Intensive Longitudinal Data Analyses With Dynamic Structural Equation Modeling

Recent developments in theories and data collection methods have made intensive longitudinal data (ILD) increasingly relevant and available for organizational research. New methods for analyzing ILD have emerged under the multilevel modeling framework. In this article, we first delineate features of ILD (including autoregressive relationships, trends, cycles/seasons, and between-subject variability in temporal trends). We discuss the analytic challenges for handling ILD using traditional analytic tools familiar to organizational researchers (e.g., growth models, single-subject time series analyses). We then introduce a statistical approach for handling ILD from the multilevel modeling framework: dynamic structural equation modeling (DSEM). We provide three examples using simulated data sets to demonstrate how to apply DSEM to examine ILD with a software program familiar to organizational researchers (i.e., Mplus). Finally, we discuss issues related to applying DSEM, including centering, missing data, and sample size.

Automated summary

This article discusses how to handle intensive longitudinal data using the dynamic structural equation modeling method under the multilevel modeling framework, as well as the challenges and solutions that may arise when analyzing ILD.