0
$\begingroup$

I'm struggling to find an appropriate way to analyse the model attached with the given data situation (see below). I would appreciate to hear some of your ideas on how to tackle this (and, if available, would appreciate references to work testing similar models).

Model

Given situation: Daily study with:

n = 500+ t = 30 daily assessments - however, not every participant responded to all 30 daily assessments

I would like to test (dark arrows) - reciprocal cross lagged links between Variable A & B - day-specific moderators C and D

I want to allow for (grey arrows) - autoregressive effects of A and B - correlations from A & B at the same point in time

Software that I considered using: R, Mplus, SPSS

I was thinking about RI-CLPM, DSEM, Multilevel modelling - but there all have their disadvantages.

My preferred way of analysis (DSEM in Mplus) unfortunately does not allow to test for time-varying Level 1 moderation.

Anyone having any other solutions how to test this model? Looking for any techniques that I might have overseen so far.

$\endgroup$
  • $\begingroup$ Given your data, this is a very complex model. I'm not sure you can use off the shelf approaches to estimate such a model. Without rolling your own Bayeisan model using JAGS or STAN, you may want to think about simplifying the model and going with one of your lesser alternatives. The cross-lagged part is doable in SEM but not MLM whereas the interactions are more feasible in MLM. Which parts of the model can you simplify and live without? $\endgroup$ – Erik Ruzek Jun 4 at 22:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.