The DOI matches its Crossref record; 46 of the 46 references it cites were found in a registry or an index.
Why the Cross-Lagged Panel Model Is Almost Never the Right Choice
Open accessAbstract
The cross-lagged panel model (CLPM) is a widely used technique for examining reciprocal causal effects using longitudinal data. Critics of the CLPM have noted that by failing to account for certain person-level associations, estimates of these causal effects can be biased. Because of this, models that incorporate stable-trait components (e.g., the random-intercept CLPM) have become popular alternatives. Debates about the merits of the CLPM have continued, however, with some researchers arguing that the CLPM is more appropriate than modern alternatives for examining common psychological questions. In this article, I discuss the ways that these defenses of the CLPM fail to acknowledge well-known limitations of the model. I propose some possible sources of confusion regarding these models and provide alternative ways of thinking about the problems with the CLPM. I then show in simulated data that with realistic assumptions, the CLPM is very likely to find spurious cross-lagged effects when they do not exist and can sometimes underestimate these effects when they do exist.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Third, much of the existing evidence relies on cross-sectional designs, which cannot adequately disentangle the temporal ordering among media consumption, attentional control, self-regulation, and learning experience”
From Algorithmic short-video music consumption and academic related functions: a three-wave longitudinal study of university students (2026) · cites this paper as Lucas, 2023
In the citing paper
“The traditional CLPM combines stable between-person differences with within-person fluctuations.”
From Algorithmic short-video music consumption and academic related functions: a three-wave longitudinal study of university students (2026) · cites this paper as Lucas, 2023
Citations
Cites
R: A Language and Environment for Statistical Computing
R Core Team (2025) · doi:10.32614/r.manuals · Cited by 352,181
null (2016) · Philosophy Study · doi:10.17265/2159-5313/2016.09.003 · Cited by 28,317
lavaan : An R Package for Structural Equation Modeling
Rosseel, Y. (2012) · Journal of Statistical Software · doi:10.18637/jss.v048.i02 · Cited by 27,955 · cited 2×
“…dplyr (Version 1.0.10; Wickham et al., 2021), ggplot2 (Version 3.3.6; Wickham, 2016), knitr (Ver sion 1.39; Xie, 2015), lavaan (Version 0.6.12; Rosseel, 2012), mnormt (Version 2.1.1; Azzalini & Genz, 2020), papaja (Version 0.1.1; Aust & Barth, 2020), and rethinking (Version 2.21; McElreath,…”
InfluentialCentering predictor variables in cross-sectional multilevel models: A new look at an old issue.
Enders, C. K., & Tofighi, D. (2007) · Psychological Methods · doi:10.1037/1082-989x.12.2.121 · Cited by 4,462
A critique of the cross-lagged panel model
Hamaker, E. L., Kuiper, R. M., & Grasman, R. P. P. P. (2015) · Psychological Methods · doi:10.1037/a0038889 · Cited by 4,005 · cited 3×
“Considering the multilevel structure of the data and applying appropriate multilevel models (as reflected in the RI-CLPM; Hamaker et al., 2015) is one way to address this issue.”
MethodInfluentialFixed Effects Regression Models
Allison, P. (2009) · doi:10.4135/9781412993869 · Cited by 3,999 · cited 2×
“(2015) pointed out, however, that the CLPM does not adequately account for stable-trait-level associations, and they proposed the random-intercept CLPM (RICLPM) as an alternative (also see Allison, 2009; Berry & Willoughby, 2017; Zyphur et al., 2020).”
BackgroundMethodStatistical Rethinking: A Bayesian Course with Examples in R and Stan
McElreath, R. (2020) · doi:10.1201/9780429029608 · Cited by 1,959
Choosing Prediction Over Explanation in Psychology: Lessons From Machine Learning
Yarkoni, T., & Westfall, J. (2017) · Perspectives on Psychological Science · doi:10.1177/1745691617693393 · Cited by 1,916 · cited 3×
“This type of prediction-focused research has historically been used most often in applied settings (Yarkoni & Westfall, 2017).”
BackgroundMethodInfluentialCausal Analysis with Panel Data
Finkel, S. (1995) · Sage eBooks · doi:10.4135/9781412983594 · Cited by 1,348
Testing prospective effects in longitudinal research: Comparing seven competing cross-lagged models.
