Responsible Research Methods talk: Clarifying Concerns About Cross-Lagged Effects in Longitudinal Analyses
- Date
- Monday 16 March 2026
- Time
- Explanation
- A Responsible Research Methods meeting for anyone to join!
- Address
- Agora (FSW)
- Room
- 0B13
Clarifying Concerns About Cross-Lagged Effects in Longitudinal Analyses
In this special edition of the Responsible Research Methods labmeeting (a research team within the psychology Methods and Statistics unit), Richard Lucas will give a brief intro about a common model used in longitudinal data and it's problems, and will then focus on the reactions to the critiques and work showing how bad things are (such as that you can always get a significant "causal effect", even with randomly selected pairs of variables).
Richard Lucas has extensive experience within the field of improving psychological science, and would love to connect with anyone at Leiden on this topic. Next to that, Richard has the following research interests: Happiness and subjective well-being, positive emotions, the effects of life events on life satisfaction, personality traits, extraversion, personality assessment.
More details about the statistical part of Richard's work: The simple lag-1 cross-lagged panel model (CLPM) is one of the most widely used models to estimate causal effects in (non-experimental) longitudinal data. Many concerns about the model have been raised, and in response to these critiques, some methodologists have proposed alternatives, whereas others have defended the model. In this talk, initial concerns with the model will be introduced and the problems with existing defenses. In addition, simulation-based and empirical analyses will be presented that illustrate the severity of the problems with the CLPM, along with new concerns about some of the proposed alternatives.
Open to all!