New blog post! explains why cross-lagged panel models are not going to save you.
Essay reminds us that longitudinal data do not allow us to infer causality
I’ve written a (casual, short & non-technical) blog post about what these models test — Granger causality — and how it’s not necessarily what people think it is. Might be helpful to understand what the results can or can’t tell you!
Very clear and valuable post from for the : longitudinal data doesn't give you a free ticket for causality Plus I've learned a new concept: granger causality!
The standard approach would be something like a cross lagged panel model, but this blog post illustrates some of the challenges therein: I think it’s not a straightforward issue, but something like this might be a good first step!
Anyone doing personality research or other research making inferences about causality from longitudinal data should know about Granger causality. Excellent accessible blog post.
New article by argues that longitudinal data can't be (typically) used to assert causality