended to RDSEM when detrending is required. In the present study, we propose a procedure to conduct ILM analysis. The first step is to decide whether one would like to get ILM effects with time changes. If yes, cross-classified DSEM should be adopted to analyze ILM effects with time changes. It is a suitable detrending approach to include time as a time-varying covariate in the cross-classified DSEM. Otherwise, one will proceed with the second step, which is to decide whether it is necessary to detrend. If detrending is necessary, RDSEM should be adopted to analyze ILM effects with individual changes. Otherwise, DSEM should be adopted to analyze ILM effects with individual changes. The third step is to check whether RDSEM or DSEM converges. If either of them converges, their result should be reported. Otherwise, MAM or RMAM should be adopted to analyze ILM effects with individual changes. This paper exemplifies how to conduct the proposed procedure and provides corresponding Mplus and R codes. Directions for future research on mediation analysis of intensive longitudinal data are discussed at the end of the paper. According to the level of the variable, there are seven ILM models: 1-1-1, 2-1-1, 2-2-1, 2-1-2, 1-2-2, 1-2-1 and 1-1-2. However, only the first four mediation models are discussed in the existing literature, and it is found that as long as there is a level-2 variable in the ILM model, the mediation effect can only occur at level-2. The analytical methods of the last three models, 1-2-2, 1-2-1 and 1-1-2 ILM, can be inferred from those of the first four.

Endogeneity is a critical concern in research methodologies, yet it has been insufficiently addressed in longitudinal cross-lagged models, leading to potentially biased outcomes. This study scrutinized the endogeneity inherent in the cross-lagged panel model (CLPM), a prevalent and representative framework in longitudinal studies. We evaluated the efficacy of the instrumental variables (IV) methods, specifically focusing on both the auxiliary IVs (AIVs) and the model-implied IVs (MIIVs), in mitigating endogeneity issues. Simulation results indicated that endogeneity induced bias in CLPM, notably overestimating cross-lagged effects and thereby amplifying the apparent causal relationships. AIV-CLPM showed a smaller, yet still unacceptably high bias, along with low robustness and elevated type I error rates. In contrast, the MIIV-CLPM produced more accurate estimates with fewer type I errors, and, given sufficient observations, it achieved moderate statistical power. An extended simulation incorporating the random-intercept CLPM supported these findings, highlighting the generalizability of this approach. Furthermore, an empirical illustration demonstrated the practicality and feasibility of the MIIV-CLPM. Overall, MIIV is proven to be a superior modeling option within cross-lagged frameworks, effectively mitigating biases caused by endogeneity.

In an era marked by the rapid expansion of Internet communication, unmediated interpersonal relationships are increasingly rare. In the workplace, however, loneliness can have a range of negative consequences. This study investigated the impact of workplace loneliness on organizational trust in a Chinese context. For data collection, 65 work teams (65 leaders and 240 members) were recruited from two private and two state-owned enterprises. Team leaders took the Loneliness at Work Scale (LAWS) while team members completed the LAWS, the Leader-Member Exchange Scale (LMX-7), and the Organizational Trust Scale. A cross-level moderated mediation model was applied using hierarchical linear modeling (HLM). Results show that members? workplace loneliness is a cross-level mediator between leaders? workplace loneliness and members? organizational trust. In addition, leader-member exchange (LMX) was a cross-level moderator in the association between leaders? and members? workplace loneliness. These findings enhance our understanding of the negative impact of workplace loneliness on organizations and shed light on the influence that team leaders exert on their members, providing valuable insights into work team management.

The investigation of relationships among variables is the main focus of empirical research in psychology and other social science disciplines. Many empirical studies based on questionnaire surveys involve the influence relationship between variables. However, the lack of a universally accepted definition for this concept has led to ambiguity, and it is often conflated with causal or correlational relationships, which may cause problems for studies on mediating effects. This article defines the influence relationship as a directional correlation, clarifying relationships among correlation, influence and causation in terms of denotation and connotation. We propose several ways to identify evidence for modeling the influence relationship. Our discussion involves the joint effects of multiple factors influencing a dependent variable. The paper provides theoretical support for studying relationships between variables in questionnaire research.

