|
|
 |
| How many assessment points (waves) in... |
 
|
| Message/Author |
|
|
|
|
Hi
Concerning the method introduced in this paper:
Hamaker, E. L., Asparouhov, T., Brose, A., Schmiedek, F., & Muthén, B. (2018). At the frontiers of modeling intensive longitudinal data... Multivariate Behavioral Research, ...
1-I wonder how many waves of longitudinal data are needed for this type of analysis? for example, does this method work with 3 or 4 time points?
2- In DSEM, only 2 autoregressive and 2 cross-lagged paths are reported (for 2 variables) even when we have more than 2 time points. Does this method constrain all the paths to be equal, or use other methods to summarize several values into a single value?
3- If equality constraints are used, can we remove the constraints in DSEM? I can imagine that in typical longitudinal research that psychologists do with less than 5 time points, this constraint is too restrictive.
many thanks |
|
|
|
|
1. You need at least 10 time points but the method is really intended for longer series with say more than 20 time points, that is, so called intensive longitudinal data. The data requirements are discussed in
Schultzberg, M. & Muthén, B. (2018). Number of subjects and time points needed for multilevel time series analysis: A simulation study of dynamic structural equation modeling. Structural Equation Modeling: A Multidisciplinary Journal, 25:4, 495-515, DOI:10.1080/10705511.2017.1392862. (Supplementary material).
For an intro article, see e.g.
McNeish, D. & Hamaker, E.L. (2018). A primer on two-level dynamic structural equation models for intensive longitudinal data. PsyArXiv. November 28. DOI:10.31234/osf.io/j56bm. (Supplementary material).
2. These paths can vary over time in DSEM but will be well-estimated only with longer time series.
3. With less than 5 time points, regular modeling (not DSEM) is more appropriate, for instance RI-CLPM as in
Hamaker, E. L., Kuiper, R. M., & Grasman, R. P. P. P. (2015). A critique of the cross-lagged panel model. Psychological Methods, 20(1), 102-116. DOI: 10.1037/a0038889 |
|
|
| Back to top |
|
|
