Beyond the Cross-Lagged Panel Model: Next-generation statistical tools for analyzing interdependencies across the life course

https://doi.org/10.1016/j.alcr.2018.10.002Get rights and content

Abstract

For decades, researchers have employed the Cross-Lagged Panel Model (CLPM) to analyze the interactions and interdependencies of a wide variety of inner- or supra-individual variables across the life course. However, in the last years the CLPM has been criticized for its underlying assumptions and several alternative models have been proposed that allow to relax these assumptions. With the Random-Intercept CLPM, the Autoregressive Latent Trajectory Model with Structured Residuals, and the Dual Change Score Model, we describe three of the most prominent alternatives to the CLPM and provide an impression about how to interpret the results obtained with these models. To this end, we illustrate the use of the presented models with an empirical example on the interplay between self-esteem and relationship satisfaction. We provide R and Mplus scripts that might help life course researchers to use these novel and powerful alternatives to the CLPM in their own research.

Section snippets

Cross-Lagged Panel Model

The CLPM is the standard model to examine rank-order changes and time-lagged associations between two longitudinally assessed variables (see Fig. 1 for a CLPM with four measurement waves). It provides two types of coefficients that are of particular interest to life course researchers. First, the autoregressive paths (a1 and a2 in Fig. 1) provide information on the rank-order stability of x or y, respectively (i.e., the stability of inter-individual differences; Mund, Zimmermann, & Neyer, 2018

Three alternatives to the CLPM

In the following, we describe the Random-Intercept CLPM (RI-CLPM), the Autoregressive Latent Trajectory Model with Structured Residuals (ALT-SR), and the Dual Change Score Model (DCSM). After having introduced these models, we will compare them to each other concerning some central aspects as well as to the multilevel growth model and the fixed effects regression model.

Empirical illustration

After having introduced the CLPM and three contemporary alternative approaches, we illustrate the interpretation of all models by an empirical example on the interplay between self-esteem (SE) and relationship satisfaction (RS). The reciprocal influences between these two variables have often been studied to investigate to what extent aspects of social relationships are influenced by and likewise further influence trait-like personality characteristics (e.g., Erol & Orth, 2014; Mund, Finn,

Conclusion

Trying to understand the life course of individuals is an ambitious endeavor that requires a tailored set of tools regarding study design, data collection, and data analysis (Bernardi et al., 2018). Across their life course, individuals navigate through and interact in different contexts. These interactions between two complex systems (individual and environment) as well as interactions within individuals create a set of interdependencies that need to be investigated when trying to understand

References (40)

  • P.D. Allison

    Fixed effects regression models

    (2009)
  • L. Bernardi et al.

    The life course cube

    Advances in Life Course Research

    (2018)
  • D. Berry et al.

    On the practical interpretability of cross-lagged panel models: Rethinking a developmental workhorse

    Child Development

    (2017)
  • J.C. Biesanz

    Autoregressive longitudinal models

  • K.A. Bollen et al.

    An overview of the autoregressive latent trajectory (ALT) model

  • J. Brüderl et al.

    Life course research with panel data

    Advances in Life Course Research

    (2018)
  • L. Castro-Schilo et al.

    Using residualized change versus difference scores for longitudinal research

    Journal of Social and Personal Relationships

    (2018)
  • P.J. Curran

    Have multilevel models been structural equation models all along?

    Multivariate Behavioral Research

    (2003)
  • P.J. Curran et al.

    The best of both worlds: Combining autoregressive and latent curve models

  • P.J. Curran et al.

    The separation of between-person and within-person components of individual change over time: A latent curve model with structured residuals

    Journal of Consulting and Clinical Psychology

    (2014)
  • P.J. Curran et al.

    Twelve frequently asked questions about growth curve modeling

    Journal of Cognition and Development

    (2010)
  • R.Y. Erol et al.

    Development of self-esteem and relationship satisfaction in couples: Two longitudinal studies

    Developmental Psychology

    (2014)
  • K.J. Grimm et al.

    Recent changes leading to subsequent changes: Extensions of multivariate latent difference score models

    Structural Equation Modeling

    (2012)
  • K.J. Grimm et al.

    Growth modeling: Structural equation and multilevel modeling approaches

    (2017)
  • F. Hamagami et al.

    Advanced studies of individual differences linear dynamic models for longitudinal data analysis

  • E.L. Hamaker

    Conditions for the equivalence of the autoregressive latent trajectory model and a latent growth curve with autoregressive disturbances

    Sociological Methods and Research

    (2005)
  • E.L. Hamaker et al.

    A critique of the cross-lagged panel model

    Psychological Methods

    (2015)
  • C. Hertzog et al.

    Assessing psychological change in adulthood: An overview of methodological issues

    Psychology and Aging

    (2003)
  • J.J. Hox

    Multilevel analysis: Techniques and applications

    (2010)
  • J. Huinink et al.

    Panel analysis of intimate relationships and family dynamics (pairfam): Conceptual framework and design

    Zeitschrift für Familienforschung

    (2011)
  • Cited by (0)

    View full text