Currently, dynamic structural equation modeling (DSEM) and residual DSEM (RDSEM) are commonly used in testing intensive longitudinal data (ILD). Researchers are interested in ILD mediation models, but their analyses are challenging. The present paper mathematically derived, empirically compared, and step-by-step demonstrated three types (i.e., 1-1-1, 2-1-1, and 2-2-1) of intensive longitudinal mediation (ILM) analyses based on DSEM and RDSEM models. Specifically, each ILM model was demonstrated with a simulated example and illustrated with the corresponding annotated Mplus codes. We compared two types of detrending methods in mediation analyses and showed that RDSEM was superior to DSEM because the latter included the time tj variable as a Level 1 predictor. Lastly, we extended ILM analyses based on DSEM and RDSEM to multilevel autoregressive mediation models, cross-classified DSEM, and intensive longitudinal moderated mediation models. KEYWORDS Dynamic structural equation modeling; intensive longitudinal data; mediation effect; moderated mediation model; residual dynamic structural equation modeling Analyses of mediation, particularly in longitudinal designs, are important in causation analyses. Recently, digital data devices (e.g., digital watch monitoring health) have made possible the collection of intensive longitudinal data (ILD) over a huge number of time points. Massive data improves construct ecological validity but is challenging in analyses. The present study mathematically examined and extended the common analytical strategies. Simulated data were then used to compare and demonstrate their proper use. Furthermore, our detailed Mplus programming codes provide step-by-step guidance to applied researchers on how ILD mediation models should be analyzed. 1. Causation and Mediation Analyses In psychological and behavioral studies, mediation analyses are important in investigating causal chains among the independent variable X, the mediator M, and the dependent variable Y (Baron & Kenny, 1986). The mediation effect of X on Y through M is often quantified and denoted by a  b, where a and b are the effects of X on M and the effect of M on Y controlling for X (MacKinnon, 2008), respectively. The part of the effect of X on Y that the mediation effect cannot explain is called the direct effect (denoted by c 0).

Background and objectives: Limited research has examined the mediating mechanisms underlying the association between procrastination in academic writing and negative emotional states during the COVID-19 pandemic. In the present study, we examined whether stress coping styles and self-efficacy for self-regulation of academic writing mediated the relationship between procrastination in academic writing and negative emotional states. Design and method: Graduate students (N = 475, 61.7% female, Mage of students at baseline = 29.02 years, SD = 5.72) completed questionnaires at Time 1 (March 2020; Procrastination in Academic Writing and Coping Inventory for Stressful Situations), and Time 2 (June 2020; The Self-Efficacy for Self-Regulation of Academic Writing Scale and Depression, Anxiety, and Stress Scale - 21). Results: Emotion-oriented coping and the self-efficacy for self-regulation of academic writing serially mediated the association between procrastination in academic writing and negative emotional states. Meanwhile, task-oriented coping and self-efficacy for self-regulation of academic writing also serially mediated the association between procrastination in academic writing and negative emotional states. Conclusions: These findings provide a plausible explanation of the roles that stress coping styles and self-efficacy for self-regulation of academic writing play in the association between procrastination in academic writing and negative emotional states.

