How to apply to Rennes School of Business
We guide you to find the programme that best meets your needs.
School News
Discover what brings the school to life every day: conferences, business partnerships, student projects, and key moments on campus.
Our campuses
Our campuses in Rennes and Paris welcome students in environments conducive to success.
In the realm of cross-lagged panel models and longitudinal data, a new approach has emerged, exploring latent interaction effects. This paper paves the way for researchers, practitioners, and data enthusiasts to navigate the intricate web of relationships within their datasets.
At the heart of this innovative approach lies the exploration of latent variables which represent complex concepts or phenomena that cannot be measured directly. Think of it as dissecting a complex puzzle into two key components: the stable (between-person) and dynamic (within-person) aspects. When examining job satisfaction, consider the between-person aspect as the inherent, enduring trait that differs among individuals—such as an individual’s baseline level of contentment in their work. On the other hand, the within-person dynamic facet captures the daily fluctuations in job satisfaction, influenced by the ever-changing tasks and the ambiance of the workplace—a person’s moment-to-moment experiences contributing to their overall satisfaction. This distinction allows us to precisely identify and understand the distinct within-person and between-person components in the context of job satisfaction. By isolating these components, we uncover purely dynamic latent interactions (WxW), purely stable interactions (BxB), and cross-level interactions (BxW).
To delve deeper into latent interaction modeling in panel data, we adopt a multi-level approach, contemplating two fundamental effects. The first effect, auto-regressive (AR), explores how a variable’s past state influences its future state. The second, cross-lagged (CL), delves into how one variable’s past state impacts another variable’s future state.
In this paper, we equip you with practical tools. We share our Mplus software code and output files, allowing you to unleash the power of latent interaction modeling on your own datasets.
This paper employs structural equation modeling (SEM) to analyze latent interaction effects in cross-lagged panel models. It dissects latent variables into stable (between-person) and dynamic (within-person) components, enabling the exploration of purely dynamic, purely stable, and cross-level interactions. The methodology combines auto-regressive (AR) and cross-lagged (CL) effects, providing a comprehensive framework for modeling latent interactions in panel data.
Here are some examples how this new tool can be applied:
Researchers in Organizational Psychology can apply this research to understand how workplace dynamics, job satisfaction, and work-family conflict evolve over time, contributing to enhanced organizational well-being.
Human Resource Professionals can gain insights into factors influencing employee satisfaction and work-family balance, guiding HR policies for improved well-being and productivity.
Marketing Professionals benefit from a powerful analysis approach. For instance, in consumer satisfaction research, identifying trait-like preferences (BxB) and dynamic fluctuations (WxW) enables marketers to tailor strategies, refining long-term product features and adjusting short-term campaigns. This nuanced approach enhances marketers’ ability to strategically meet evolving consumer needs.
Ozkok, O., Vaulont, M. J., Zyphur, M. J., Zhang, Z., Preacher, K. J., Koval, P., & Zheng, Y. (2022). Interaction Effects in Cross-Lagged Panel Models: SEM with Latent Interactions Applied to Work-Family Conflict, Job Satisfaction, and Gender. Organizational Research Methods, 25(4), 673-715.
Consult the research paper
Ozkok O, Vaulont MJ. Interaction Effects in Cross Lagged Panel Models. Organizational research methods Youtube
