search

jflournoy / riclpmr

Generate lavaan syntax for RI-CLPM
https://jflournoy.github.io/riclpmr/
Other
11 stars 1 forks source link

Pooling results from multiple models #2

Open sccollitt opened 3 years ago

sccollitt commented 3 years ago

Hi John, This isn't an issue per se, so I apologize if this isn't the right place. I have multiple imputed datasets that I would like to run riclpm on and then pool the results. However I'm having some issues with this.

The Semtools package provides a handy way to pool lavaan model results from multiply imputed datasets, but I can't find a good way to do this with your riclpm function. I have generated a list of model results, 1 for each imputed dataframe (and the original), but don't have a good idea as to how to pool these results - using Rubin's rules - into some sensible output. Please let me know if you have any advice on this. Thank you for this package. It's saved me from many opportunities for user error by explicitly modeling with lavaan.

Sam

jflournoy commented 3 years ago

@sccollitt I'm not very familiar with the nuts and bolts of multiple imputation. If you have time, perhaps you could give a more detailed description of exactly what you need from each model. For example, do you need the coefficients and standard errors for all effects of interest across models in a single data frame? What does Semtools do exactly?