本书是一本研究中介效应、调节效应、有调节的中介效应和有中介的调节效应分析方法的专著，主要涉及四种效应模型的新近发展。例如，类别变量的中介效应和调节效应、追踪数据的中介效应和调节效应、稳健中介效应、因果中介效应、探索性中介效应、成对数据的主客体互倚性中介效应、两层回归模型的调节效应、多个调节变量的调节效应、两水平被试内设计的调节效应、两层回归模型的有调节的中介效应、有调节的多重中介模型、多个调节变量的有调节的中介模型、有中介的调节模型Ⅱ等。本书的读者对象是有关专业研究人员、应用工作者和研究生。本书的定位是应用取向的，在介绍模型时仅保留了基本的、必要的数学公式，让惧怕数学公式的读者也能轻松阅读。 本书共9章，论述了四种效应的基本概念、统计原理、估计和检验方法及其在社会科学研究中的应用。四种效应模型都拓展到类别变量分析、潜变量分析、多层（嵌套）数据分析、交叉嵌套数据分析、纵向（追踪）数据分析、密集追踪数据分析。对所用的模型，提供了SPSS插件PROCESS（显变量建模）和MLMED（多层数据的显变量中介模型）操作步骤，或者Mplus和R程序（各种数据类型的潜变量建模），以及结果解释。 中介效应和调节效应分析是当前社会科学研究很常用的统计分析方法，目的是研究自变量对因变量的影响机制。中介效应解释了自变量为什么会影响因变量，研究的是影响过程；调节效应解释了什么时候自变量会影响因变量或何时影响较大，研究的是影响条件。还可以将中介变量和调节变量整合到一个模型中，做更加深入的分析。 本书作为中介效应和调节效应的进阶专著，主要讨论的是中介效应和调节效应的模型发展及其应用。中介效应和调节效应模型发展主要有三类：第一类是增加变量从简单模型到复杂模型拓展，例如多个中介变量的中介模型（即多重中介模型）、多个调节变量的调节模型、同时包含中介变量和调节变量的整合模型；第二类是显变量建模发展到潜变量建模，在考虑变量之间影响关系的同时考虑测量误差，以结构方程模型为基础；第三类是针对不同数据类型的建模，即从连续数据发展到类别数据，从常规（单层）数据发展到多层（多水平）数据，从截面（横向）数据发展到追踪（纵向）数据。十年来，方杰教授与我合作，在上面三个类别上都做了不少工作，发表的系列论文受到广大读者的关注（有较高的下载次数和引用次数），多次获得发文期刊的年度优秀论文奖。本书不仅是这些工作的系统梳理和总结，而且尽可能包含国内外同行的新近研究成果。 本书的结构采用树形分类。首先，按数据类型分类：常规数据篇、多层数据篇、纵向数据篇。然后，在每篇中都会讨论中介模型、调节模型、中介与调节并存的整合模型。最后，在每种模型中，适当情形都会包含显变量建模和潜变量建模。 第一篇常规数据篇，包括第一章至第四章，介绍常用的中介变量、调节变量以及两者兼有的整合模型。第一章是简单中介效应分析及其相关议题，包括中介效应检验方法、中介效应量和检验力分析、因果中介分析、类别变量的中介模型、稳健中介效应分析、两水平重复测量的中介模型、潜变量中介模型。第二章是多重中介效应分析，包括并行、链式和混合多重中介模型、类别变量的多重中介模型、两水平重复测量的多重中介模型、潜变量多重中介模型、成对数据的主客体互倚性中介模型、探索性中介效应分析。第三章是调节效应分析，包括中心化策略、标准化估计、简单斜率检验、类别变量的调节效应、两水平重复测量的调节效应、多个调节变量的调节效应、基于两层回归模型的调节效应分析、稳健调节效应分析、潜变量调节效应。第四章是中介和调节的整合模型，即有调节的中介模型、有中介的调节模型（Ι和II型），包括类别变量的有调节的中介模型、有调节的多重中介模型、多个调节变量的有调节的中介模型、潜变量的有调节的中介模型、基于两层回归模型的有调节的中介模型和有中介的调节模型。 第二篇是多层数据篇，包括第五章和第六章，讨论多层数据的中介、调节及其整合模型。第五章是多层中介效应分析，包括基于多层线性模型的情境效应分析和多层中介效应分析、基于多层结构方程模型的情境效应和多层中介效应分析、交叉嵌套数据的情境效应分析和多层中介效应分析。第六章是多层调节效应分析、多层中介和调节的整合模型分析，包括基于多层线性模型和多层结构方程模型的多层调节效应分析、有调节的多层中介效应分析、有中介的多层调节效应分析。 第三篇是纵向数据篇，包括第七章至第九章，讨论纵向数据的中介、调节及其整合模型。第七章是纵向数据分析，包括基于交叉滞后模型、潜变量增长模型、潜变化分数模型、动态结构方程模型、潜在转变分析的十多种纵向数据分析模型。第八章是纵向数据的中介效应分析，包括纵向数据的4种常用中介效应模型，以及这些模型的6种拓展模型。第九章是纵向数据的调节效应分析、纵向数据的中介和调节的整合模型分析，包括6种纵向数据的调节效应模型、4种纵向数据的有调节的中介效应模型。 本书的内容具有三个特色。一是新颖性，参考文献中有三分之一是近5年（2018年-2022年）的。二是系统性，涵盖了（连续或类别的）单层数据、成对数据、多层数据、纵向数据的中介效应模型、调节效应模型、中介和调节的整合模型。既有显变量建模，又有潜变量建模。三是实用性。对于每种数据和模型，基本上都有一个应用实例，展示分析过程和结果解释；给出了适当的Mplus和R分析程序，以及SPSS的PROCESS（显变量建模）和MLMED（多层数据的显变量中介模型）插件的操作步骤；总结出多个流程图，便于应用研究者按图索骥，选择合适的分析方法。

Employees' innovative work behavior (IWB) is one of the key factors in improving organizational competitiveness. Previous studies show that challenge and hindrance stress can impact employees' IWB, but our understanding of the exact mechanism underlying the impact is still limited. The present study employed four scales (Challenge and Hindrance Stress Scale, Thriving at Work Scale, Chinese Emotional Exhaustion Scale, and Employee Innovative Behavior Scale) to collect questionnaire data from 789 employees in diverse organizations via an online platform. A two‐path mediation model was constructed. The results show that: (a) challenge stress positively predicted thriving at work and IWB; (b) thriving at work played a partial mediation effect between challenge stress and IWB; (c) hindrance stress negatively predicted thriving at work and positively predicted emotional exhaustion; and (4) hindrance stress did not directly impact IWB while thriving at work and emotional exhaustion were main mediators in the relationship between hindrance stress and IWB. These findings suggest that employees should sensibly cope with different work stresses, while managers should plan work tasks scientifically and give employees adequate opportunities to learn and rest in order to keep them in a positive state to solve problems and work creatively.