Orth, U., Clark, D. A., Donnellan, M. B., & Robins, R. W. (2021) · Journal of Personality and Social Psychology · doi:10.1037/pspp0000358 · Cited by 801 · cited 1×
“…a straightforward way to the type of data that contribute to the effect.7 The RI-CLPM does separate between-persons associations from within-persons associations, but it does not do so by “[relegating] between-person differences . . . to the random intercept factors” (Orth et al., 2021, p. 1026).”
Background
Research works printed but not linked: 2
- Rogosa D. R., 1995, The analysis of change, 3
- Watson N., 2012, Longitudinal and Life Course Studies, 3, 369
Cited by
Self-Determination Theory and Workplace Outcomes: A Conceptual Review and Future Research Directions
McAnally, K., & Hagger, M. S. (2024) · Behavioral Sciences · doi:10.3390/bs14060428 · Cited by 296
“It is important to note that panel designs have increased in complexity and sophistication in recent years to better allow the inference of directional effects while controlling for multiple stability components [65].”
BackgroundDerakhshan, A., & Fathi, J. (2024) · Journal of Multilingual and Multicultural Development · doi:10.1080/01434632.2024.2393705 · Cited by 59
“Additionally, while the cross-lagged panel design (CLPD) offered insights into potential temporal relationships among the variables, it has limitations in establishing accurate causality (Berry and Willoughby 2017; Lucas 2023).”
On Partial Versus Full Mediation and the Importance of Effect Sizes
Ledermann, T., Rudaz, M., & Fritz, M. S. (2025) · Advances in Methods and Practices in Psychological Science · doi:10.1177/25152459251355585 · Cited by 51
“However, both these measures are limited to direct effects, and their calculation can be challenging if models are complex, such as those involving cross-lagged effects and three or more waves (e.g., Cole & Maxwell, 2003; Lucas, 2023).”
The Causal Cookbook: Recipes for Propensity Scores, G-Computation, and Doubly Robust Standardization
Chatton, A., & Rohrer, J. M. (2024) · Advances in Methods and Practices in Psychological Science · doi:10.1177/25152459241236149 · Cited by 57
“…al., 2022), work focusing on aspects such as mediation analysis (e.g., Nguyen et al., 2020; Rohrer et al., 2022) or longitudinal data modeling (e.g., Lucas, 2023; Rohrer & Murayama, 2023), and articles trying to build bridges between frameworks (e.g., Deffner et al., 2022; West & Thoemmes, 2010).”
BackgroundHe, G., Chen, S. H. A., Lin, H., & Su, A. (2024) · Educational Psychology Review · doi:10.1007/s10648-024-09922-w · Cited by 50
“However, it still has some shortcomings in examining reciprocal relationships, thus the causal conclusions drawn need to be considered with caution (Lucas, 2023).”
Metadata
- Authors
- Lucas, R. E.
- Affiliations
- Michigan State University (US)
- Corresponding author
- Lucas, Richard E.
- ORCID
- Lucas, Richard E.
- Venue
- Advances in Methods and Practices in Psychological Science
- Publisher
- SAGE Publications
- Published
- 2023-01
- Volume
- 6
- Issue
- 1
- ISSN
- 2515-2459 (ISSN-L)
- 2515-2467
- OpenAlex
- W4361799298
- Cited by
- 361
- References
- 42
- Citations by year
- 2026: 123 · 2025: 117 · 2024: 86 · 2023: 35
- FWCI
- 73.77
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Advances in Methods and Practices in Psychological Science (publishedVersion, cc-by-nc)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Bayesian Inference
- Social Sciences › Decision Sciences › Management Science and Operations Research › Psychometric Methodologies and Testing
- Social Sciences › Psychology › Experimental and Cognitive Psychology › Mental Health Research Topics
- Keywords
- cross-lagged panel model, reciprocal causation, random intercept cross-lagged panel model, longitudinal data, simulation study
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Why the Cross-Lagged Panel Model Is Almost Never the Right Choice. https://app.cochat.ai/papers/10.1177/25152459231158378
- The paper
- Lucas, R. E. (2023). Why the Cross-Lagged Panel Model Is Almost Never the Right Choice. Advances in Methods and Practices in Psychological Science. https://doi.org/10.1177/25152459231158378
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.